An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

John Nicot; Ludivine Fadel; Thierry Taris

doi:10.1155/2021/6642938

An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

Nicot, John;Fadel, Ludivine;Taris, Thierry 2021-08-03 00:00:00 Hindawi Wireless Power Transfer Volume 2021, Article ID 6642938, 25 pages https://doi.org/10.1155/2021/6642938 Research Article An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting 1,2,3 1,2,3 1,2,3 John Nicot , Ludivine Fadel , and Thierry Taris University of Bordeaux, IMS, UMR 5218, 33405 Talence, Bordeaux, France CNRS, IMS, UMR 5218, 33405 Talence, Bordeaux, France Bordeaux INP, IMS, UMR 5218, 33405 Talence, Bordeaux, France Correspondence should be addressed to John Nicot; john.nicot@ims-bordeaux.fr Received 19 October 2020; Revised 2 June 2021; Accepted 10 July 2021; Published 3 August 2021 Academic Editor: Jiafeng Zhou Copyright © 2021 John Nicot et al. )is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. )e widespread deployment of the Internet of )ings (IoT) requires the development of new embedded systems, which will provide a diverse array of diﬀerent intelligent functionalities. However, these devices must also meet environmental, maintenance, and longevity constraints, while maintaining extremely low-power consumption. In this work, a batteryless, low-power con- sumption, compact embedded system for IoT applications is presented. )is system is capable of using a combination of hybrid solar and radiofrequency power sources and operates in the 900 MHz ISM band. It is capable of receiving OOK or ASK modulated data and measuring environmental data and can transmit information back to the requester using GFSK modulated data. )e total consumption of the system during its sleep state is 920 nW. Minimum power required to operate is −15.1 dBm or 70 lux, when using only radiofrequency or solar powering, respectively. )e system is fully designed with components oﬀ the shelf (COTS). It is important to diﬀerentiate three types of RF power 1. Introduction harvesting: opportunistic power harvesting, which is usually With the ever-increasing growth of wireless sensors and unreliable, as power levels can vary greatly depending on the their associated networks, in the context of the deployment environment and are generally low [4]; dedicated power of the IoT, it is of great importance for such devices to harvesting, which can provide high levels of power at the minimize their power consumption. On top of this con- cost of specialized hardware and, possibly, proprietary straint, modern sensor nodes not only must maintain low- communication and power transmission protocols [5]; or cost deployment, but also are expected to minimize their semidedicated power harvesting, which can involve diverse environmental impact. )erefore, it is crucial to explore new methods of powering, including piggy-backing over existing methods and techniques in order to power and use these standards or devices currently in place (possibly with minor nodes. )e rise of diverse energy harvesting methods (solar modiﬁcations to implement communications) or remotely powering the device at regular intervals, in order to keep the energy, thermoelectric energy, and vibrational energy) makes it possible to use the environment to power these IoT sensor node operational [6]. nodes and does not require them to be permanently con- Solar energy, on the other hand, depending on the size of nected to the classical electrical grid. the solar panel and the ambient luminosity levels, can easily )e range of power consumption, for most IoT sensor provide several milliwatts of power in an outdoor conﬁg- nodes, varies between several tens to hundreds of micro- uration down to several tens of microwatts of power in watts, depending on their range of operation and func- indoor environments [7]. tionality [1, 2]. Radiofrequency (RF) powering can be used in )is work investigates a hybrid-powered solar and RF order to provide up to a milliwatt at close range or several harvesting wireless sensor node, on top of which bidirec- microwatts at further distances [3]. tional communications are implemented. In order to 2 Wireless Power Transfer minimize the power consumption of the node, data re- using the Greinacher topology, Figure 3, in order to achieve ception is implemented under the form of an ASK or OOK large voltage ampliﬁcation during RF to DC conversion. Schottky diodes are preferred due to their low forward modulated, low-power wake-up radio, and data transmis- sion is ensured by using a specialized, low-power transceiver voltage—typically 80 mV for HSMS-285C diodes [19]. using GFSK modulation, for the transmission of data col- )e highly capacitive nature of diode-based rectiﬁer lected by the node. circuits makes them diﬃcult to match at radiofrequencies. It )is work is an extension of previous work [8]. )e is important that the matching network exhibits a high- extensions implemented include the addition of hybrid quality factor, Q, in order to maximize the voltage at the solar/radiofrequency power harvesting, data retransmission input of the diodes, thus minimizing the losses due to their capability from the node, several additional scenarios of forward voltage [20]. operation, and an overview of a simulator used to model the )e S parameters of the n-stage rectiﬁers are shown in aforementioned scenarios. Figure 4, before matching, from 700 MHz to 1.1 GHz, to )is work is structured into three parts. Firstly, the node emphasize the high capacitance present on the circuit’s is presented, and each module composing it is visited, in input. order to demonstrate how each node module is designed. )e matching networks are conﬁgured with an inductor- )e modules are further characterized, in order to analyze based, double “L” network topology, Figure 5. their performance in terms of metrics. Secondly, an overview )e values for 50 Ω matching were simulated using ADS, of power consumption of the node in diﬀerent states of and further adjustments were done in situ through an it- operation is presented, in order to characterize power erative, dichotomous approach. )e values determined for consumption. )is is used in order to develop a simple the rectiﬁers are presented in Table 1. simulator which can predetermine the performance of the )is ensures an input return loss below −10 dB over node in a given scenario of operation. )irdly, these diﬀerent more than 30 MHz of bandwidth at 935 MHz at lower power scenarios of operation are presented, and the performance of levels (−20 to −35 dBm), Figure 6. the node is compared between data generated through the Due to varying diode capacitance as a function of power simulator and in situ measurements. Finally, this work input P , the center frequency of the matched rectiﬁers IN concludes the presented work and compares it to other work tends to shift, though not signiﬁcantly. present in this ﬁeld. As illustrated in Figure 5, the centered frequency of the input return loss shifts with the input power, due to the dependence of the diode capacitance on the latter. )is 2. Presentation of the Node variation remains negligible regarding the achievable bandwidth over the considered range of power and targeted )e node is made exclusively from COTS. )e architecture application. Indeed, for low-power IoT, the broadcasted data of the node is presented in Figure 1; it includes a rectenna, is sent at low bitrates, requiring narrow bandwidth, and based on HSMS-285C Schottky diodes [9], to collect both powering methods via RF often use Continuous Wave (CW) harvested RF power and modulated data to control the node carriers [20]. (operating at 935 MHz); a generic solar panel; a bq25570 Rectiﬁer performance is evaluated by two metrics: the power harvesting IC to manage RF energy [10]; a bq25504 sensitivity, S (1), which is the unloaded voltage output, and power harvesting IC to manage solar energy [11]; ceramic the eﬃciency, ƞ (2), which is the ratio of the output DC capacitors as storage elements (multiples of 4V, 470 μF power (P ) to the input power (P ), depending on the AMK432BJ477MM-T capacitors for RF energy [12] and a OUT IN load R : 470 mF supercapacitor for solar energy [13]); a PIC16LF1559 L microcontroller for functionality [14]; an MR45V100 FRAM S � V P 􏼁 , (1) O IN for storage [15]; a TS811 comparator for demodulation of received data and the wake-up circuit [16]; and a SPIRIT1 transceiver for data transmission, operating at 869 MHz [17]. P R 􏼁 V OUT L O (2) η � � . RF power and instructions sent to the node are provided P R · P IN L IN through a B200 Software-Deﬁned Radio (SDR) and am- Increasing the number of stages improves the sensi- pliﬁed by a ZHL-42W ampliﬁer in order to provide power tivity at low levels of power, Figure 7, but degrades the outputs ranging from 0 to 28 dBm at the antenna. )is is rectiﬁers’ eﬃciency accordingly, Figures 7(a) and 7(b). typical for semidedicated RF power harvesting systems. )e Interestingly, the increase of cascaded stages makes the system is operated though a custom-developed software eﬃciency increasingly independent of the load value R , package, operating under MATLAB. L Figure 8. Similar work has yielded the same conclusions [21]. Based on this analysis, we will use 4 stages to receive 2.1. Rectenna. Harvesting of power and reception of data data and only one stage to harvest RF energy. Four stages through radio-waves require antennas. A multibranch di- were chosen for data reception in order to permit the node pole antenna, Figure 2, is developed [18]. It exhibits a return to receive data at a suﬃcient distance (up to 20 meters), loss of −15 dB and a directivity of 2 dBi at 935 MHz. while avoiding overloading the node’s reception module, as Due to the low-power levels of collected RF signals, the the sensitivity (and therefore outputted voltage) increases rectiﬁcation is based on cascaded stages of voltage doublers, with the number of stages. Wireless Power Transfer 3 Solar energy Energy Power 935 MHz storage mgmt. Rectification Energy Power storage mgmt. Power + data TX (OOK/ASK) Demod. + Processing wake-up Microprocessor Sensor module(s) 869 MHz Storage Feedback (FRAM) module Data RX (GFSK) Power path Data path Figure 1: Overview of the IoT sensor node’s diﬀerent modules, including external modules used for data transmission and reception. –5 –10 –15 –20 900 920 940 960 980 Frequency (MHz) (a) (b) Figure 2: Details of the rectenna developed for data reception and power harvesting. (a) Antenna connected to the rectiﬁer, with a euro coin for reference. (b) Antenna S parameters from 885 MHz to 985 MHz. out Matching network Stage 1 Stage 2 Stage N Figure 3: Full-wave, N-stage rectiﬁer, Greinacher topology used. Amplitude (dB) 4 Wireless Power Transfer Table 1: Component values chosen for the matching network +j1 between the antenna and the rectiﬁer. +j2 +j0.5 Stage (s) Value (nH) 1 2 3 4 L1 13 5 18 Short +j5 +j0.2 L2 6.2 24 68 Open L3 10 24 12 27 L4 91 13 9.1 7.5 +j30 0.0 ∞ –j30 )e unloaded maximum voltage (open voltage) and maximum DC output power (available power) under full sunlight (L> 2000 lux) are 3.5 V and 1 mW for the small –j0.2 –j5 panel and 4 V and 2.5 mW for the square panel. In an indoor environment featuring windows (600< L< 1500 lux), the open voltage ranges from 2.8 V to more than 3 V, and the power is between 500 μW and 1.5 mW for the square panel –j0.5 –j2 and between 250 μW and 600 μW for the small panel. In an –j1 oﬃce with only artiﬁcial light (L< 500 lux), the square panel still yields more than 50 μW and has an open voltage of more 1 stage than 2 V. Yellow LEDs provide a larger open voltage and a 4 stages higher output power, as they are more spectrally rich than Figure 4: )e S parameters of the 1-stage and 4-stage rectiﬁers, white LEDs [22]; however, the diﬀerence is not signiﬁcant. from 700 MHz to 1.1 GHz, without a matching network, using Typical illumination levels were sourced from [23]. Z � 50Ω. 2.3. Power Management RF RF IN OUT L3 L1 2.3.1. Rectiﬁers. Rectiﬁcation eﬃciency is dependent on L2 L4 load, and therefore it is not appropriate to directly supply digital circuitry, which may have diﬀerent power con- sumption proﬁles, when the rectiﬁer operates. )is issue is remedied through the use of Maximum Power Point Tracking (MPPT), which loads the rectiﬁer appropriately, Figure 5: )e “double-L” matching network topology used to permitting an extraction of the most amount of power adapt the rectiﬁers. possible at any given time [24]. With the use of a boost converter, MPPT can be used to store energy into a ce- ramic capacitor-based storage element, thus providing adequate power when required. Ceramic-based capacitors Figure 9 shows the amount of available output DC are used for rectiﬁer energy harvesting, due to their ex- power, P , of a 1-stage rectiﬁer, as a function of the input OUT tremely low leakage proﬁle, in consideration of the power, P , with an optimal load applied for all power levels. IN Available power ranges from approximatively 200 nW at amount of energy that can be harvested from RF energy sources [25]. −30 dBm up to 450 μW at 0 dBm. To deliver 1 μW, a single stage rectiﬁer requires an input power level of −21 dBm. A bq25570 power harvesting IC was chosen for this purpose, as it provides MPPT, a boost converter, and a buck converter with a variable output voltage and has a low 2.2. Solar Panels. Solar energy is used as an alternative quiescent current. Figure 13 shows the general operation and source of energy to complement RF harvesting when it is wiring diagram used for this work with the bq25570. possible. Solar harvesting signiﬁcantly increases the amount )e bq25570 has several diﬀerent modes of operation. If the main storage element (C of energy available, enabling the use of an RF transmitter to ) is depleted, it needs to BIGSTOR provide data back to the requester. be charged to a certain threshold (V ) before operation CHGEN Two solar panels, a small one with circle shape and a of the MPPT and the boost converter, through the use of a large one with square shape, have been considered for the “cold-start” charge pump, with low eﬃciency (approx- system, Figures 10(a) and 10(b), respectively. )e opera- imatively 5%). )e storage element is then charged to an tional areas of these panels are, respectively, 13.9 cm and acceptable voltage level, in order to allow the main parts of 30.5 cm . )e open voltage, Figure 11, and available out DC the IC to operate. In order to avoid rapidly regressing into a power, Figure 12, of the solar panels are characterized in a cold-start state (especially due to the activation of the load), controlled environment using white and yellow LEDs as once the desired voltage is reached, a storage element illumination sources, in order to simulate realistic illumi- threshold indicator (with hysteresis), V , is provided by BATOK nation conditions for the scenarios of application. the IC. It is used in this work in order to turn on or oﬀ the 0.2 0.5 30 Wireless Power Transfer 5 0 0 –5 –5 –10 –10 –15 –15 –20 –20 –25 –25 –30 –30 –35 –35 880 900 920 940 960 980 880 900 920 940 960 980 Frequency (MHz) Frequency (MHz) –15 dBm –15 dBm –25 dBm –25 dBm –35 dBm –35 dBm (a) (b) Figure 6: )e S parameters of the 1-stage (a) and 4-stage (b) rectiﬁers, from 870 MHz to 970 MHz, matched, using Z � 50Ω, for varying 11 0 input power levels. –30 –25 –20 –15 –10 –5 0 Input power (dBm) 1 stage 3 stages 2 stages 4 stages Figure 7: Measured rectiﬁcation sensitivity S versus input power P , for power levels ranging from −30 dBm to 0 dBm. IN supply to the digital section of the node if power is scarce. cold-start state should there be an interruption in the RF )e minimum storage voltage required to power on the energy source, in order to preserve an entry point into the system is 2.5 V, and the hysteresis shutoﬀ is deﬁned at 2.2 V. high eﬃciency boost mode, when power is restored. )e values of the resistors to program the device are chosen In order to perform MPPT, the bq25570 periodically using [26]. samples the sensitivity, S, of the 1-stage rectiﬁer through the Although V is approximatively 1.7 V, the values R /R divider bridge and applies a load according to the CHGEN OC1 OC2 for the hysteresis circuit were programmed higher, to take voltage present between R and R (S ). )e OC1 OC2 MPPT into account the possibility of the absence of power and to measured eﬃciency, reported in Figure 7, depending on the insert a security margin to avoid regression into a “cold- optimal loads from −30 dBm to 0 dBm, was determined for start” state. )e minimum storage voltage required to power MPPT. From these loads, the optimum voltage levels that on the system was set at 2.5 V, and the hysteresis shutoﬀ set should be delivered to the bq25570 through the voltage at 2.2 V. Although, theoretically, these values could both be divider are calculated. )is is presented as the ratio of S / MPPT equal to V to ensure maximum energy usage if power S in Figure 14. It can be seen that a value of around 35% CHGEN should become unavailable, the limits were chosen to be presents optimal power transfer, especially at lower input higher than the former in order to avoid a regression into a power; therefore, R and R were set accordingly. OC1 OC2 Amplitude (dB) Sensitivity (V) Amplitude (dB) 6 Wireless Power Transfer 4 4 ×10 ×10 5.5 5.5 5 5 4.5 4.5 4 25 4 3.5 3.5 3 3 2.5 2.5 2 2 1.5 1.5 1 1 0.5 0.5 –30 –25 –20 –15 –10 –5 0 –30 –25 –20 –15 –10 –5 0 Power input (dBm) Power input (dBm) (a) (b) Figure 8: Measured rectiﬁcation eﬃciencyƞ versus input power P and load. (a) )e eﬃciency ƞ for the 1-stage rectiﬁer. (b) )e eﬃciency IN ƞ for the 4-stage rectiﬁer. –2 –30 –25 –20 –15 –10 –5 0 Power input (dBm) Figure 9: Measured power output P versus input power P , for the 1-stage rectiﬁer, for input power levels ranging from −30 dBm to OUT IN 0 dBm. 3.2cm 5cm 4cm 5.5cm (a) (b) Figure 10: )e solar panels with a 1-euro coin for reference. (a) )e small solar panel. (b) )e large solar panel. 2.3.2. Solar Panels. Similarly to rectiﬁers, solar panels also bq25570, with the exception that it does not have an inte- require the use of MPPT in order for the maximum available grated buck converter and therefore cannot provide a reg- power to be extracted from them. For this purpose, a ulated output voltage to a load. Figure 15 shows the general bq25504 power harvesting IC was used. It is similar to the operation and wiring diagram used for this work with the Load (ohms) Power out (μW) Efficiency (%) Load (ohms) Wireless Power Transfer 7 3.5 2.5 1.5 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 11: Measured open voltage of the two solar panels, depending on both the luminosity available and the type of LED lighting. Supermarkets commercial areas Offices Full daylight Homes and above Dark Overcast day areas 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 12: Measured power output, P , of the two solar panels, depending on both the luminosity available and the type of LED lighting, OUT with illustrations of diﬀerent luminosity levels, L. bq25504. )e storage element chosen for solar power har- )e same voltage thresholds and hysteresis values were vesting is a 470 mF supercapacitor, which can store con- used as in the previous subsection. Likewise, the MPPT ratio siderably larger amounts of energy than ceramic-based S /S can also be inferred from measured data, Figure 16. MPPT capacitors. However, these elements do have a higher It can be seen that a value of around 70% presents optimal leakage current [13, 27]. power transfer, with a variation of around plus or minus 5%, In order to combine both RF and solar energy together, a depending on the luminosity and solar panel type. BAT17-04 Schottky diode [28] was introduced between the supercapacitor serving as a storage element for the bq25504 and the ceramic capacitor array serving as a storage element 2.4. Energy Storage for the bq25570. )e use of a Schottky diode permits a completely passive 2.4.1. Rectiﬁers. )e bq25570 has two main modes of solution permitting energy to be transferred from the solar charging operation: “cold-start” (CS) mode, where a charge energy circuit to the RF energy circuit, while avoiding a pump with low eﬃciency charges up the main storage el- signiﬁcant reverse drain of energy from the RF energy circuit ement to about 1.7 V (VCHGEN), and a “warm-start” (WS) if no solar power is available. )e low threshold voltage of this mode, where the MPPT and boost converter are active and diode also makes it possible to avoid signiﬁcant energy loss charge the main storage element up to a conﬁgured voltage during energy transfer. Similar techniques have been used in (4.0 V) [10]. )e combination of both the CS and WS modes other work with combined RF/solar harvesters [29, 30]. has been deﬁned as “full-start” (FS) mode. Available power (μW) Sensitivity (V) 8 Wireless Power Transfer MPPT To storage C C STOR BYP configuration element R R OC1 OC2 V V V OC SAMP STOR BAT V V IN BOOST From solar panel C L REF SAMP IN L 2 V BUCK OUT Boost controller OUT REF To node V C SS OUT Buck VCC supply controller MPPT SS IN DC Cold start Nano-power management nEN OUT EN To μC V bq25570 BAT_OK RF harvesting R R OK3 OV2 OUT2 On/off hysteresis buck output voltage and battery level OK2 configuration OV1 OUT1 OK1 Figure 13: Schematic describing the bq25570 as used in this work, used for RF power harvesting and voltage regulation for the load (adapted from [10]). –30 –25 –20 –15 –10 –5 0 Power input (dBm) Figure 14: Optimum MPPT calculation, in order to size R and R to extract maximum power from the 1-stage rectiﬁer. OC1 OC2 CS mode requires a minimum of 15 μW to be able to when the storage element reaches 2.5 V. Discharge can be charge the storage element to VCHGEN, leakages not- seen from 650 s to about 1300 s, during which the bq25570 withstanding. From Figure 9, it can be determined that a provides power to the digital section of the node until reaching the shutoﬀ voltage of 2.2 V. minimum RF power input of −13.1 dBm is required to bootstrap the system if the main RF energy storage element )e amount of time taken to fully charge the system from is empty. A typical charge and discharge curve of the system an empty ceramic capacitor storage element of 470 μF was during normal operation was measured, Figure 17, without measured as a function of RF power input in Figure 18. No active communications. )e digital section of the node’s communications were active during this time. output is regulated to 1.8 V. Charge times in CS and WS modes increased quasi- )e CS phase of the bq25570 can be seen from 0 s to linearly with the size of the ceramic capacitor storage ele- 500 s, and WS phase can be seen from 500 s to 620 s, with the ments. Discharge times were also measured and determined output to the digital section of the node turning on at 550 s, to be 425 s, 971 s, 1810 s, and 3505 s, respectively, for the % Open voltage (%) RDIV BAT OV OK PROG OK HYST OUT SET Wireless Power Transfer 9 MPPT C C To storage STOR BYP configuration element OC1 OC2 V V OC SAMP STOR V L IN BOOST From solar panel C REF SAMP IN Boost controller REF SS MPPT AV IN DC SS Cold start Nano-power BAT_OK To μC management OT PROG bq25504 solar harvesting R OK3 UV2 OV2 On/off hysteresis and R OK2 battery level OV1 configuration UV1 OK1 Figure 15: Schematic describing the bq25504 as used in this work, used for solar power harvesting (adapted from [11]). 0 500 1000 1500 2000 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 16: Optimum MPPT calculation, in order to size R and R for the extraction of maximum power from the solar panels. OC1 OC2 2 2 470 μF, 940 μF, 1860 μF, and 3720 μF ceramic capacitor- C 􏼐V − V 􏼑 BIGSTOR INIT FINAL (3) based storage elements. t � , 2P CONSUMPTION )e quasi-linearity of the charge and discharge times is logical, as the device is operating at a constant power dis- charge. )is is because, during sleep mode, the node exhibits where t is the amount of time taken to discharge capacitor static, unvarying power consumption, and the bq25570’s C , between the initial and ﬁnal voltages BIGSTOR quiescent current varies little with voltage, so charge and V and V , respectively. P is the power CONSUMPTION INIT FINAL discharge times will tend to obey: consumption that is applied to the capacitor. Percent of open voltage (%) RDIV BAT OV OK PROG OK HYST BAT UV 10 Wireless Power Transfer 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Time (s) Storage voltage Digital section voltage Figure 17: bq25570 charge and discharge cycle, with a 470 μF ceramic capacitor storage element, and P � −10 dBm during charging. IN 2.4.2. Solar Panels. )e bq25504 also has two modes of several days, minimizing the leakage current, assuming no operation, the CS and WS modes, similarly to the bq25570. communications were present. Like the bq25504, CS mode on the bq25570 requires at least 15 μW to be able to charge the storage element in CS mode. However, when using a supercapacitor, the leakage current 2.5. Communications cannot be neglected [27]. 2.5.1. Node Reception. As this system is designed to con- )e initial but worst-case leakage current of such devices can be as high as 100 μA (whether charged or discharged), sume as little power as possible, the use of a standard RF and therefore the CS requires approximatively 20 μW of receiver is not feasible due to the high-power consumption power in order to be able to charge the bq25504 to WS mode of such devices which requires at least a couple of mW to [11]. )e leakage current, though initially high, eventually operate [31]. )erefore, the node is always in deep sleep stabilizes to less than 1 μA after several days. )is can be mode when not being solicited, and reception (RX) com- observed through the rate at which the voltage drops, munications are ﬁrst processed using a 4-stage rectiﬁer Figure 19. circuit to perform the envelop detection of an incoming )is exhibits the worst-case scenario of self-discharge, as OOK, or ASK, modulated RF signal at 935 MHz. it was charged instantly, using a standard power supply for )e wake-up radio demodulator circuit is presented in approximatively 5 minutes. It should be noted that this Figure 21. It provides a good trade-oﬀ between robustness and low-power consumption to demodulate data while leakage current does diminish after several days down to about 1 μA to 2 μA, due to the charging characteristics of minimizing false wakeups. Indeed, amplitude modulation such devices [27]. schemes are prone to large and abrupt variations of the A charge and discharge cycle of the supercapacitor dynamic signal due the environment of propagation, which through solar energy was measured over a period of 100 contribute to corruption of the demodulation process. hours, Figure 20, using the large solar panel, where the )e main component used for data demodulation is a environmental luminosity varied between 100 lux during TS881 comparator, chosen for its low quiescent current. V IN nighttime hours and 1000 lux during daylight hours. is the output of the 4-stage rectiﬁer. R and C form a data 2 3 It can be seen that the initial charge time in CS mode is very slicer (low pass ﬁlter), which is used to demodulate incoming slow, due to the ineﬃciency of the charge pump in the power Manchester-encoded data. )e RC time constant was chosen management IC, but is still possible, even if only 100 lux is to be 5 times the baud rate in order to provide suﬃcient permanently present. Once the WS mode is active, the charging settling time while remaining adaptable to varying power of the supercapacitor becomes signiﬁcantly faster, due to the levels. activation of MPPT, which has a much higher eﬃciency C and C , associated with R /R /R and R /R /R , re- 1 2 1 3 4 2 5 6 (measured at least 70%). From Figure 20, it can be noted that, as spectively, form high-pass ﬁlters in order to avoid false data long as there is enough luminous energy to charge the capacitor detection which may arise from slow power level variations. to at least its previous voltage before the next luminous cycle )e “+” and “− ” input nodes of the comparator are polarized (i.e., work hours or daytime), the operation of the node may be to half the supply voltage through R /R and R /R , thus 3 4 5 6 maintained indeﬁnitely in warm-start mode, which can permit ensuring proper operation of the comparator. More spe- the use of the node in low-luminosity environments due to the ciﬁcally, the “+” node is slightly biased below the “−” node to increased eﬃciency of WS mode. avoid chattering, at the cost of decreased sensitivity. Measurements indicate that the minimum amount of In the case of ASK modulation, data is sent along with luminosity required to keep the node functional without power, whereas OOK modulation implies that the system is discharge is 70 lux, after the supercapacitor was charged for not powered during communications. RX communications Voltage (V) Wireless Power Transfer 11 –14 –12 –10 –8 –6 –4 –2 0 Rectiﬁer input power (dBm) Cold start Warm start Full start Figure 18: Time taken for the bq25570 to accomplish cold, warm, and full start using RF energy only, as a function of input power. 02 468 10 12 Time (days) Figure 19: Self-discharge curve of the 470 mF supercapacitor over several days, showing the voltage present. 470 mF storage - charge and discharge Daylight hours 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Time (days) Storage voltage Digital section voltage Figure 20: bq25504 charge and discharge cycle, with a 470 mF supercapacitor storage element, with the luminosity varying between 100 lux and 1000 lux. are done at a speed of 800 bauds, but the algorithm in the it avoids long sequences of binary zeroes and ones which can microcontroller can be adapted to run from 40 bauds to 25.6 induce demodulation errors in the data slicer. It does, kilobauds. however, increase the spectrum usage for an equivalent )e frame format of messages sent to the node was number of sent bits. implemented using Manchester encoding, with a synchro- )e synchronization subframe length, s, was chosen to nization preamble, Figure 22. )is form of encoding pro- be 16 bits, in order to allow for proper wake-up generation vides several advantages. First, the synchronization and baud rate detection. )e actual message subframe preamble can be used to lock onto the baud rate with a length, n, is two bytes long (m � 16 bits). )e ﬁrst byte of the certain error tolerance (clock reconstruction). Furthermore, message subframe is the address of the node, which permits Time (s) Voltage (V) Voltage (V) 12 Wireless Power Transfer VCC VCC the basic packet format was used in order to minimize overhead on the microcontroller. )e frequency at which the C R 1 1 data was emitted is 868 MHz. )e frequency deviation for Rectiﬁer Data the modulation is 19 kHz. )e output power level chosen for CC output data transmission is −30 dBm. A transmission of the basic C R packet was collected via SDR and demodulated in order to C D 2 2 4 2 WUP show the digital data sent, using the default settings of the R D C R 3 transmitter, Figure 25. 6 1 5 7 )e packet is split into 4 main parts: the preamble, which is 64 bits long; the sync word, which is 32 bits long; the Figure 21: Comparator circuit for data reception; wake-up signal message, which is 3 bytes (or 24 bits) long; and the CRC generation. checksum, which is 1 byte (or 8 bits) long. )e packet format can be modiﬁed through registers present in the module, and additional ﬁelds can be added, depending on the usage an addressing of up to 256 nodes, and the second byte is a scenario, which includes addressing capabilities, multi- command, subdivided into two nibbles. )e ﬁrst nibble packet, or variable-length messages [17]. )e demodulation indicates the operation, and the second contains an optional argument for the microcontroller to process. )e frame is of the received packets is performed on the same device transmitting data to the node. )e supplied software is used therefore 33 bits long. )e RF signal is ﬁrst rectiﬁed by the 4-stage rectiﬁer, to receive data packets. An example of packet reception, containing temperature data as the third byte, is shown in Figure 23, where it is then available for processing by the comparator and wake-up circuit. )e wake-up detector has Figure 26. been adapted from [32]. It is designed to allow data rates above a certain fre- 2.6. Functionality. In order to use harvested energy and to quency to activate the wake-up circuit, which makes it manage the concurrent operation of diﬀerent modules, the robust against low-frequency, brutal changes in power levels data demodulation capabilities and intelligent functions of or against external interference from other communications this sensor node are provided by a PIC16LF1559 micro- in the same band, which may be interpreted by the com- controller, with an integrated temperature sensor [14]. parator as data. Storage capabilities are implemented with an FRAM IC, )e output of the comparator is a normalized repre- using the SPI bus [15]. sentation of received binary data to V (1.8 V). )is is used CC )e PIC spends most of its time in sleeping mode and in two ways: to generate a wake-up signal with C , D , D , C , 4 1 2 5 only wakes up when an event is received from the com- and R (WUP) and also give data to the microcontroller parator/data slicer module. )e Manchester decoding al- (DATA) to be processed when the wake-up signal is gen- gorithm is implemented in a purely software-based manner, erated, Figure 24. using interrupts and timer loops. )e algorithm can be )e sensitivity of the circuit was measured, and the node adapted for diﬀerent synchronization formats, message messages could still be properly processed by the micro- sizes, and data transmission speeds and is presented in controller with power input down to −31 dBm at the input of Figure 27. Once a message is received, it is then processed the 4-stage rectiﬁer circuit. and, if successful, an action is performed, as illustrated in Figure 28. Table 2 presents a brief overview of currently imple- 2.5.2. Node Transmission. Transmission (TX) communica- mented commands that can be sent to the node with the tions from the node are done using a low-power radio basic, 2-byte frame format. )e PIC only processes these transmitter IC, the SPIRIT1. )is IC is capable of sending commands if the ﬁrst byte of the frame matches its pro- data with diverse modulation formats, including OOK, ASK, grammed address. FSK, and GFSK. Power levels used to transmit the data can be conﬁgured from 11 dBm down to −30 dBm. )e baud rate of data transmission can also be conﬁgured from 1 kbps up to 500 kbps. 2.7. Assembled System. )e diﬀerent modules featured in the node were independently manufactured on 2- or 4-layer FR- It can use either direct modulation or a standardized packet format. Using the standardized packet format, data 4 substrate and soldered. )e node, Figure 29, is further transmission integrity can be veriﬁed using a Cyclic Re- enhanced by using modular signal connectors. dundancy Check (CRC). )e Advanced Encryption Stan- )is modular, multiboard approach allows the experi- dard (AES) can also be used in order to secure data mentation with various assemblies of diﬀerent subcircuit transmission. )e module also includes a Carrier Sense generations, in order to evaluate their performance inde- Multiple Access (CSMA) algorithm, which can be exploited pendently. )e presence of connectors also facilitates easy determination of power consumption by inserting mea- to detect if another module or device is already transmitting, thus avoiding packet collisions [17]. surement devices between the pins and permits easy debugging of various modules with measurement For the purposes of this work, 2-GFSK modulation was selected in order to transmit data, at a rate of 38.4 kbps, and instrumentation. Wireless Power Transfer 13 Byte 1 Byte 2 Byte n Synchronization Sync stop Data frame (s bits) bit (m bits) (1 bit) Figure 22: )e frame format for the node’s low-power wake-up radio, used in RX communications. 1.5 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) Figure 23: Data frame sent to node, after rectiﬁcation. 1.5 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) DATA WUP Figure 24: Data frame normalization to 1.8 V and wake-up signal generation for the microcontroller. –1 –2 –3 Sync (32 bits) CRC (8 bits) –4 Message (24 bits) Preamble (64 bits) –5 0 500 1000 1500 2000 2500 3000 Time (μs) Figure 25: Standard “basic” data frame from the low-power TX module. Voltage (V) Voltage Frequency deviation (abitrary) 14 Wireless Power Transfer Figure 26: Reception of data from the node. )e third byte is ambient temperature as measured by the microcontroller, in hexadecimal ° ° format, between 25 C and 27 C in this case. )e ﬁrst two bytes are “T?” in ASCII. WUP interrupt Abort Process RX generated RX message Receive sync bits Yes Required Required No Timeout Timeout No message sync bits bits received? received? Yes Average sync bit time for clock reconstruction Receive message bits using reconstructed Wait for sync stop clock time bit No Timeout Sync stop bit? Yes Figure 27: Software algorithm used in order to decode the Manchester-encoded message received by the node. Yes Perform ADC Get Initialize PIC and conversion and temp.? peripherals store in RAM No Output Yes Sleep Say temperature in temp.? RAM on test point No Wake- No up? Yes Store Store temperature Yes temp.? in FRAM Receive data frame No Yes Retrieve Retrieve temperature from No Correct temp.? FRAM address? No Yes Yes Send temperature TX? Process command via TX module No Figure 28: Software algorithm used after reception of a correct message, showing a sample of available commands. Wireless Power Transfer 15 Table 2: Commands available within the node. Byte Description of command instructions 0 × F0 Sample the temperature of the node 0 × F1 Show the temperature on a test point 0 × F2 Turn on the FRAM power supply; make the FRAM sleep 0 × F3 Turn oﬀ the FRAM power supply 0 × F4 Turn on the comparator power supply 0 × F5 Turn oﬀ the comparator power supply 0 × F6 Show the command on a test point 0 × F7 Reset the temperature in the node to 25 C 0 × F8 Send the node temperature via the TX module 0 × FB Carry out sanity check of FRAM and TX modules (verify device IDs) 0 × Dn Store the temperature in FRAM address n 0 × Cn Retrieve the temperature in FRAM address n 1,2 0 × 3n Set the high nibble of the temperature byte in RAM with n 1,2 0 × 4n Set the low nibble of the temperature byte in RAM with n 1 2 )ese commands are used for debugging purposes. For these commands, n is a variable, ranging from 0 × 0 to 0 × F. Large solar panel Antenna (data RX) Microcontroller and FRAM Comparator Power management wake-up (stacked) Power and data Energy storage rectiﬁers capacitors Antenna (power harvesting) Antenna (data TX) RF transceiver Figure 29: )e whole system, with all modules present. A 1-euro coin is provided for reference. the voltage delivered to the digital subsection and is con- 3. Power Consumption Analysis ﬁgured at 1.8 V. A simulator that is constructed from )e estimated power consumption of the overall system is measured data is then presented. necessary in order to determine the viability of applicative scenarios of usage of the node. Indeed, it may be necessary to limit certain power consuming features of the node (notably 3.1. Analog Modules the broadcasting of measured data) when limited power is 3.1.1. Comparator. )e comparator/wake-up module is the available. Typically, when RF harvesting is the only source of energy, the system must operate in a restricted mode from a device that consumes the most power in sleep. )is is power consumption perspective. primarily for two reasons. Firstly, in order to polarize the )is section fully characterizes the power consumption inputs of the comparator to (V /2), after the high-pass CC of the diﬀerent modules of the node, depending on their ﬁlter, Figure 20, the resistor network uses approximatively phase of operation. )e modules are categorized as digital 180 nA, or 324 nW. Secondly, the comparator itself uses and analog modules, as digital modules have more phases of 220 nA, or 396 nW. Additional current used, when the operation than analog modules, with the latter having more WUP and DATA pins change due to the reception of data, predictable and static power consumption over time. VCC is was measured to be about 10 nA to 20 nA at most, or 16 Wireless Power Transfer a message for 40 ms, it consumes 16.6 μJ to receive 33 bits (a 18–36 nW. )e module therefore consumes, at most, about 850 nW of power, continuously. message of 16 bits and a synchronization subframe of 17 bits), which is approximatively 503 nJ per bit received. 3.1.2. Power Management ICs. )e bq25504 and bq25570 must also be considered when doing a power consumption 3.2.3. TX Module. )e data TX module must be operated in analysis. During CS mode, the necessary minimum amount a particular order so as to ensure data transmission. It must of power required to start up both devices is 15 μW, below be turned on, reset, and then set into its ready operating which the device will never manage to enter WS mode, due state. Afterwards, the Voltage Controlled Oscillator (VCO) to the ineﬃciency of the charge pump, which is approx- operating within the integrated PLL must be calibrated, after imatively 5%. which it must also be returned to a ready operating state. Once in WS mode, the power consumption of the de- After this initialization phase, data may be sent to the vices varies relatively little, and it was measured, with no transmitter buﬀer, and the device may send data. At the end power on the input, to be around 500 nA, independently of of the transmission, it returns into ready state and may then voltage present on the storage elements. Depending on be shut down. storage element voltage, this can range from a power con- )e device in shutdown mode consumes 2 nA of current, sumption of 2 μW down to 850 nW. )e eﬃciency of the or 2.6 nW. In ready mode, it consumes 430 μA, or 774 μW. buck converter, Figure 30, is measured with the storage When calibrating its VCO, it consumes 6.72 mA, or element fully charged, using the methodology provided in its 12.33 μW. Finally, when transmitting data, it consumes datasheet [10]. 9.58 mA, or 17.56 mW. Buck converter eﬃciency exceeded 80% for load currents As the module sends messages at 38.4 kbps and con- above 10 μA. However, for currents between 1 μA and 10 μA, sumes 17.56 mW while sending a message for 3.75 ms, it eﬃciency dropped down to 60%, and for load currents below consumes approximatively 65.9 μJ to transmit 128 bits, 1 μA, the eﬃciency is approximatively 20% or less. which is 514 nJ per bit sent during the actual transmission state. If VCO calibration and module setup are to be in- cluded, then the power consumption per bit increases to 3.2. Digital Modules 592 nJ per bit. 3.2.1. FRAM. )e FRAM IC has several modes of operation: read mode, write mode, sleep mode, and idle mode. In sleep 3.3. Module Power Consumption. )e modules comprising mode, current consumption is measured at 30 nA, or about the node are split into two categories: analog modules and 55 nW. During this mode, the FRAM is not available to digital modules. Analog modules generally present an in- process commands and must be woken up into idle mode in variant power consumption rate whatever the node’s op- order to function. Wake-up time is measured to be 15 μs. erating state is, though the warm- and cold-start modes of During idle mode, the FRAM module consumes 4 μA of the power management ICs are exceptions. current or 7.2 μW. In this mode, the FRAM module is ready Table 3 presents an overview of the power consumption to accept read and write commands. Reading data from one of the analog modules. )e power consumption of the power address consumes 2.13 mA, or 3.83 mW of energy. Writing management ICs is variable, as they are supplied by the data consumes 2.32 mA, or 4.17 μW of energy. )e power storage elements, whose voltage may vary from 1.8 V to 4 V consumption of FRAM modules is signiﬁcantly lower than in warm-start mode. ﬂash devices [33]. Table 4 presents an overview of the power consumption of the digital modules, depending on their diﬀerent states of operation. As they are supplied by the power management 3.2.2. Microcontroller. )e PIC microcontroller, similarly to modules with a constant voltage of 1.8 V, their consumption the FRAM, has several modes of operation: sleep mode, only varies depending on the state of operation. message receiving mode (which can be further separated into false wakeups and full messages), data processing mode (which includes SPI transactions), and ADC measurement mode. 3.4. Node Operation. Excluding the ineﬃciencies and qui- )e power consumption in sleep mode is measured to be escent current consumption of the power management 40 nA or 72 nW. In message receiving mode, whether due to modules, the static consumption of the node in sleep mode is false wakeups or actual messages, the consumption of the deﬁned as the sum of the comparator, FRAM, micro- microcontroller is the same, at 230 μA, or 415 μW, though controller, and transceiver power consumption in sleep the time spent processing these diﬀerent modes is diﬀerent. mode. All other power consumption states are a combination of In data processing mode, the power consumption is 410 μA, or 738 μW, and, ﬁnally, in temperature measurement mode, the modules’ operating states as shown in Tables 3 and 4, whose operation is organized by the microcontroller as fast the power consumption is 800 μA, or 1.44 mW. Another metric that is of importance is the consumed as possible, in order to minimize power consumption. )e energy per bit received. As the microcontroller receives average power consumption of the node while performing messages at 800 bauds and consumes 415 μW while receiving select commands was measured and is presented in Table 5. Wireless Power Transfer 17 –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Load current (μA) Figure 30: Eﬃciency of the buck converter of the bq25570, as a function of output current. Table 3: Power consumption of the node’s analog modules. node performance has negligible power consumption in comparison to sending data, Figure 31. Device State Current (μA) Power (μW) WS/FS 500 850–2000 bq25570 CS — 15 3.5. Simulator. )e power consumption simulator is derived WS/FS 500 850–2000 bq25504 from power consumption data measured in situ. )e sim- CS — 15 ulator uses an iterative approach to determine power con- Comparator Static 0.42 0.756 sumption over a small time interval,Δt. Once the calculation of the power consumption during Δt is complete for all elements, the states of operation of each device are updated. Table 4: Power Consumption of the node’s digital modules. )e general equation modeling the consumption of any given module of the node in the simulator is shown below: Device State Current (μA) Power (μW) Sleep 0.04 0.072 P(t) � P(t −Δt) + M(Δt, S), (4) Message RX 230 423 PIC Processing 410 738 where P is the power consumption, or generation of any Temp. measurement 800 1440 module in the node, M is the change in power consumption False wakeup 230 423 over the previous Δt time increment, and S is the current Sleep 0.03 0.054 state of operation of the node. )is includes any changes in Idle 4.00 7.200 solar and RF levels, power storage element levels, and any FRAM Storage (read) 2130 3834 operations undertaken by the node that may modify the Storage (write) 2315 4167 power consumption of the node. )e changes in power Sleep 0.02 0.036 consumption are then summed or subtracted, depending on Ready state 430 774 whether the device is a power consumer or generator, in a SPIRIT1 VCO calibration 6720 12334 chain that corresponds to power ﬂow. TX (−30 dBm) 9580 17558 An overview of the simulated algorithm is presented in Figure 32: Static power consumption of the node is 920 nW, or less than a microwatt. Most messages have similar power 4. Scenarios of Operation consumption proﬁles, around 17 μJ per message, including any operation which uses the diﬀerent modules on the Several applicative scenarios were envisioned for this type of node, due to the speed at which the operations are device. Scenarios have been classiﬁed according to the type conducted. of power mainly used to power the node, being that of RF )ere are, however, two exceptions to this generaliza- energy or of solar energy. Using data derived from part III of tion. )e ﬁrst one is when the node receives a false wakeup. this work, simulation of expected results was theoretically Most received messages are interfering signals, which are not calculated. Measured data is also presented and compared to theoretical predictions. OOK/ASK modulated, or use an incompatible frame format. With this in mind, the message reception code was designed )eoretical predictions of power consumption were to time-out after approximatively 16.8 ms if an incorrect estimated by taking into account the leakage current of message is received. storage devices, the power consumption of the power )e second exception is when the node is asked to management modules, and their ineﬃciencies, especially in transmit data, as this is a time and energy consuming op- regard to the low performance of the bq25570’s buck eration, during which the node consumes about 97.8 μJ of converter. )ese values were integrated into a program energy. When transmitting, however, other operations, the capable of simulating the power consumption of the node Efficiency (%) 18 Wireless Power Transfer Table 5: Average power consumption of the node during diﬀerent phases of operation. Command Duration (ms) Average power (μW) Energy used (μJ) Sleep mode — 0.92 — False wakeup 16.8 416 7.00 Temperature measurement 40.6 422 17.1 FRAM data storage 42.1 420 17.7 FRAM data retrieval 42.2 417 17.6 Data transmission 55.9 1750 97.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Time (ms) (a) 0.4 0.2 0 5 10 15 20 25 30 35 40 45 50 Time (ms) (b) 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) (c) 0 102030405060 Time (ms) (d) Figure 31: Power consumption of the node for diﬀerent commands. (a) )e power consumption when the node is idle. (b) )e power consumption when a message is received. (c) )e power consumption when the FRAM is solicited for a read or write operation. (d) )e power consumption when a data TX command is issued. Power (mW) Power (mW) Power (mW) Power (μW) Wireless Power Transfer 19 can be performed (maximum measurement speed) without Input scenario the main storage capacitor discharging, which will depend information on the power input to the rectiﬁer. Fixed Scenario )is scenario will be played out in two parts: for the ﬁrst conﬁguration parameters Initialize all devices part, the node will operate in RX-only mode, where the only instructions it will receive are that of data measurement and Calculate Calculate storage; during the second part, in addition to RX mode, it Variable consumer’s power generator’s power parameters will also operate in TX mode. usage creation Calculate power 4.1.1. RX-Only Operation. )e calculated and measured ﬂow balances speeds at selected points of this metric are reported in Figure 33. Update all states ∆t increment At measurement intervals of 10, 50, 100, 500, and 1000 seconds, the node requires, respectively, −6.2, −11.3, −12.5, Yes Store simulation Plot −14.2, and −14.5 dBm of input power in order to remain End? results results operational. It can be seen that as the frequency of measurements No becomes increasingly close to the theoretical 6.66 Hz Generated data maximum possible, the amount of power required for maintaining the system operational increases rapidly. It is Figure 32: Overview of the simulator algorithm. not possible to reach the theoretical limit of 6.66 Hz, as this would require more instantaneous energy than the rectiﬁer can provide, 10 mW, if it is supposed that over time, in order to determine the accuracy of the sim- a maximum input power of 0 dBm is available for ulations in regard to the actual power consumption. this scenario. As the interval between measurements For these scenarios, a data measurement operation shall increases, the power level trends towards the hard limit be deﬁned as the combination of a temperature measure- of −15.1 dBm. ment command with an FRAM write command. A data )e higher levels of power required as the frequency of transmission operation shall be deﬁned as an FRAM read measurements increases are likely due to cumulated inter- command combined with a data transmission command. action of ineﬃciencies (buck converter, MPPT out of range It should be noted that the maximum speed at which the as power input levels increase,. . .) in the node as power node can receive commands is approximatively at 6.67 Hz, consumption rises. or 150 ms. Although the microcontroller takes approx- )e typical distance at which the node can operate in imatively 50 to 60 ms to fully process a command (Fig- RX-only continuous RF mode ranges from 1 m to 6 m, ure 30), the wake-up radio takes approximatively 150 ms to though the amount of time between measurements increases return to a “known state,” wherein the wake-up generation with distance, in order to give enough time for the storage circuit is fully discharged (Figures 20 and 23), thereby element to recharge and return to its previous level. )e limiting the speed to a maximum of 6.67 requests per recharge time for 1, 2, 5, and 6 m distances was measured to second. be 10, 21, 344, and 1505 s, respectively. Operation at 10 m was also observed as a result of multipath propagation in the 4.1. Continuous RF Operation. Continuous RF operation of measurement environment. the sensor node means that it will be constantly powered, but only through RF energy, from a 28 dBm emitter, with a 3 dBi 4.1.2. RX and TX Operation. Similarly to the ﬁrst part of the patch antenna, operating at 935 MHz. )e ceramic capacitor scenario, the node will now be instructed to perform a data storage elements recharge back up to 4 V if powered with at measurement and a data transmission operation. )is will least −15.1 dBm of power input, which is the minimum reduce the eﬀective maximal frequency rate of measure- amount of power required to keep the PIC, comparator, and ments down to 3.33 Hz. )e calculated and measured speeds bq25570 operational (while sleeping and without commu- at selected points of this metric are shown in Figure 34. nications). )is corresponds to a power usage of 7 μW, of At measurement intervals of 10, 50, 100, 500, and 1000 which the bq25504 uses 2 μW. )e high usage of the digital seconds, the node requires, respectively, −1.9, −7.6, −9.7, subsection can be explained by the low eﬃciency of the buck −13.2, and −14 dBm of input power in order to remain converter for very small supply currents (Figure 30). )e operational. bq25504 will be considered to be nonoperational in this )e node is more restricted in terms of measurement mode and therefore will not consume any energy. )e speed, due to the power consumption of transmission op- storage element for this scenario is a 3720 μF ceramic ca- erations, which are nearly an order of magnitude larger than pacitor-based element. reception and measurement only operations; however, op- )e main criterion to evaluate the performance of eration is still possible with only RF power. continuous operation of the node is how fast measurements 20 Wireless Power Transfer –2 –4 –6 –8 –10 –12 –14 –16 0 1 2 3 4 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 33: Maximum measurement speed as a function of input power, with a 3720 μF ceramic capacitor-based storage element, for data measurement operations. –2 –4 –6 –8 –10 –12 –14 –16 0 1 2 3 4 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 34: Maximum measurement speed as a function of input power, with a 3720 μF ceramic capacitor-based storage element, for data measurement and data transmission operations. 4.2. Fully Autonomous Operation. As autonomous operation the energy drain is considered to be only supercapacitor self- would require at least autonomy of a day (or more); it is discharge, in combination with the quiescent current of the therefore necessary to use a supercapacitor to store the node in sleep mode. )e worst-case scenario, on the other energy required to power the circuit during sleep mode. hand, is when the PIC is nearly continuously solicited. In this second scenario, a 470 mF capacitor is charged to full capacity at 4 V, for 30 minutes, using a standard power supply, and hot-plugged into the bq25504 circuit. No power, 4.2.1. RX-Only Operation. )e calculated and measured speeds at selected points of this metric are reported in Figure 35. whether solar or RF energy, was given to the device. Both the At measurement intervals of 1 s, 5 s, 10 s, 1 min, 1 hour, bq25570 and the bq25504 operated in WS mode during the length of the scenario. and 2.8 hours, the autonomy of the node is, respectively, 12, 29, 37, 49, 53, and 53 hours. )is is close to, but slightly less )e main criterion to evaluate the performance of this autonomous operation of the node is how long it could eﬃcient than, theoretically predicted results. As measure- ment intervals increase to more than one hour, the static operate with main storage supercapacitor discharging, depending on the frequency of measurements. power consumption of the system overtakes that of the power used during normal operations, and so the autonomy Discharge time was measured to be 2.25 days with no measurements. )is is considered to be the case scenario, as varies little. Input power (dBm) Input power (dBm) Wireless Power Transfer 21 –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 35: Maximum node autonomy is a function of command frequency, with a 470 mF supercapacitor storage element, for data measurement operations. As measurements increase in frequency, the autonomy of recharge the node during the day, supposing that it is solicited during the night. In regard to the duration of a the node decreases, though with a measurement every ten seconds, the node still has autonomy of 37 hours. If the node is day, a pessimistic case of 8 hours of light and 16 hours of not solicited, the autonomy of the node is approximatively 54 darkness will be supposed. Furthermore, it will be sup- hours. )is is a hard limit which depends on the supercapacitor posed that the node will always be operating in warm- leakage and the power consumption of the whole node. start mode. Using a larger supercapacitor, or higher maximum Figure 37 shows the amount of time required for the storage element voltages, would increase the autonomy of node to charge from the beginning of warm-start mode until the node. However, larger storage elements would come at a the storage element (supercapacitor) is full, as a function of cost of longer cold-start times. luminosity, supposing that the node is not solicited. It can be observed from Figure 35 that, for the white For longer autonomy requirements, it may be necessary to implement a cold-start charge method that will rapidly LEDs, at least 155 lux is required to recharge the node in 8 hours and that, for the yellow LEDs, 135 lux is required. )is kick-start the system into warm-start mode, which may use existing energy harvesting methods (i.e., leave the node supposes that the node is fully depleted in terms of the under a very bright light or leave the node next to a high- warm-start mode but has not yet entered cold-start mode. power transmitter, until it indicates that it is operational) or In the context of this scenario, if the node is solicited at require a cable-based solution. night, then 16 hours of autonomy is required. From Fig- ures 35 and 36, it can be determined that the node must not be sent commands more than once every 1.5 seconds and 4.2.2. RX and TX Operation. )e calculated and measured every 7 seconds, respectively, for RX and TX/RX operation speeds at selected points of this metric are reported in modes, in order to stay in warm-start mode. )erefore, it is Figure 36. possible to conclude that the node may operate in envi- At measurement intervals of 1 s, 5 s, 10 s, 1 min, 1 hour, ronments with luminosity as low as 150 to 200 lux for an and 2.8 hours, the autonomy of the node is, respectively, 4, indeﬁnite amount of time, with reasonable environmental 13.5, 21, 41, 52, and 53 hours. )e autonomy of the node has sensing capabilities. decreased compared to the previous subscenario, due to the high consumption of data transmission operations. How- 5. Analysis ever, at measurement intervals of over an hour, the node still RF harvesting techniques can produce work which is presents similar autonomy. capable of generating energy with low levels of power input, such as [4], where charging of storage occurs with 4.3. Solar Operation. )e autonomy and power consump- power levels of −29 dBm, without any load present. Other tion of the node have been characterized in autonomous RF-only harvesting systems have produced similar per- mode. Solar energy is analyzed, using the large solar panel formance with this work, such as [19, 21, 34], with power and both LED types. harvesting being usable, without load, with power input As solar energy can easily provide suﬃcient energy to levels around −20 to −25 dBm. Distances at which the power the node, the main criterion for this scenario will node could operate from a 28 dBm power source ranged analyze the amount of luminosity required in order to from 1 m to 6 m. Autonomy (h) 22 Wireless Power Transfer –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 36: Maximum node autonomy is a function of command frequency, with a 470 mF supercapacitor storage element, for data measurement and data transmission operations. 100 200 300 400 500 600 700 800 900 1000 Luminosity (lux) Yellow LEDs White LEDs Figure 37: Node warm-start charge time as a function of luminous intensity. Combining RF and solar energy leads to diverse per- As sleep current is not negligible compared to the power formance depending on the architecture used, such as consumption of storage element leakages and power man- [29, 30, 35, 36]. Due to diﬀerent topologies and solutions agement IC quiescent currents, it is normal that perfor- used, it is diﬃcult to compare such work homogeneously; mance metrics are worse than power harvesting systems in however, it can be noted that solar energy delivers more which no load is present. energy than RF harvesting and that the performance of this Integration of a single antenna with both data and power work was similar to [30]. In regard to the previously de- reception was attempted. However, the integration was not veloped node [8], the minimal power requirements to power possible, due to interference generated by the boost con- verter used for RF power harvesting. Indeed, the constant this improved node in sleep mode are quasi-identical, as the added elements have very little sleep current, in the range of modiﬁcation of the load impedance presented to the rec- several nA. )e addition of a solar-powered energy source in tiﬁers, in order to perform MPPT, caused interference in the this work extends the autonomy of the node in environment data reception (comparator) circuit, creating a situation in where solar energy, even at ambient levels, is available. which false wakeups were continuously generated. Time (hours) Autonomy (h) Wireless Power Transfer 23 considered in order to increase the diversity of powering Furthermore, this work adds data retransmission ca- pabilities compared to the previously developed node [8]. methods, to improve redundancy in terms of power availability. )is is useful, as it makes measurement data retrieval possible without having to make any physical connections to the node. 6. Conclusion and Perspectives A custom-based interface using SDR was also created for sending commands to the node. Furthermore, this work )is work demonstrates the possibility of creating a COTS- does not only analyze metrics of the power harvesting el- based, hybrid RF, and solar-powered IoT sensor node, ca- ements of IoT nodes but develops details on the more pable of environmental sensing with a restrained power functional aspects of the node. )is includes what the node budget. )e active section of the node consumes 920 nW of can do, in terms of functionality, how it does this, and a list power in sleep mode, of which 870 nW is used for the wake- of the performance of diﬀerent modules, in order to see up circuitry. where optimization is required in order to reduce power During RF-only power operation (in the 900 MHz ISM consumption. band), the device may function with power levels down to )ese additions have also permitted the creation of more −15.1 dBm, albeit with a cold-start requirement of complex scenarios of operation for the node. Scenarios in the −13.1 dBm, with a 28 dBm transmitter capable of providing previous work [8] were limited to the amount of mea- both data and energy. surement operations that could be done with a set RF power In solar mode, the device may function with luminosity input, and the autonomy of the node was, at best, limited to levels down to 70 lux, albeit with a cold-start requirement of several hours. 100 to 150 lux, depending on supercapacitor leakage current, In this work, the node can now operate for several days which is a typical lighting situation in an oﬃce or a house. without any power source present (under the assumption )e system is fully operational, able to handle both that its storage elements are full), and the varied power communications and RF energy transfer over the same sources increase energy redundancy, should one of the band, and, charge time notwithstanding, operate over a sources of power be not available. A greater variety of period of one or two days with the presence of a super- scenarios of operation is also possible with the further ad- capacitor as a storage element. Additionally, the node is dition of a data retransmission module. capable of wirelessly transmitting back data when A power consumption simulator, which was not present requested to do so. in our previous work [8], is also developed based on )is work characterizes various operational scenarios measured performance data and is used to compare mea- that IoT nodes could be used in, giving an additional sured performance in real settings with theoretical perfor- insight into the abilities and limitations that are possible mance. )eoretical predictions and real data match closely. in a small subset of potential operating environments. Furthermore, the simulator can be used in order to create Besides cartography of diﬀerent modes of operation, at the and test other hypothetical scenarios before they are system level, a model is proposed based on the analysis implemented, which can be used to test their viability before and characterization of the power consumption of the requiring measurements to be undertaken. functions featured in the proposed hybrid node. )e )ere are several limitations in this system that could be resulting implementation exhibits a good correlation with improved upon. On rare occasions, false wakeups can also actual measurements, thus allowing for prediction of the trigger the microcontroller, which will waste energy trying to node’s performance in various environments, or other process data. )is is partly remedied by using a timeout hypothetical scenarios. counter in the software loop, to minimize time spent decoding erroneous/false messages. A diﬀerent reception Data Availability architecture or modulation scheme should be considered, as in [37]. More energy eﬃcient transceivers could also be used )e data used to support the ﬁndings of this study are in- [38]. cluded within the article. Other data used to support the Another limitation is that of the bq25570 and bq25504 ﬁndings of this study are available from the corresponding power management ICs, which are being used near their author upon request. lower limits in terms of energy input, and therefore they do not always provide highly eﬃcient energy transfer, which can easily triple the static energy consumption in sleep mode Conflicts of Interest or cause long cold-start charging times. )is could be )e authors declare that there are no conﬂicts of interest mitigated by using custom-designed power management regarding the publication of this paper. ICs, as in [39], or by using extremely low-power techniques to create equivalent replacements [36]. Future additions that could be explored for this work Acknowledgments include adding a larger variety of environmental sensors, which remains possible using the SPI bus present on the )e authors would like to thank M. De Matos for her help in microcontroller. Additionally, the use of other power har- using certain measurement instrumentation and Y. Faize for vesting methods (such as thermoelectric modules) should be the loan of his Joulescope. 24 Wireless Power Transfer [20] J. Nicot and T. Taris, “Remote RF powering of ambient References sensors,” in Proceedings of the 2016 IEEE International [1] B. Martinez, M. Monton, ´ I. Vilajosana, and J. D. Prades, “)e Conference on Electronics, Circuits and Systems (ICECS), power of models: modeling power consumption for IoT de- pp. 760–763, Monte Carlo, Monaco, 2016. vices,” IEEE Sensors Journal, vol. 15, no. 10, pp. 5777–5789, [21] K. Kotani, S. Inoue, T. Komiyama, Y. Chonan, and H. Yamaguchi, “RF rectiﬁer absolute ﬁgure of merit based on [2] A. Hussein, “Energy consumption of wireless IoT nodes,” relative comparison with reference rectiﬁer,” in Proceedings of Master’s thesis, Norwegian University of Science and Tech- the 2019 IEEE International Symposium on Circuits and nology, Trondheim, Norway, 2017. Systems (ISCAS), pp. 1–4, Sapporo, Japan, 2019. [3] J. A. Paradiso and T. Starner, “Energy scavenging for mobile [22] P. T. Tolnai and E. Racz, ´ “Preliminary spectral design of a and wireless electronics,” IEEE Pervasive Computing, vol. 4, LED- and light bulb-based sun spectrum simulator,” in no. 1, pp. 18–27, 2005. Proceedings of the 2019 IEEE 17th World Symposium on [4] M. Piñuela, P. D. Mitcheson, and S. Lucyszyn, “Ambient RF Applied Machine Intelligence and Informatics (SAMI), energy harvesting in urban and semi-urban environments,” pp. 217–222, Herlany, Slovakia, 2019. IEEE Transactions on Microwave :eory and Techniques, [23] “Recommended light levels (illuminance) for outdoor and vol. 61, no. 7, pp. 2715–2726, 2013. indoor venues,” 2019, https://www.noao.edu/education/ [5] C. Greene, “Wireless power, energy harvesting, and power QLTkit/ACTIVITY_Documents/Safety/LightLevels_ management solutions for sensors and the IoT,” 2017, http:// outdoor+indoor.pdf. www.powercastco.com/wp-content/uploads/. [24] T. Esram and P. L. Chapman, “Comparison of photovoltaic [6] B. Calhoun and D. Wentzloﬀ, “Ultra-low power wireless SoCs array maximum power point tracking techniques,” IEEE enabling a batteryless IoT,” in Proceedings of the 2015 IEEE Transactions on Energy Conversion, vol. 22, no. 2, pp. 439–449, Hot Chips 27 Symposium (HCS), pp. 1–45, Cupertino, CA, USA, 2015. [25] H. Lee, K.-C. Lee, J. Schunke, and L. Burton, “Leakage cur- [7] V. Raghunathan, A. Kansal, J. Hsu, J. Friedman, and rents in multilayer ceramic capacitors,” IEEE Transactions on M. Srivastava, “Design considerations for solar energy har- Components, Hybrids, and Manufacturing Technology, vol. 7, vesting wireless embedded systems,” in Proceedings of the 4th no. 4, pp. 443–453, 1984. International Symposium on Information Processing in Sensor [26] Y. Ramadash, “bq25505/70 design help V1.3,” 2013, http:// Networks, pp. 457–462, Boise, ID, USA, 2005. www.ti.com/lit/zip/sluc484. [8] N. John, F. )ierry, and T. Taris, “An RF-powered IoT node [27] A. S. Weddell, G. V. Merrett, T. J. Kazmierski, and B. M. Al- for environment sensoring,” in Proceedings of the 2019 IEEE Hashimi, “Accurate supercapacitor modeling for energy Wireless Power Transfer Conference (WPTC), London, UK, harvesting wireless sensor nodes,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 58, no. 12, [9] Broadcom, “HSMS-285x series surface mount zero bias pp. 911–915, 2011. Schottky detector diodes,” 2009, https://docs.broadcom.com/ [28] RF Schottky Diodes, “Inﬁneon,” 2016, https://www.inﬁneon. doc/AV02-1377EN. com/dgdl/Inﬁneon-RF%20Schottky%20Diodes-PB-v01_01- [10] Texas Instruments, “bq25570 nano power boost charger and EN.pdf?ﬁleId=5546d462557e6e890155917b964163fb. buck converter for energy harvester powered applications,” [29] K. Niotaki, F. Giuppi, A. Georgiadis, and A. Collado, “Solar/ 2018, http://www.ti.com/lit/ds/symlink/bq25570.pdf. EM energy harvester for autonomous operation of a moni- [11] Texas Instruments, “bq25504 ultra low-power boost converter toring sensor platform,” Wireless Power Transfer, vol. 1, no. 1, with battery management for energy harvester applications,” pp. 44–50, 2014. 2019, http://www.ti.com/lit/ds/symlink/bq25504.pdf. [30] J. Bito, R. Bahr, J. G. Hester, S. A. Nauroze, A. Georgiadis, and [12] Taiyo Yuden, “Multi layer ceramic capacitor,” 2020, https://ds. M. M. Tentzeris, “A novel solar and electromagnetic energy yuden.co.jp/TYCOMPAS/or/specSheet? harvesting system with a 3-D printed package for energy pn=AMK432BJ477MM-T. eﬃcient internet-of-things wireless sensors,” IEEE Transac- [13] SCM, “Series-connected super capacitor modules,” 2020, tions on Microwave :eory and Techniques, vol. 65, no. 5, http://datasheets.avx.com/AVX-SCM.pdf. pp. 1831–1842, 2017. [14] PIC16LF1554/1559, “20-Pin ﬂash, 8-bit microcontrollers with [31] W. S. Wang, T. O’Donnell, L. Ribetto, B. O’Flynn, M. Hayes, XLP technology,” 2016, http://ww1.microchip.com/ and C. O’Mathuna, “Energy harvesting embedded wireless downloads/en/DeviceDoc/40001761E.pdf. sensor system for building environment applications,” in [15] FeRAM MR45V100A, “Lapis semiconductor,” 2017, http:// Proceedings of the 2009 1st International Conference on www.lapis-semi.com/en/data/datasheet-ﬁle_db/Memory/ Wireless Communication, Vehicular Technology, Information FEDR45V100A-01.pdf. :eory and Aerospace & Electronic Systems Technology, [16] “TS881 rail-to-rail 0.9 V nanopower comparator,” 2013, pp. 36–41, Aalborg, Denmark, 2009. https://www.st.com/resource/en/datasheet/ts881.pdf. [32] S. J. Marinkovic and E. M. Popovici, “Nano-power wireless [17] SPIRIT1, “Low data rate, low power sub-1 GHz transceiver,” wake-up receiver with serial peripheral interface,” IEEE 2016, https://www.st.com/resource/en/datasheet/spirit1.pdf. Journal on Selected Areas in Communications, vol. 29, no. 8, [18] D. Karolak, Syst`eme de Radiocommunication T´el´e-Aliment´e pp. 1641–1647, 2011. par voie Radiofr´equence a` 2.45 GHz, Universite´ de Bordeaux, [33] H. Medu, “Energy comparison of cypress F-ram and Bordeaux, France, 2015. EEPROM, cypress semiconductor,” 2017, https://www. [19] T. Taris, V. Vigneras, and L. Fadel, “A 900 MHz RF energy cypress.com/ﬁle/46746/download. harvesting module,” in Proceedings of the 10th IEEE Inter- [34] R. Usami, T. Komiyama, Y. Chonan, H. Yamaguchi, and national NEWCAS Conference, pp. 445–448, Montreal, K. Kotani, “Photovoltaic-assisted self-vth-cancellation CMOS Canada, 2012. RF rectiﬁer for wide power range operation,” in Proceedings of Wireless Power Transfer 25 the 2018 IEEE Asian Solid-State Circuits Conference (A-SSCC), pp. 115–118, Tainan, Taiwan, 2018. [35] A. P. Sample, J. Braun, A. Parks, and J. R. Smith, “Photovoltaic enhanced UHF RFID tag antennas for dual purpose energy harvesting,” in Proceedings of the 2011 IEEE International Conference on RFID, pp. 146–153, Orlando, FL, USA, 2011. [36] J. Bito, J. G. Hester, and M. M. Tentzeris, “A fully autonomous ultralow power hybrid RF/photovoltaic energy harvesting system with −25 dBm sensitivity,” in Proceedings of the 2017 IEEE Wireless Power Transfer Conference (WPTC), Taipei, Taiwan, 2017. [37] S. Oh, N. E. Roberts, and D. D. Wentzloﬀ, “A 116 nW multi- band wake-up receiver with 31-bit correlator and interference rejection,” in Proceedings of the IEEE 2013 Custom Integrated Circuits Conference, pp. 1–4, San Jose, CA, USA, 2013. [38] A. Molnar, B. Lu, S. Lanzisera, B. W. Cook, and K. S. J. Pister, “An ultra-low power 900 MHz RF transceiver for wireless sensor networks,” in Proceedings of the IEEE 2004 Custom Integrated Circuits Conference, pp. 401–404, Orlando, FL, USA, 2004. [39] T. Paing, E. Falkenstein, R. Zane, and Z. Popovic, “Custom IC for ultra-low power RF energy harvesting,” in Proceedings of the 2009 24th Annual IEEE Applied Power Electronics Con- ference and Exposition, pp. 1239–1245, Washington, DC, USA, 2009. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Wireless Power Transfer Hindawi Publishing Corporation http://www.deepdyve.com/lp/hindawi-publishing-corporation/an-autonomous-wireless-sensor-node-based-on-hybrid-rf-solar-energy-MOaFkBbAMx

Loading next page...

References (40)

Dean Karolak (2015)
Système de radiocommunication télé-alimenté par voie radiofréquence à 2.45 GHz
(2017)
Lapis semiconductor
T. Esram, P. Chapman (2007)
Comparison of Photovoltaic Array Maximum Power Point Tracking Techniques
IEEE Transactions on Energy Conversion, 22
Texas Instruments
bq25570 nano power boost charger and buck converter for energy harvester powered applications,
K. Kotani, S. Inoue, T. Komiyama, Y. Chonan, H. Yamaguchi (2019)
RF Rectifier Absolute Figure of Merit Based on Relative Comparison with Reference Rectifier
2019 IEEE International Symposium on Circuits and Systems (ISCAS)
(2017)
Wireless power, energy harvesting, and power management solutions for sensors and the IoT
F. Giuppi K. Niotaki (2014)
Solar/EM energy harvester for autonomous operation of a monitoring sensor platform,
Wireless Power Transfer, 1
Hee-Kab Lee, Kyoungho Lee, J. Schunke, L. Burton (1984)
Leakage Currents in Multilayer Ceramic Capacitors
IEEE Transactions on Components, Hybrids, and Manufacturing Technology, 7
(2019)
bq25504 ultra low-power boost converter with battery management for energy harvester applications
(2013)
TS881 rail-to-rail 0.9 V nanopower comparator
A. Sample, J. Braun, Aaron Parks, Joshua Smith (2011)
Photovoltaic enhanced UHF RFID tag antennas for dual purpose energy harvesting
2011 IEEE International Conference on RFID
Taiyo Yuden
Multi layer ceramic capacitor,
(2019)
Recommended light levels (illuminance) for outdoor and indoor venues
A. Molnar, Benson Lu, S. Lanzisera, B. Cook, Kristofer Pister (2004)
An ultra-low power 900 MHz RF transceiver for wireless sensor networks
Proceedings of the IEEE 2004 Custom Integrated Circuits Conference (IEEE Cat. No.04CH37571)
Ren Usami, T. Komiyama, Y. Chonan, H. Yamaguchi, K. Kotani (2018)
Photovoltaic-Assisted Self-Vth-Cancellation CMOS RF Rectifier for Wide Power Range Operation
2018 IEEE Asian Solid-State Circuits Conference (A-SSCC)
K. Niotaki, F. Giuppi, A. Georgiadis, A. Collado (2014)
Solar/EM energy harvester for autonomous operation of a monitoring sensor platform
Water Practice & Technology, 1
Pál Tolnai, Ervin Rácz (2019)
Preliminary spectral design of a LED- and light bulb-based sun spectrum simulator
2019 IEEE 17th World Symposium on Applied Machine Intelligence and Informatics (SAMI)
B. Calhoun, D. Wentzloff (2015)
Ultra-low power wireless SoCs enabling a batteryless IoT
2015 IEEE Hot Chips 27 Symposium (HCS)
Amen Hussain (2017)
Energy Consumption of Wireless IoT Nodes
T. Paing, E. Falkenstein, R. Zane, Z. Popovic (2009)
Custom IC for Ultra-low Power RF Energy Harvesting
2009 Twenty-Fourth Annual IEEE Applied Power Electronics Conference and Exposition
A. Weddell, G. Merrett, T. Kazmierski, B. Al-Hashimi (2011)
Accurate Supercapacitor Modeling for Energy Harvesting Wireless Sensor Nodes
IEEE Transactions on Circuits and Systems II: Express Briefs, 58
J. Bito, J. Hester, M. Tentzeris (2017)
A fully autonomous ultra-low power hybrid RF/photovoltaic energy harvesting system with −25 dBm sensitivity
2017 IEEE Wireless Power Transfer Conference (WPTC)
W. Wang, T. O'Donnell, L. Ribetto, B. O’flynn, M. Hayes, C. O'Mathúna (2009)
Energy harvesting embedded wireless sensor system for building environment applications
2009 1st International Conference on Wireless Communication, Vehicular Technology, Information Theory and Aerospace & Electronic Systems Technology
(2017)
Energy comparison of cypress F-ram and EEPROM, cypress semiconductor
V. Raghunathan, A. Kansal, Jason Hsu, J. Friedman, M. Srivastava (2005)
Design considerations for solar energy harvesting wireless embedded systems
IPSN 2005. Fourth International Symposium on Information Processing in Sensor Networks, 2005.
Borja Martínez, M. Montón, I. Vilajosana, J. Prades (2015)
The Power of Models: Modeling Power Consumption for IoT Devices
IEEE Sensors Journal, 15
M. Piñuela, P. Mitcheson, S. Lucyszyn (2013)
Ambient RF Energy Harvesting in Urban and Semi-Urban Environments
IEEE Transactions on Microwave Theory and Techniques, 61
(2016)
20-Pin flash, 8-bit microcontrollers with XLP technology
(2016)
Low data rate, low power sub-1 GHz transceiver
J. Paradiso, Thad Starner (2005)
Energy scavenging for mobile and wireless electronics
IEEE Pervasive Computing, 4
S. Marinkovic, E. Popovici (2011)
Nano-Power Wireless Wake-Up Receiver With Serial Peripheral Interface
IEEE Journal on Selected Areas in Communications, 29
(2020)
Series-connected super capacitor modules
(2016)
Infineon
(2009)
HSMS-285x series surface mount zero bias Schottky detector diodes
J. Bito, R. Bahr, J. Hester, S. Nauroze, A. Georgiadis, M. Tentzeris (2017)
A Novel Solar and Electromagnetic Energy Harvesting System With a 3-D Printed Package for Energy Efficient Internet-of-Things Wireless Sensors
IEEE Transactions on Microwave Theory and Techniques, 65
J. Nicot, T. Taris (2016)
Remote RF powering of ambient sensors
2016 IEEE International Conference on Electronics, Circuits and Systems (ICECS)
Seunghyun Oh, N. Roberts, D. Wentzloff (2013)
A 116nW multi-band wake-up receiver with 31-bit correlator and interference rejection
Proceedings of the IEEE 2013 Custom Integrated Circuits Conference
T. Taris, V. Vigneras, L. Fadel (2012)
A 900MHz RF energy harvesting module
10th IEEE International NEWCAS Conference
(2013)
bq25505/70 design help V1.3
Nicot John, Fadel Ludivine, T. Thierry (2019)
An RF-Powered IoT Node for Environment Sensoring
2019 IEEE Wireless Power Transfer Conference (WPTC)

Publisher: Hindawi Publishing Corporation
Copyright: Copyright © 2021 John Nicot et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
eISSN: 2052-8418
DOI: 10.1155/2021/6642938
Publisher site: See Article on Publisher Site

Abstract

Hindawi Wireless Power Transfer Volume 2021, Article ID 6642938, 25 pages https://doi.org/10.1155/2021/6642938 Research Article An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting 1,2,3 1,2,3 1,2,3 John Nicot , Ludivine Fadel , and Thierry Taris University of Bordeaux, IMS, UMR 5218, 33405 Talence, Bordeaux, France CNRS, IMS, UMR 5218, 33405 Talence, Bordeaux, France Bordeaux INP, IMS, UMR 5218, 33405 Talence, Bordeaux, France Correspondence should be addressed to John Nicot; john.nicot@ims-bordeaux.fr Received 19 October 2020; Revised 2 June 2021; Accepted 10 July 2021; Published 3 August 2021 Academic Editor: Jiafeng Zhou Copyright © 2021 John Nicot et al. )is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. )e widespread deployment of the Internet of )ings (IoT) requires the development of new embedded systems, which will provide a diverse array of diﬀerent intelligent functionalities. However, these devices must also meet environmental, maintenance, and longevity constraints, while maintaining extremely low-power consumption. In this work, a batteryless, low-power con- sumption, compact embedded system for IoT applications is presented. )is system is capable of using a combination of hybrid solar and radiofrequency power sources and operates in the 900 MHz ISM band. It is capable of receiving OOK or ASK modulated data and measuring environmental data and can transmit information back to the requester using GFSK modulated data. )e total consumption of the system during its sleep state is 920 nW. Minimum power required to operate is −15.1 dBm or 70 lux, when using only radiofrequency or solar powering, respectively. )e system is fully designed with components oﬀ the shelf (COTS). It is important to diﬀerentiate three types of RF power 1. Introduction harvesting: opportunistic power harvesting, which is usually With the ever-increasing growth of wireless sensors and unreliable, as power levels can vary greatly depending on the their associated networks, in the context of the deployment environment and are generally low [4]; dedicated power of the IoT, it is of great importance for such devices to harvesting, which can provide high levels of power at the minimize their power consumption. On top of this con- cost of specialized hardware and, possibly, proprietary straint, modern sensor nodes not only must maintain low- communication and power transmission protocols [5]; or cost deployment, but also are expected to minimize their semidedicated power harvesting, which can involve diverse environmental impact. )erefore, it is crucial to explore new methods of powering, including piggy-backing over existing methods and techniques in order to power and use these standards or devices currently in place (possibly with minor nodes. )e rise of diverse energy harvesting methods (solar modiﬁcations to implement communications) or remotely powering the device at regular intervals, in order to keep the energy, thermoelectric energy, and vibrational energy) makes it possible to use the environment to power these IoT sensor node operational [6]. nodes and does not require them to be permanently con- Solar energy, on the other hand, depending on the size of nected to the classical electrical grid. the solar panel and the ambient luminosity levels, can easily )e range of power consumption, for most IoT sensor provide several milliwatts of power in an outdoor conﬁg- nodes, varies between several tens to hundreds of micro- uration down to several tens of microwatts of power in watts, depending on their range of operation and func- indoor environments [7]. tionality [1, 2]. Radiofrequency (RF) powering can be used in )is work investigates a hybrid-powered solar and RF order to provide up to a milliwatt at close range or several harvesting wireless sensor node, on top of which bidirec- microwatts at further distances [3]. tional communications are implemented. In order to 2 Wireless Power Transfer minimize the power consumption of the node, data re- using the Greinacher topology, Figure 3, in order to achieve ception is implemented under the form of an ASK or OOK large voltage ampliﬁcation during RF to DC conversion. Schottky diodes are preferred due to their low forward modulated, low-power wake-up radio, and data transmis- sion is ensured by using a specialized, low-power transceiver voltage—typically 80 mV for HSMS-285C diodes [19]. using GFSK modulation, for the transmission of data col- )e highly capacitive nature of diode-based rectiﬁer lected by the node. circuits makes them diﬃcult to match at radiofrequencies. It )is work is an extension of previous work [8]. )e is important that the matching network exhibits a high- extensions implemented include the addition of hybrid quality factor, Q, in order to maximize the voltage at the solar/radiofrequency power harvesting, data retransmission input of the diodes, thus minimizing the losses due to their capability from the node, several additional scenarios of forward voltage [20]. operation, and an overview of a simulator used to model the )e S parameters of the n-stage rectiﬁers are shown in aforementioned scenarios. Figure 4, before matching, from 700 MHz to 1.1 GHz, to )is work is structured into three parts. Firstly, the node emphasize the high capacitance present on the circuit’s is presented, and each module composing it is visited, in input. order to demonstrate how each node module is designed. )e matching networks are conﬁgured with an inductor- )e modules are further characterized, in order to analyze based, double “L” network topology, Figure 5. their performance in terms of metrics. Secondly, an overview )e values for 50 Ω matching were simulated using ADS, of power consumption of the node in diﬀerent states of and further adjustments were done in situ through an it- operation is presented, in order to characterize power erative, dichotomous approach. )e values determined for consumption. )is is used in order to develop a simple the rectiﬁers are presented in Table 1. simulator which can predetermine the performance of the )is ensures an input return loss below −10 dB over node in a given scenario of operation. )irdly, these diﬀerent more than 30 MHz of bandwidth at 935 MHz at lower power scenarios of operation are presented, and the performance of levels (−20 to −35 dBm), Figure 6. the node is compared between data generated through the Due to varying diode capacitance as a function of power simulator and in situ measurements. Finally, this work input P , the center frequency of the matched rectiﬁers IN concludes the presented work and compares it to other work tends to shift, though not signiﬁcantly. present in this ﬁeld. As illustrated in Figure 5, the centered frequency of the input return loss shifts with the input power, due to the dependence of the diode capacitance on the latter. )is 2. Presentation of the Node variation remains negligible regarding the achievable bandwidth over the considered range of power and targeted )e node is made exclusively from COTS. )e architecture application. Indeed, for low-power IoT, the broadcasted data of the node is presented in Figure 1; it includes a rectenna, is sent at low bitrates, requiring narrow bandwidth, and based on HSMS-285C Schottky diodes [9], to collect both powering methods via RF often use Continuous Wave (CW) harvested RF power and modulated data to control the node carriers [20]. (operating at 935 MHz); a generic solar panel; a bq25570 Rectiﬁer performance is evaluated by two metrics: the power harvesting IC to manage RF energy [10]; a bq25504 sensitivity, S (1), which is the unloaded voltage output, and power harvesting IC to manage solar energy [11]; ceramic the eﬃciency, ƞ (2), which is the ratio of the output DC capacitors as storage elements (multiples of 4V, 470 μF power (P ) to the input power (P ), depending on the AMK432BJ477MM-T capacitors for RF energy [12] and a OUT IN load R : 470 mF supercapacitor for solar energy [13]); a PIC16LF1559 L microcontroller for functionality [14]; an MR45V100 FRAM S � V P 􏼁 , (1) O IN for storage [15]; a TS811 comparator for demodulation of received data and the wake-up circuit [16]; and a SPIRIT1 transceiver for data transmission, operating at 869 MHz [17]. P R 􏼁 V OUT L O (2) η � � . RF power and instructions sent to the node are provided P R · P IN L IN through a B200 Software-Deﬁned Radio (SDR) and am- Increasing the number of stages improves the sensi- pliﬁed by a ZHL-42W ampliﬁer in order to provide power tivity at low levels of power, Figure 7, but degrades the outputs ranging from 0 to 28 dBm at the antenna. )is is rectiﬁers’ eﬃciency accordingly, Figures 7(a) and 7(b). typical for semidedicated RF power harvesting systems. )e Interestingly, the increase of cascaded stages makes the system is operated though a custom-developed software eﬃciency increasingly independent of the load value R , package, operating under MATLAB. L Figure 8. Similar work has yielded the same conclusions [21]. Based on this analysis, we will use 4 stages to receive 2.1. Rectenna. Harvesting of power and reception of data data and only one stage to harvest RF energy. Four stages through radio-waves require antennas. A multibranch di- were chosen for data reception in order to permit the node pole antenna, Figure 2, is developed [18]. It exhibits a return to receive data at a suﬃcient distance (up to 20 meters), loss of −15 dB and a directivity of 2 dBi at 935 MHz. while avoiding overloading the node’s reception module, as Due to the low-power levels of collected RF signals, the the sensitivity (and therefore outputted voltage) increases rectiﬁcation is based on cascaded stages of voltage doublers, with the number of stages. Wireless Power Transfer 3 Solar energy Energy Power 935 MHz storage mgmt. Rectification Energy Power storage mgmt. Power + data TX (OOK/ASK) Demod. + Processing wake-up Microprocessor Sensor module(s) 869 MHz Storage Feedback (FRAM) module Data RX (GFSK) Power path Data path Figure 1: Overview of the IoT sensor node’s diﬀerent modules, including external modules used for data transmission and reception. –5 –10 –15 –20 900 920 940 960 980 Frequency (MHz) (a) (b) Figure 2: Details of the rectenna developed for data reception and power harvesting. (a) Antenna connected to the rectiﬁer, with a euro coin for reference. (b) Antenna S parameters from 885 MHz to 985 MHz. out Matching network Stage 1 Stage 2 Stage N Figure 3: Full-wave, N-stage rectiﬁer, Greinacher topology used. Amplitude (dB) 4 Wireless Power Transfer Table 1: Component values chosen for the matching network +j1 between the antenna and the rectiﬁer. +j2 +j0.5 Stage (s) Value (nH) 1 2 3 4 L1 13 5 18 Short +j5 +j0.2 L2 6.2 24 68 Open L3 10 24 12 27 L4 91 13 9.1 7.5 +j30 0.0 ∞ –j30 )e unloaded maximum voltage (open voltage) and maximum DC output power (available power) under full sunlight (L> 2000 lux) are 3.5 V and 1 mW for the small –j0.2 –j5 panel and 4 V and 2.5 mW for the square panel. In an indoor environment featuring windows (600< L< 1500 lux), the open voltage ranges from 2.8 V to more than 3 V, and the power is between 500 μW and 1.5 mW for the square panel –j0.5 –j2 and between 250 μW and 600 μW for the small panel. In an –j1 oﬃce with only artiﬁcial light (L< 500 lux), the square panel still yields more than 50 μW and has an open voltage of more 1 stage than 2 V. Yellow LEDs provide a larger open voltage and a 4 stages higher output power, as they are more spectrally rich than Figure 4: )e S parameters of the 1-stage and 4-stage rectiﬁers, white LEDs [22]; however, the diﬀerence is not signiﬁcant. from 700 MHz to 1.1 GHz, without a matching network, using Typical illumination levels were sourced from [23]. Z � 50Ω. 2.3. Power Management RF RF IN OUT L3 L1 2.3.1. Rectiﬁers. Rectiﬁcation eﬃciency is dependent on L2 L4 load, and therefore it is not appropriate to directly supply digital circuitry, which may have diﬀerent power con- sumption proﬁles, when the rectiﬁer operates. )is issue is remedied through the use of Maximum Power Point Tracking (MPPT), which loads the rectiﬁer appropriately, Figure 5: )e “double-L” matching network topology used to permitting an extraction of the most amount of power adapt the rectiﬁers. possible at any given time [24]. With the use of a boost converter, MPPT can be used to store energy into a ce- ramic capacitor-based storage element, thus providing adequate power when required. Ceramic-based capacitors Figure 9 shows the amount of available output DC are used for rectiﬁer energy harvesting, due to their ex- power, P , of a 1-stage rectiﬁer, as a function of the input OUT tremely low leakage proﬁle, in consideration of the power, P , with an optimal load applied for all power levels. IN Available power ranges from approximatively 200 nW at amount of energy that can be harvested from RF energy sources [25]. −30 dBm up to 450 μW at 0 dBm. To deliver 1 μW, a single stage rectiﬁer requires an input power level of −21 dBm. A bq25570 power harvesting IC was chosen for this purpose, as it provides MPPT, a boost converter, and a buck converter with a variable output voltage and has a low 2.2. Solar Panels. Solar energy is used as an alternative quiescent current. Figure 13 shows the general operation and source of energy to complement RF harvesting when it is wiring diagram used for this work with the bq25570. possible. Solar harvesting signiﬁcantly increases the amount )e bq25570 has several diﬀerent modes of operation. If the main storage element (C of energy available, enabling the use of an RF transmitter to ) is depleted, it needs to BIGSTOR provide data back to the requester. be charged to a certain threshold (V ) before operation CHGEN Two solar panels, a small one with circle shape and a of the MPPT and the boost converter, through the use of a large one with square shape, have been considered for the “cold-start” charge pump, with low eﬃciency (approx- system, Figures 10(a) and 10(b), respectively. )e opera- imatively 5%). )e storage element is then charged to an tional areas of these panels are, respectively, 13.9 cm and acceptable voltage level, in order to allow the main parts of 30.5 cm . )e open voltage, Figure 11, and available out DC the IC to operate. In order to avoid rapidly regressing into a power, Figure 12, of the solar panels are characterized in a cold-start state (especially due to the activation of the load), controlled environment using white and yellow LEDs as once the desired voltage is reached, a storage element illumination sources, in order to simulate realistic illumi- threshold indicator (with hysteresis), V , is provided by BATOK nation conditions for the scenarios of application. the IC. It is used in this work in order to turn on or oﬀ the 0.2 0.5 30 Wireless Power Transfer 5 0 0 –5 –5 –10 –10 –15 –15 –20 –20 –25 –25 –30 –30 –35 –35 880 900 920 940 960 980 880 900 920 940 960 980 Frequency (MHz) Frequency (MHz) –15 dBm –15 dBm –25 dBm –25 dBm –35 dBm –35 dBm (a) (b) Figure 6: )e S parameters of the 1-stage (a) and 4-stage (b) rectiﬁers, from 870 MHz to 970 MHz, matched, using Z � 50Ω, for varying 11 0 input power levels. –30 –25 –20 –15 –10 –5 0 Input power (dBm) 1 stage 3 stages 2 stages 4 stages Figure 7: Measured rectiﬁcation sensitivity S versus input power P , for power levels ranging from −30 dBm to 0 dBm. IN supply to the digital section of the node if power is scarce. cold-start state should there be an interruption in the RF )e minimum storage voltage required to power on the energy source, in order to preserve an entry point into the system is 2.5 V, and the hysteresis shutoﬀ is deﬁned at 2.2 V. high eﬃciency boost mode, when power is restored. )e values of the resistors to program the device are chosen In order to perform MPPT, the bq25570 periodically using [26]. samples the sensitivity, S, of the 1-stage rectiﬁer through the Although V is approximatively 1.7 V, the values R /R divider bridge and applies a load according to the CHGEN OC1 OC2 for the hysteresis circuit were programmed higher, to take voltage present between R and R (S ). )e OC1 OC2 MPPT into account the possibility of the absence of power and to measured eﬃciency, reported in Figure 7, depending on the insert a security margin to avoid regression into a “cold- optimal loads from −30 dBm to 0 dBm, was determined for start” state. )e minimum storage voltage required to power MPPT. From these loads, the optimum voltage levels that on the system was set at 2.5 V, and the hysteresis shutoﬀ set should be delivered to the bq25570 through the voltage at 2.2 V. Although, theoretically, these values could both be divider are calculated. )is is presented as the ratio of S / MPPT equal to V to ensure maximum energy usage if power S in Figure 14. It can be seen that a value of around 35% CHGEN should become unavailable, the limits were chosen to be presents optimal power transfer, especially at lower input higher than the former in order to avoid a regression into a power; therefore, R and R were set accordingly. OC1 OC2 Amplitude (dB) Sensitivity (V) Amplitude (dB) 6 Wireless Power Transfer 4 4 ×10 ×10 5.5 5.5 5 5 4.5 4.5 4 25 4 3.5 3.5 3 3 2.5 2.5 2 2 1.5 1.5 1 1 0.5 0.5 –30 –25 –20 –15 –10 –5 0 –30 –25 –20 –15 –10 –5 0 Power input (dBm) Power input (dBm) (a) (b) Figure 8: Measured rectiﬁcation eﬃciencyƞ versus input power P and load. (a) )e eﬃciency ƞ for the 1-stage rectiﬁer. (b) )e eﬃciency IN ƞ for the 4-stage rectiﬁer. –2 –30 –25 –20 –15 –10 –5 0 Power input (dBm) Figure 9: Measured power output P versus input power P , for the 1-stage rectiﬁer, for input power levels ranging from −30 dBm to OUT IN 0 dBm. 3.2cm 5cm 4cm 5.5cm (a) (b) Figure 10: )e solar panels with a 1-euro coin for reference. (a) )e small solar panel. (b) )e large solar panel. 2.3.2. Solar Panels. Similarly to rectiﬁers, solar panels also bq25570, with the exception that it does not have an inte- require the use of MPPT in order for the maximum available grated buck converter and therefore cannot provide a reg- power to be extracted from them. For this purpose, a ulated output voltage to a load. Figure 15 shows the general bq25504 power harvesting IC was used. It is similar to the operation and wiring diagram used for this work with the Load (ohms) Power out (μW) Efficiency (%) Load (ohms) Wireless Power Transfer 7 3.5 2.5 1.5 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 11: Measured open voltage of the two solar panels, depending on both the luminosity available and the type of LED lighting. Supermarkets commercial areas Offices Full daylight Homes and above Dark Overcast day areas 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 12: Measured power output, P , of the two solar panels, depending on both the luminosity available and the type of LED lighting, OUT with illustrations of diﬀerent luminosity levels, L. bq25504. )e storage element chosen for solar power har- )e same voltage thresholds and hysteresis values were vesting is a 470 mF supercapacitor, which can store con- used as in the previous subsection. Likewise, the MPPT ratio siderably larger amounts of energy than ceramic-based S /S can also be inferred from measured data, Figure 16. MPPT capacitors. However, these elements do have a higher It can be seen that a value of around 70% presents optimal leakage current [13, 27]. power transfer, with a variation of around plus or minus 5%, In order to combine both RF and solar energy together, a depending on the luminosity and solar panel type. BAT17-04 Schottky diode [28] was introduced between the supercapacitor serving as a storage element for the bq25504 and the ceramic capacitor array serving as a storage element 2.4. Energy Storage for the bq25570. )e use of a Schottky diode permits a completely passive 2.4.1. Rectiﬁers. )e bq25570 has two main modes of solution permitting energy to be transferred from the solar charging operation: “cold-start” (CS) mode, where a charge energy circuit to the RF energy circuit, while avoiding a pump with low eﬃciency charges up the main storage el- signiﬁcant reverse drain of energy from the RF energy circuit ement to about 1.7 V (VCHGEN), and a “warm-start” (WS) if no solar power is available. )e low threshold voltage of this mode, where the MPPT and boost converter are active and diode also makes it possible to avoid signiﬁcant energy loss charge the main storage element up to a conﬁgured voltage during energy transfer. Similar techniques have been used in (4.0 V) [10]. )e combination of both the CS and WS modes other work with combined RF/solar harvesters [29, 30]. has been deﬁned as “full-start” (FS) mode. Available power (μW) Sensitivity (V) 8 Wireless Power Transfer MPPT To storage C C STOR BYP configuration element R R OC1 OC2 V V V OC SAMP STOR BAT V V IN BOOST From solar panel C L REF SAMP IN L 2 V BUCK OUT Boost controller OUT REF To node V C SS OUT Buck VCC supply controller MPPT SS IN DC Cold start Nano-power management nEN OUT EN To μC V bq25570 BAT_OK RF harvesting R R OK3 OV2 OUT2 On/off hysteresis buck output voltage and battery level OK2 configuration OV1 OUT1 OK1 Figure 13: Schematic describing the bq25570 as used in this work, used for RF power harvesting and voltage regulation for the load (adapted from [10]). –30 –25 –20 –15 –10 –5 0 Power input (dBm) Figure 14: Optimum MPPT calculation, in order to size R and R to extract maximum power from the 1-stage rectiﬁer. OC1 OC2 CS mode requires a minimum of 15 μW to be able to when the storage element reaches 2.5 V. Discharge can be charge the storage element to VCHGEN, leakages not- seen from 650 s to about 1300 s, during which the bq25570 withstanding. From Figure 9, it can be determined that a provides power to the digital section of the node until reaching the shutoﬀ voltage of 2.2 V. minimum RF power input of −13.1 dBm is required to bootstrap the system if the main RF energy storage element )e amount of time taken to fully charge the system from is empty. A typical charge and discharge curve of the system an empty ceramic capacitor storage element of 470 μF was during normal operation was measured, Figure 17, without measured as a function of RF power input in Figure 18. No active communications. )e digital section of the node’s communications were active during this time. output is regulated to 1.8 V. Charge times in CS and WS modes increased quasi- )e CS phase of the bq25570 can be seen from 0 s to linearly with the size of the ceramic capacitor storage ele- 500 s, and WS phase can be seen from 500 s to 620 s, with the ments. Discharge times were also measured and determined output to the digital section of the node turning on at 550 s, to be 425 s, 971 s, 1810 s, and 3505 s, respectively, for the % Open voltage (%) RDIV BAT OV OK PROG OK HYST OUT SET Wireless Power Transfer 9 MPPT C C To storage STOR BYP configuration element OC1 OC2 V V OC SAMP STOR V L IN BOOST From solar panel C REF SAMP IN Boost controller REF SS MPPT AV IN DC SS Cold start Nano-power BAT_OK To μC management OT PROG bq25504 solar harvesting R OK3 UV2 OV2 On/off hysteresis and R OK2 battery level OV1 configuration UV1 OK1 Figure 15: Schematic describing the bq25504 as used in this work, used for solar power harvesting (adapted from [11]). 0 500 1000 1500 2000 Luminosity (lux) Small panel, white LED Large panel, white LED Small panel, yellow LED Large panel, yellow LED Figure 16: Optimum MPPT calculation, in order to size R and R for the extraction of maximum power from the solar panels. OC1 OC2 2 2 470 μF, 940 μF, 1860 μF, and 3720 μF ceramic capacitor- C 􏼐V − V 􏼑 BIGSTOR INIT FINAL (3) based storage elements. t � , 2P CONSUMPTION )e quasi-linearity of the charge and discharge times is logical, as the device is operating at a constant power dis- charge. )is is because, during sleep mode, the node exhibits where t is the amount of time taken to discharge capacitor static, unvarying power consumption, and the bq25570’s C , between the initial and ﬁnal voltages BIGSTOR quiescent current varies little with voltage, so charge and V and V , respectively. P is the power CONSUMPTION INIT FINAL discharge times will tend to obey: consumption that is applied to the capacitor. Percent of open voltage (%) RDIV BAT OV OK PROG OK HYST BAT UV 10 Wireless Power Transfer 0 200 400 600 800 1000 1200 1400 1600 1800 2000 Time (s) Storage voltage Digital section voltage Figure 17: bq25570 charge and discharge cycle, with a 470 μF ceramic capacitor storage element, and P � −10 dBm during charging. IN 2.4.2. Solar Panels. )e bq25504 also has two modes of several days, minimizing the leakage current, assuming no operation, the CS and WS modes, similarly to the bq25570. communications were present. Like the bq25504, CS mode on the bq25570 requires at least 15 μW to be able to charge the storage element in CS mode. However, when using a supercapacitor, the leakage current 2.5. Communications cannot be neglected [27]. 2.5.1. Node Reception. As this system is designed to con- )e initial but worst-case leakage current of such devices can be as high as 100 μA (whether charged or discharged), sume as little power as possible, the use of a standard RF and therefore the CS requires approximatively 20 μW of receiver is not feasible due to the high-power consumption power in order to be able to charge the bq25504 to WS mode of such devices which requires at least a couple of mW to [11]. )e leakage current, though initially high, eventually operate [31]. )erefore, the node is always in deep sleep stabilizes to less than 1 μA after several days. )is can be mode when not being solicited, and reception (RX) com- observed through the rate at which the voltage drops, munications are ﬁrst processed using a 4-stage rectiﬁer Figure 19. circuit to perform the envelop detection of an incoming )is exhibits the worst-case scenario of self-discharge, as OOK, or ASK, modulated RF signal at 935 MHz. it was charged instantly, using a standard power supply for )e wake-up radio demodulator circuit is presented in approximatively 5 minutes. It should be noted that this Figure 21. It provides a good trade-oﬀ between robustness and low-power consumption to demodulate data while leakage current does diminish after several days down to about 1 μA to 2 μA, due to the charging characteristics of minimizing false wakeups. Indeed, amplitude modulation such devices [27]. schemes are prone to large and abrupt variations of the A charge and discharge cycle of the supercapacitor dynamic signal due the environment of propagation, which through solar energy was measured over a period of 100 contribute to corruption of the demodulation process. hours, Figure 20, using the large solar panel, where the )e main component used for data demodulation is a environmental luminosity varied between 100 lux during TS881 comparator, chosen for its low quiescent current. V IN nighttime hours and 1000 lux during daylight hours. is the output of the 4-stage rectiﬁer. R and C form a data 2 3 It can be seen that the initial charge time in CS mode is very slicer (low pass ﬁlter), which is used to demodulate incoming slow, due to the ineﬃciency of the charge pump in the power Manchester-encoded data. )e RC time constant was chosen management IC, but is still possible, even if only 100 lux is to be 5 times the baud rate in order to provide suﬃcient permanently present. Once the WS mode is active, the charging settling time while remaining adaptable to varying power of the supercapacitor becomes signiﬁcantly faster, due to the levels. activation of MPPT, which has a much higher eﬃciency C and C , associated with R /R /R and R /R /R , re- 1 2 1 3 4 2 5 6 (measured at least 70%). From Figure 20, it can be noted that, as spectively, form high-pass ﬁlters in order to avoid false data long as there is enough luminous energy to charge the capacitor detection which may arise from slow power level variations. to at least its previous voltage before the next luminous cycle )e “+” and “− ” input nodes of the comparator are polarized (i.e., work hours or daytime), the operation of the node may be to half the supply voltage through R /R and R /R , thus 3 4 5 6 maintained indeﬁnitely in warm-start mode, which can permit ensuring proper operation of the comparator. More spe- the use of the node in low-luminosity environments due to the ciﬁcally, the “+” node is slightly biased below the “−” node to increased eﬃciency of WS mode. avoid chattering, at the cost of decreased sensitivity. Measurements indicate that the minimum amount of In the case of ASK modulation, data is sent along with luminosity required to keep the node functional without power, whereas OOK modulation implies that the system is discharge is 70 lux, after the supercapacitor was charged for not powered during communications. RX communications Voltage (V) Wireless Power Transfer 11 –14 –12 –10 –8 –6 –4 –2 0 Rectiﬁer input power (dBm) Cold start Warm start Full start Figure 18: Time taken for the bq25570 to accomplish cold, warm, and full start using RF energy only, as a function of input power. 02 468 10 12 Time (days) Figure 19: Self-discharge curve of the 470 mF supercapacitor over several days, showing the voltage present. 470 mF storage - charge and discharge Daylight hours 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Time (days) Storage voltage Digital section voltage Figure 20: bq25504 charge and discharge cycle, with a 470 mF supercapacitor storage element, with the luminosity varying between 100 lux and 1000 lux. are done at a speed of 800 bauds, but the algorithm in the it avoids long sequences of binary zeroes and ones which can microcontroller can be adapted to run from 40 bauds to 25.6 induce demodulation errors in the data slicer. It does, kilobauds. however, increase the spectrum usage for an equivalent )e frame format of messages sent to the node was number of sent bits. implemented using Manchester encoding, with a synchro- )e synchronization subframe length, s, was chosen to nization preamble, Figure 22. )is form of encoding pro- be 16 bits, in order to allow for proper wake-up generation vides several advantages. First, the synchronization and baud rate detection. )e actual message subframe preamble can be used to lock onto the baud rate with a length, n, is two bytes long (m � 16 bits). )e ﬁrst byte of the certain error tolerance (clock reconstruction). Furthermore, message subframe is the address of the node, which permits Time (s) Voltage (V) Voltage (V) 12 Wireless Power Transfer VCC VCC the basic packet format was used in order to minimize overhead on the microcontroller. )e frequency at which the C R 1 1 data was emitted is 868 MHz. )e frequency deviation for Rectiﬁer Data the modulation is 19 kHz. )e output power level chosen for CC output data transmission is −30 dBm. A transmission of the basic C R packet was collected via SDR and demodulated in order to C D 2 2 4 2 WUP show the digital data sent, using the default settings of the R D C R 3 transmitter, Figure 25. 6 1 5 7 )e packet is split into 4 main parts: the preamble, which is 64 bits long; the sync word, which is 32 bits long; the Figure 21: Comparator circuit for data reception; wake-up signal message, which is 3 bytes (or 24 bits) long; and the CRC generation. checksum, which is 1 byte (or 8 bits) long. )e packet format can be modiﬁed through registers present in the module, and additional ﬁelds can be added, depending on the usage an addressing of up to 256 nodes, and the second byte is a scenario, which includes addressing capabilities, multi- command, subdivided into two nibbles. )e ﬁrst nibble packet, or variable-length messages [17]. )e demodulation indicates the operation, and the second contains an optional argument for the microcontroller to process. )e frame is of the received packets is performed on the same device transmitting data to the node. )e supplied software is used therefore 33 bits long. )e RF signal is ﬁrst rectiﬁed by the 4-stage rectiﬁer, to receive data packets. An example of packet reception, containing temperature data as the third byte, is shown in Figure 23, where it is then available for processing by the comparator and wake-up circuit. )e wake-up detector has Figure 26. been adapted from [32]. It is designed to allow data rates above a certain fre- 2.6. Functionality. In order to use harvested energy and to quency to activate the wake-up circuit, which makes it manage the concurrent operation of diﬀerent modules, the robust against low-frequency, brutal changes in power levels data demodulation capabilities and intelligent functions of or against external interference from other communications this sensor node are provided by a PIC16LF1559 micro- in the same band, which may be interpreted by the com- controller, with an integrated temperature sensor [14]. parator as data. Storage capabilities are implemented with an FRAM IC, )e output of the comparator is a normalized repre- using the SPI bus [15]. sentation of received binary data to V (1.8 V). )is is used CC )e PIC spends most of its time in sleeping mode and in two ways: to generate a wake-up signal with C , D , D , C , 4 1 2 5 only wakes up when an event is received from the com- and R (WUP) and also give data to the microcontroller parator/data slicer module. )e Manchester decoding al- (DATA) to be processed when the wake-up signal is gen- gorithm is implemented in a purely software-based manner, erated, Figure 24. using interrupts and timer loops. )e algorithm can be )e sensitivity of the circuit was measured, and the node adapted for diﬀerent synchronization formats, message messages could still be properly processed by the micro- sizes, and data transmission speeds and is presented in controller with power input down to −31 dBm at the input of Figure 27. Once a message is received, it is then processed the 4-stage rectiﬁer circuit. and, if successful, an action is performed, as illustrated in Figure 28. Table 2 presents a brief overview of currently imple- 2.5.2. Node Transmission. Transmission (TX) communica- mented commands that can be sent to the node with the tions from the node are done using a low-power radio basic, 2-byte frame format. )e PIC only processes these transmitter IC, the SPIRIT1. )is IC is capable of sending commands if the ﬁrst byte of the frame matches its pro- data with diverse modulation formats, including OOK, ASK, grammed address. FSK, and GFSK. Power levels used to transmit the data can be conﬁgured from 11 dBm down to −30 dBm. )e baud rate of data transmission can also be conﬁgured from 1 kbps up to 500 kbps. 2.7. Assembled System. )e diﬀerent modules featured in the node were independently manufactured on 2- or 4-layer FR- It can use either direct modulation or a standardized packet format. Using the standardized packet format, data 4 substrate and soldered. )e node, Figure 29, is further transmission integrity can be veriﬁed using a Cyclic Re- enhanced by using modular signal connectors. dundancy Check (CRC). )e Advanced Encryption Stan- )is modular, multiboard approach allows the experi- dard (AES) can also be used in order to secure data mentation with various assemblies of diﬀerent subcircuit transmission. )e module also includes a Carrier Sense generations, in order to evaluate their performance inde- Multiple Access (CSMA) algorithm, which can be exploited pendently. )e presence of connectors also facilitates easy determination of power consumption by inserting mea- to detect if another module or device is already transmitting, thus avoiding packet collisions [17]. surement devices between the pins and permits easy debugging of various modules with measurement For the purposes of this work, 2-GFSK modulation was selected in order to transmit data, at a rate of 38.4 kbps, and instrumentation. Wireless Power Transfer 13 Byte 1 Byte 2 Byte n Synchronization Sync stop Data frame (s bits) bit (m bits) (1 bit) Figure 22: )e frame format for the node’s low-power wake-up radio, used in RX communications. 1.5 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) Figure 23: Data frame sent to node, after rectiﬁcation. 1.5 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) DATA WUP Figure 24: Data frame normalization to 1.8 V and wake-up signal generation for the microcontroller. –1 –2 –3 Sync (32 bits) CRC (8 bits) –4 Message (24 bits) Preamble (64 bits) –5 0 500 1000 1500 2000 2500 3000 Time (μs) Figure 25: Standard “basic” data frame from the low-power TX module. Voltage (V) Voltage Frequency deviation (abitrary) 14 Wireless Power Transfer Figure 26: Reception of data from the node. )e third byte is ambient temperature as measured by the microcontroller, in hexadecimal ° ° format, between 25 C and 27 C in this case. )e ﬁrst two bytes are “T?” in ASCII. WUP interrupt Abort Process RX generated RX message Receive sync bits Yes Required Required No Timeout Timeout No message sync bits bits received? received? Yes Average sync bit time for clock reconstruction Receive message bits using reconstructed Wait for sync stop clock time bit No Timeout Sync stop bit? Yes Figure 27: Software algorithm used in order to decode the Manchester-encoded message received by the node. Yes Perform ADC Get Initialize PIC and conversion and temp.? peripherals store in RAM No Output Yes Sleep Say temperature in temp.? RAM on test point No Wake- No up? Yes Store Store temperature Yes temp.? in FRAM Receive data frame No Yes Retrieve Retrieve temperature from No Correct temp.? FRAM address? No Yes Yes Send temperature TX? Process command via TX module No Figure 28: Software algorithm used after reception of a correct message, showing a sample of available commands. Wireless Power Transfer 15 Table 2: Commands available within the node. Byte Description of command instructions 0 × F0 Sample the temperature of the node 0 × F1 Show the temperature on a test point 0 × F2 Turn on the FRAM power supply; make the FRAM sleep 0 × F3 Turn oﬀ the FRAM power supply 0 × F4 Turn on the comparator power supply 0 × F5 Turn oﬀ the comparator power supply 0 × F6 Show the command on a test point 0 × F7 Reset the temperature in the node to 25 C 0 × F8 Send the node temperature via the TX module 0 × FB Carry out sanity check of FRAM and TX modules (verify device IDs) 0 × Dn Store the temperature in FRAM address n 0 × Cn Retrieve the temperature in FRAM address n 1,2 0 × 3n Set the high nibble of the temperature byte in RAM with n 1,2 0 × 4n Set the low nibble of the temperature byte in RAM with n 1 2 )ese commands are used for debugging purposes. For these commands, n is a variable, ranging from 0 × 0 to 0 × F. Large solar panel Antenna (data RX) Microcontroller and FRAM Comparator Power management wake-up (stacked) Power and data Energy storage rectiﬁers capacitors Antenna (power harvesting) Antenna (data TX) RF transceiver Figure 29: )e whole system, with all modules present. A 1-euro coin is provided for reference. the voltage delivered to the digital subsection and is con- 3. Power Consumption Analysis ﬁgured at 1.8 V. A simulator that is constructed from )e estimated power consumption of the overall system is measured data is then presented. necessary in order to determine the viability of applicative scenarios of usage of the node. Indeed, it may be necessary to limit certain power consuming features of the node (notably 3.1. Analog Modules the broadcasting of measured data) when limited power is 3.1.1. Comparator. )e comparator/wake-up module is the available. Typically, when RF harvesting is the only source of energy, the system must operate in a restricted mode from a device that consumes the most power in sleep. )is is power consumption perspective. primarily for two reasons. Firstly, in order to polarize the )is section fully characterizes the power consumption inputs of the comparator to (V /2), after the high-pass CC of the diﬀerent modules of the node, depending on their ﬁlter, Figure 20, the resistor network uses approximatively phase of operation. )e modules are categorized as digital 180 nA, or 324 nW. Secondly, the comparator itself uses and analog modules, as digital modules have more phases of 220 nA, or 396 nW. Additional current used, when the operation than analog modules, with the latter having more WUP and DATA pins change due to the reception of data, predictable and static power consumption over time. VCC is was measured to be about 10 nA to 20 nA at most, or 16 Wireless Power Transfer a message for 40 ms, it consumes 16.6 μJ to receive 33 bits (a 18–36 nW. )e module therefore consumes, at most, about 850 nW of power, continuously. message of 16 bits and a synchronization subframe of 17 bits), which is approximatively 503 nJ per bit received. 3.1.2. Power Management ICs. )e bq25504 and bq25570 must also be considered when doing a power consumption 3.2.3. TX Module. )e data TX module must be operated in analysis. During CS mode, the necessary minimum amount a particular order so as to ensure data transmission. It must of power required to start up both devices is 15 μW, below be turned on, reset, and then set into its ready operating which the device will never manage to enter WS mode, due state. Afterwards, the Voltage Controlled Oscillator (VCO) to the ineﬃciency of the charge pump, which is approx- operating within the integrated PLL must be calibrated, after imatively 5%. which it must also be returned to a ready operating state. Once in WS mode, the power consumption of the de- After this initialization phase, data may be sent to the vices varies relatively little, and it was measured, with no transmitter buﬀer, and the device may send data. At the end power on the input, to be around 500 nA, independently of of the transmission, it returns into ready state and may then voltage present on the storage elements. Depending on be shut down. storage element voltage, this can range from a power con- )e device in shutdown mode consumes 2 nA of current, sumption of 2 μW down to 850 nW. )e eﬃciency of the or 2.6 nW. In ready mode, it consumes 430 μA, or 774 μW. buck converter, Figure 30, is measured with the storage When calibrating its VCO, it consumes 6.72 mA, or element fully charged, using the methodology provided in its 12.33 μW. Finally, when transmitting data, it consumes datasheet [10]. 9.58 mA, or 17.56 mW. Buck converter eﬃciency exceeded 80% for load currents As the module sends messages at 38.4 kbps and con- above 10 μA. However, for currents between 1 μA and 10 μA, sumes 17.56 mW while sending a message for 3.75 ms, it eﬃciency dropped down to 60%, and for load currents below consumes approximatively 65.9 μJ to transmit 128 bits, 1 μA, the eﬃciency is approximatively 20% or less. which is 514 nJ per bit sent during the actual transmission state. If VCO calibration and module setup are to be in- cluded, then the power consumption per bit increases to 3.2. Digital Modules 592 nJ per bit. 3.2.1. FRAM. )e FRAM IC has several modes of operation: read mode, write mode, sleep mode, and idle mode. In sleep 3.3. Module Power Consumption. )e modules comprising mode, current consumption is measured at 30 nA, or about the node are split into two categories: analog modules and 55 nW. During this mode, the FRAM is not available to digital modules. Analog modules generally present an in- process commands and must be woken up into idle mode in variant power consumption rate whatever the node’s op- order to function. Wake-up time is measured to be 15 μs. erating state is, though the warm- and cold-start modes of During idle mode, the FRAM module consumes 4 μA of the power management ICs are exceptions. current or 7.2 μW. In this mode, the FRAM module is ready Table 3 presents an overview of the power consumption to accept read and write commands. Reading data from one of the analog modules. )e power consumption of the power address consumes 2.13 mA, or 3.83 mW of energy. Writing management ICs is variable, as they are supplied by the data consumes 2.32 mA, or 4.17 μW of energy. )e power storage elements, whose voltage may vary from 1.8 V to 4 V consumption of FRAM modules is signiﬁcantly lower than in warm-start mode. ﬂash devices [33]. Table 4 presents an overview of the power consumption of the digital modules, depending on their diﬀerent states of operation. As they are supplied by the power management 3.2.2. Microcontroller. )e PIC microcontroller, similarly to modules with a constant voltage of 1.8 V, their consumption the FRAM, has several modes of operation: sleep mode, only varies depending on the state of operation. message receiving mode (which can be further separated into false wakeups and full messages), data processing mode (which includes SPI transactions), and ADC measurement mode. 3.4. Node Operation. Excluding the ineﬃciencies and qui- )e power consumption in sleep mode is measured to be escent current consumption of the power management 40 nA or 72 nW. In message receiving mode, whether due to modules, the static consumption of the node in sleep mode is false wakeups or actual messages, the consumption of the deﬁned as the sum of the comparator, FRAM, micro- microcontroller is the same, at 230 μA, or 415 μW, though controller, and transceiver power consumption in sleep the time spent processing these diﬀerent modes is diﬀerent. mode. All other power consumption states are a combination of In data processing mode, the power consumption is 410 μA, or 738 μW, and, ﬁnally, in temperature measurement mode, the modules’ operating states as shown in Tables 3 and 4, whose operation is organized by the microcontroller as fast the power consumption is 800 μA, or 1.44 mW. Another metric that is of importance is the consumed as possible, in order to minimize power consumption. )e energy per bit received. As the microcontroller receives average power consumption of the node while performing messages at 800 bauds and consumes 415 μW while receiving select commands was measured and is presented in Table 5. Wireless Power Transfer 17 –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Load current (μA) Figure 30: Eﬃciency of the buck converter of the bq25570, as a function of output current. Table 3: Power consumption of the node’s analog modules. node performance has negligible power consumption in comparison to sending data, Figure 31. Device State Current (μA) Power (μW) WS/FS 500 850–2000 bq25570 CS — 15 3.5. Simulator. )e power consumption simulator is derived WS/FS 500 850–2000 bq25504 from power consumption data measured in situ. )e sim- CS — 15 ulator uses an iterative approach to determine power con- Comparator Static 0.42 0.756 sumption over a small time interval,Δt. Once the calculation of the power consumption during Δt is complete for all elements, the states of operation of each device are updated. Table 4: Power Consumption of the node’s digital modules. )e general equation modeling the consumption of any given module of the node in the simulator is shown below: Device State Current (μA) Power (μW) Sleep 0.04 0.072 P(t) � P(t −Δt) + M(Δt, S), (4) Message RX 230 423 PIC Processing 410 738 where P is the power consumption, or generation of any Temp. measurement 800 1440 module in the node, M is the change in power consumption False wakeup 230 423 over the previous Δt time increment, and S is the current Sleep 0.03 0.054 state of operation of the node. )is includes any changes in Idle 4.00 7.200 solar and RF levels, power storage element levels, and any FRAM Storage (read) 2130 3834 operations undertaken by the node that may modify the Storage (write) 2315 4167 power consumption of the node. )e changes in power Sleep 0.02 0.036 consumption are then summed or subtracted, depending on Ready state 430 774 whether the device is a power consumer or generator, in a SPIRIT1 VCO calibration 6720 12334 chain that corresponds to power ﬂow. TX (−30 dBm) 9580 17558 An overview of the simulated algorithm is presented in Figure 32: Static power consumption of the node is 920 nW, or less than a microwatt. Most messages have similar power 4. Scenarios of Operation consumption proﬁles, around 17 μJ per message, including any operation which uses the diﬀerent modules on the Several applicative scenarios were envisioned for this type of node, due to the speed at which the operations are device. Scenarios have been classiﬁed according to the type conducted. of power mainly used to power the node, being that of RF )ere are, however, two exceptions to this generaliza- energy or of solar energy. Using data derived from part III of tion. )e ﬁrst one is when the node receives a false wakeup. this work, simulation of expected results was theoretically Most received messages are interfering signals, which are not calculated. Measured data is also presented and compared to theoretical predictions. OOK/ASK modulated, or use an incompatible frame format. With this in mind, the message reception code was designed )eoretical predictions of power consumption were to time-out after approximatively 16.8 ms if an incorrect estimated by taking into account the leakage current of message is received. storage devices, the power consumption of the power )e second exception is when the node is asked to management modules, and their ineﬃciencies, especially in transmit data, as this is a time and energy consuming op- regard to the low performance of the bq25570’s buck eration, during which the node consumes about 97.8 μJ of converter. )ese values were integrated into a program energy. When transmitting, however, other operations, the capable of simulating the power consumption of the node Efficiency (%) 18 Wireless Power Transfer Table 5: Average power consumption of the node during diﬀerent phases of operation. Command Duration (ms) Average power (μW) Energy used (μJ) Sleep mode — 0.92 — False wakeup 16.8 416 7.00 Temperature measurement 40.6 422 17.1 FRAM data storage 42.1 420 17.7 FRAM data retrieval 42.2 417 17.6 Data transmission 55.9 1750 97.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Time (ms) (a) 0.4 0.2 0 5 10 15 20 25 30 35 40 45 50 Time (ms) (b) 0.5 0 5 10 15 20 25 30 35 40 45 50 Time (ms) (c) 0 102030405060 Time (ms) (d) Figure 31: Power consumption of the node for diﬀerent commands. (a) )e power consumption when the node is idle. (b) )e power consumption when a message is received. (c) )e power consumption when the FRAM is solicited for a read or write operation. (d) )e power consumption when a data TX command is issued. Power (mW) Power (mW) Power (mW) Power (μW) Wireless Power Transfer 19 can be performed (maximum measurement speed) without Input scenario the main storage capacitor discharging, which will depend information on the power input to the rectiﬁer. Fixed Scenario )is scenario will be played out in two parts: for the ﬁrst conﬁguration parameters Initialize all devices part, the node will operate in RX-only mode, where the only instructions it will receive are that of data measurement and Calculate Calculate storage; during the second part, in addition to RX mode, it Variable consumer’s power generator’s power parameters will also operate in TX mode. usage creation Calculate power 4.1.1. RX-Only Operation. )e calculated and measured ﬂow balances speeds at selected points of this metric are reported in Figure 33. Update all states ∆t increment At measurement intervals of 10, 50, 100, 500, and 1000 seconds, the node requires, respectively, −6.2, −11.3, −12.5, Yes Store simulation Plot −14.2, and −14.5 dBm of input power in order to remain End? results results operational. It can be seen that as the frequency of measurements No becomes increasingly close to the theoretical 6.66 Hz Generated data maximum possible, the amount of power required for maintaining the system operational increases rapidly. It is Figure 32: Overview of the simulator algorithm. not possible to reach the theoretical limit of 6.66 Hz, as this would require more instantaneous energy than the rectiﬁer can provide, 10 mW, if it is supposed that over time, in order to determine the accuracy of the sim- a maximum input power of 0 dBm is available for ulations in regard to the actual power consumption. this scenario. As the interval between measurements For these scenarios, a data measurement operation shall increases, the power level trends towards the hard limit be deﬁned as the combination of a temperature measure- of −15.1 dBm. ment command with an FRAM write command. A data )e higher levels of power required as the frequency of transmission operation shall be deﬁned as an FRAM read measurements increases are likely due to cumulated inter- command combined with a data transmission command. action of ineﬃciencies (buck converter, MPPT out of range It should be noted that the maximum speed at which the as power input levels increase,. . .) in the node as power node can receive commands is approximatively at 6.67 Hz, consumption rises. or 150 ms. Although the microcontroller takes approx- )e typical distance at which the node can operate in imatively 50 to 60 ms to fully process a command (Fig- RX-only continuous RF mode ranges from 1 m to 6 m, ure 30), the wake-up radio takes approximatively 150 ms to though the amount of time between measurements increases return to a “known state,” wherein the wake-up generation with distance, in order to give enough time for the storage circuit is fully discharged (Figures 20 and 23), thereby element to recharge and return to its previous level. )e limiting the speed to a maximum of 6.67 requests per recharge time for 1, 2, 5, and 6 m distances was measured to second. be 10, 21, 344, and 1505 s, respectively. Operation at 10 m was also observed as a result of multipath propagation in the 4.1. Continuous RF Operation. Continuous RF operation of measurement environment. the sensor node means that it will be constantly powered, but only through RF energy, from a 28 dBm emitter, with a 3 dBi 4.1.2. RX and TX Operation. Similarly to the ﬁrst part of the patch antenna, operating at 935 MHz. )e ceramic capacitor scenario, the node will now be instructed to perform a data storage elements recharge back up to 4 V if powered with at measurement and a data transmission operation. )is will least −15.1 dBm of power input, which is the minimum reduce the eﬀective maximal frequency rate of measure- amount of power required to keep the PIC, comparator, and ments down to 3.33 Hz. )e calculated and measured speeds bq25570 operational (while sleeping and without commu- at selected points of this metric are shown in Figure 34. nications). )is corresponds to a power usage of 7 μW, of At measurement intervals of 10, 50, 100, 500, and 1000 which the bq25504 uses 2 μW. )e high usage of the digital seconds, the node requires, respectively, −1.9, −7.6, −9.7, subsection can be explained by the low eﬃciency of the buck −13.2, and −14 dBm of input power in order to remain converter for very small supply currents (Figure 30). )e operational. bq25504 will be considered to be nonoperational in this )e node is more restricted in terms of measurement mode and therefore will not consume any energy. )e speed, due to the power consumption of transmission op- storage element for this scenario is a 3720 μF ceramic ca- erations, which are nearly an order of magnitude larger than pacitor-based element. reception and measurement only operations; however, op- )e main criterion to evaluate the performance of eration is still possible with only RF power. continuous operation of the node is how fast measurements 20 Wireless Power Transfer –2 –4 –6 –8 –10 –12 –14 –16 0 1 2 3 4 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 33: Maximum measurement speed as a function of input power, with a 3720 μF ceramic capacitor-based storage element, for data measurement operations. –2 –4 –6 –8 –10 –12 –14 –16 0 1 2 3 4 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 34: Maximum measurement speed as a function of input power, with a 3720 μF ceramic capacitor-based storage element, for data measurement and data transmission operations. 4.2. Fully Autonomous Operation. As autonomous operation the energy drain is considered to be only supercapacitor self- would require at least autonomy of a day (or more); it is discharge, in combination with the quiescent current of the therefore necessary to use a supercapacitor to store the node in sleep mode. )e worst-case scenario, on the other energy required to power the circuit during sleep mode. hand, is when the PIC is nearly continuously solicited. In this second scenario, a 470 mF capacitor is charged to full capacity at 4 V, for 30 minutes, using a standard power supply, and hot-plugged into the bq25504 circuit. No power, 4.2.1. RX-Only Operation. )e calculated and measured speeds at selected points of this metric are reported in Figure 35. whether solar or RF energy, was given to the device. Both the At measurement intervals of 1 s, 5 s, 10 s, 1 min, 1 hour, bq25570 and the bq25504 operated in WS mode during the length of the scenario. and 2.8 hours, the autonomy of the node is, respectively, 12, 29, 37, 49, 53, and 53 hours. )is is close to, but slightly less )e main criterion to evaluate the performance of this autonomous operation of the node is how long it could eﬃcient than, theoretically predicted results. As measure- ment intervals increase to more than one hour, the static operate with main storage supercapacitor discharging, depending on the frequency of measurements. power consumption of the system overtakes that of the power used during normal operations, and so the autonomy Discharge time was measured to be 2.25 days with no measurements. )is is considered to be the case scenario, as varies little. Input power (dBm) Input power (dBm) Wireless Power Transfer 21 –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 35: Maximum node autonomy is a function of command frequency, with a 470 mF supercapacitor storage element, for data measurement operations. As measurements increase in frequency, the autonomy of recharge the node during the day, supposing that it is solicited during the night. In regard to the duration of a the node decreases, though with a measurement every ten seconds, the node still has autonomy of 37 hours. If the node is day, a pessimistic case of 8 hours of light and 16 hours of not solicited, the autonomy of the node is approximatively 54 darkness will be supposed. Furthermore, it will be sup- hours. )is is a hard limit which depends on the supercapacitor posed that the node will always be operating in warm- leakage and the power consumption of the whole node. start mode. Using a larger supercapacitor, or higher maximum Figure 37 shows the amount of time required for the storage element voltages, would increase the autonomy of node to charge from the beginning of warm-start mode until the node. However, larger storage elements would come at a the storage element (supercapacitor) is full, as a function of cost of longer cold-start times. luminosity, supposing that the node is not solicited. It can be observed from Figure 35 that, for the white For longer autonomy requirements, it may be necessary to implement a cold-start charge method that will rapidly LEDs, at least 155 lux is required to recharge the node in 8 hours and that, for the yellow LEDs, 135 lux is required. )is kick-start the system into warm-start mode, which may use existing energy harvesting methods (i.e., leave the node supposes that the node is fully depleted in terms of the under a very bright light or leave the node next to a high- warm-start mode but has not yet entered cold-start mode. power transmitter, until it indicates that it is operational) or In the context of this scenario, if the node is solicited at require a cable-based solution. night, then 16 hours of autonomy is required. From Fig- ures 35 and 36, it can be determined that the node must not be sent commands more than once every 1.5 seconds and 4.2.2. RX and TX Operation. )e calculated and measured every 7 seconds, respectively, for RX and TX/RX operation speeds at selected points of this metric are reported in modes, in order to stay in warm-start mode. )erefore, it is Figure 36. possible to conclude that the node may operate in envi- At measurement intervals of 1 s, 5 s, 10 s, 1 min, 1 hour, ronments with luminosity as low as 150 to 200 lux for an and 2.8 hours, the autonomy of the node is, respectively, 4, indeﬁnite amount of time, with reasonable environmental 13.5, 21, 41, 52, and 53 hours. )e autonomy of the node has sensing capabilities. decreased compared to the previous subscenario, due to the high consumption of data transmission operations. How- 5. Analysis ever, at measurement intervals of over an hour, the node still RF harvesting techniques can produce work which is presents similar autonomy. capable of generating energy with low levels of power input, such as [4], where charging of storage occurs with 4.3. Solar Operation. )e autonomy and power consump- power levels of −29 dBm, without any load present. Other tion of the node have been characterized in autonomous RF-only harvesting systems have produced similar per- mode. Solar energy is analyzed, using the large solar panel formance with this work, such as [19, 21, 34], with power and both LED types. harvesting being usable, without load, with power input As solar energy can easily provide suﬃcient energy to levels around −20 to −25 dBm. Distances at which the power the node, the main criterion for this scenario will node could operate from a 28 dBm power source ranged analyze the amount of luminosity required in order to from 1 m to 6 m. Autonomy (h) 22 Wireless Power Transfer –1 0 1 2 3 4 5 10 10 10 10 10 10 10 Measurement interval (s) Calculated Measured Figure 36: Maximum node autonomy is a function of command frequency, with a 470 mF supercapacitor storage element, for data measurement and data transmission operations. 100 200 300 400 500 600 700 800 900 1000 Luminosity (lux) Yellow LEDs White LEDs Figure 37: Node warm-start charge time as a function of luminous intensity. Combining RF and solar energy leads to diverse per- As sleep current is not negligible compared to the power formance depending on the architecture used, such as consumption of storage element leakages and power man- [29, 30, 35, 36]. Due to diﬀerent topologies and solutions agement IC quiescent currents, it is normal that perfor- used, it is diﬃcult to compare such work homogeneously; mance metrics are worse than power harvesting systems in however, it can be noted that solar energy delivers more which no load is present. energy than RF harvesting and that the performance of this Integration of a single antenna with both data and power work was similar to [30]. In regard to the previously de- reception was attempted. However, the integration was not veloped node [8], the minimal power requirements to power possible, due to interference generated by the boost con- verter used for RF power harvesting. Indeed, the constant this improved node in sleep mode are quasi-identical, as the added elements have very little sleep current, in the range of modiﬁcation of the load impedance presented to the rec- several nA. )e addition of a solar-powered energy source in tiﬁers, in order to perform MPPT, caused interference in the this work extends the autonomy of the node in environment data reception (comparator) circuit, creating a situation in where solar energy, even at ambient levels, is available. which false wakeups were continuously generated. Time (hours) Autonomy (h) Wireless Power Transfer 23 considered in order to increase the diversity of powering Furthermore, this work adds data retransmission ca- pabilities compared to the previously developed node [8]. methods, to improve redundancy in terms of power availability. )is is useful, as it makes measurement data retrieval possible without having to make any physical connections to the node. 6. Conclusion and Perspectives A custom-based interface using SDR was also created for sending commands to the node. Furthermore, this work )is work demonstrates the possibility of creating a COTS- does not only analyze metrics of the power harvesting el- based, hybrid RF, and solar-powered IoT sensor node, ca- ements of IoT nodes but develops details on the more pable of environmental sensing with a restrained power functional aspects of the node. )is includes what the node budget. )e active section of the node consumes 920 nW of can do, in terms of functionality, how it does this, and a list power in sleep mode, of which 870 nW is used for the wake- of the performance of diﬀerent modules, in order to see up circuitry. where optimization is required in order to reduce power During RF-only power operation (in the 900 MHz ISM consumption. band), the device may function with power levels down to )ese additions have also permitted the creation of more −15.1 dBm, albeit with a cold-start requirement of complex scenarios of operation for the node. Scenarios in the −13.1 dBm, with a 28 dBm transmitter capable of providing previous work [8] were limited to the amount of mea- both data and energy. surement operations that could be done with a set RF power In solar mode, the device may function with luminosity input, and the autonomy of the node was, at best, limited to levels down to 70 lux, albeit with a cold-start requirement of several hours. 100 to 150 lux, depending on supercapacitor leakage current, In this work, the node can now operate for several days which is a typical lighting situation in an oﬃce or a house. without any power source present (under the assumption )e system is fully operational, able to handle both that its storage elements are full), and the varied power communications and RF energy transfer over the same sources increase energy redundancy, should one of the band, and, charge time notwithstanding, operate over a sources of power be not available. A greater variety of period of one or two days with the presence of a super- scenarios of operation is also possible with the further ad- capacitor as a storage element. Additionally, the node is dition of a data retransmission module. capable of wirelessly transmitting back data when A power consumption simulator, which was not present requested to do so. in our previous work [8], is also developed based on )is work characterizes various operational scenarios measured performance data and is used to compare mea- that IoT nodes could be used in, giving an additional sured performance in real settings with theoretical perfor- insight into the abilities and limitations that are possible mance. )eoretical predictions and real data match closely. in a small subset of potential operating environments. Furthermore, the simulator can be used in order to create Besides cartography of diﬀerent modes of operation, at the and test other hypothetical scenarios before they are system level, a model is proposed based on the analysis implemented, which can be used to test their viability before and characterization of the power consumption of the requiring measurements to be undertaken. functions featured in the proposed hybrid node. )e )ere are several limitations in this system that could be resulting implementation exhibits a good correlation with improved upon. On rare occasions, false wakeups can also actual measurements, thus allowing for prediction of the trigger the microcontroller, which will waste energy trying to node’s performance in various environments, or other process data. )is is partly remedied by using a timeout hypothetical scenarios. counter in the software loop, to minimize time spent decoding erroneous/false messages. A diﬀerent reception Data Availability architecture or modulation scheme should be considered, as in [37]. More energy eﬃcient transceivers could also be used )e data used to support the ﬁndings of this study are in- [38]. cluded within the article. Other data used to support the Another limitation is that of the bq25570 and bq25504 ﬁndings of this study are available from the corresponding power management ICs, which are being used near their author upon request. lower limits in terms of energy input, and therefore they do not always provide highly eﬃcient energy transfer, which can easily triple the static energy consumption in sleep mode Conflicts of Interest or cause long cold-start charging times. )is could be )e authors declare that there are no conﬂicts of interest mitigated by using custom-designed power management regarding the publication of this paper. ICs, as in [39], or by using extremely low-power techniques to create equivalent replacements [36]. Future additions that could be explored for this work Acknowledgments include adding a larger variety of environmental sensors, which remains possible using the SPI bus present on the )e authors would like to thank M. De Matos for her help in microcontroller. Additionally, the use of other power har- using certain measurement instrumentation and Y. Faize for vesting methods (such as thermoelectric modules) should be the loan of his Joulescope. 24 Wireless Power Transfer [20] J. Nicot and T. Taris, “Remote RF powering of ambient References sensors,” in Proceedings of the 2016 IEEE International [1] B. Martinez, M. Monton, ´ I. Vilajosana, and J. D. Prades, “)e Conference on Electronics, Circuits and Systems (ICECS), power of models: modeling power consumption for IoT de- pp. 760–763, Monte Carlo, Monaco, 2016. vices,” IEEE Sensors Journal, vol. 15, no. 10, pp. 5777–5789, [21] K. Kotani, S. Inoue, T. Komiyama, Y. Chonan, and H. Yamaguchi, “RF rectiﬁer absolute ﬁgure of merit based on [2] A. Hussein, “Energy consumption of wireless IoT nodes,” relative comparison with reference rectiﬁer,” in Proceedings of Master’s thesis, Norwegian University of Science and Tech- the 2019 IEEE International Symposium on Circuits and nology, Trondheim, Norway, 2017. Systems (ISCAS), pp. 1–4, Sapporo, Japan, 2019. [3] J. A. Paradiso and T. Starner, “Energy scavenging for mobile [22] P. T. Tolnai and E. Racz, ´ “Preliminary spectral design of a and wireless electronics,” IEEE Pervasive Computing, vol. 4, LED- and light bulb-based sun spectrum simulator,” in no. 1, pp. 18–27, 2005. Proceedings of the 2019 IEEE 17th World Symposium on [4] M. Piñuela, P. D. Mitcheson, and S. Lucyszyn, “Ambient RF Applied Machine Intelligence and Informatics (SAMI), energy harvesting in urban and semi-urban environments,” pp. 217–222, Herlany, Slovakia, 2019. IEEE Transactions on Microwave :eory and Techniques, [23] “Recommended light levels (illuminance) for outdoor and vol. 61, no. 7, pp. 2715–2726, 2013. indoor venues,” 2019, https://www.noao.edu/education/ [5] C. Greene, “Wireless power, energy harvesting, and power QLTkit/ACTIVITY_Documents/Safety/LightLevels_ management solutions for sensors and the IoT,” 2017, http:// outdoor+indoor.pdf. www.powercastco.com/wp-content/uploads/. [24] T. Esram and P. L. Chapman, “Comparison of photovoltaic [6] B. Calhoun and D. Wentzloﬀ, “Ultra-low power wireless SoCs array maximum power point tracking techniques,” IEEE enabling a batteryless IoT,” in Proceedings of the 2015 IEEE Transactions on Energy Conversion, vol. 22, no. 2, pp. 439–449, Hot Chips 27 Symposium (HCS), pp. 1–45, Cupertino, CA, USA, 2015. [25] H. Lee, K.-C. Lee, J. Schunke, and L. Burton, “Leakage cur- [7] V. Raghunathan, A. Kansal, J. Hsu, J. Friedman, and rents in multilayer ceramic capacitors,” IEEE Transactions on M. Srivastava, “Design considerations for solar energy har- Components, Hybrids, and Manufacturing Technology, vol. 7, vesting wireless embedded systems,” in Proceedings of the 4th no. 4, pp. 443–453, 1984. International Symposium on Information Processing in Sensor [26] Y. Ramadash, “bq25505/70 design help V1.3,” 2013, http:// Networks, pp. 457–462, Boise, ID, USA, 2005. www.ti.com/lit/zip/sluc484. [8] N. John, F. )ierry, and T. Taris, “An RF-powered IoT node [27] A. S. Weddell, G. V. Merrett, T. J. Kazmierski, and B. M. Al- for environment sensoring,” in Proceedings of the 2019 IEEE Hashimi, “Accurate supercapacitor modeling for energy Wireless Power Transfer Conference (WPTC), London, UK, harvesting wireless sensor nodes,” IEEE Transactions on Circuits and Systems II: Express Briefs, vol. 58, no. 12, [9] Broadcom, “HSMS-285x series surface mount zero bias pp. 911–915, 2011. Schottky detector diodes,” 2009, https://docs.broadcom.com/ [28] RF Schottky Diodes, “Inﬁneon,” 2016, https://www.inﬁneon. doc/AV02-1377EN. com/dgdl/Inﬁneon-RF%20Schottky%20Diodes-PB-v01_01- [10] Texas Instruments, “bq25570 nano power boost charger and EN.pdf?ﬁleId=5546d462557e6e890155917b964163fb. buck converter for energy harvester powered applications,” [29] K. Niotaki, F. Giuppi, A. Georgiadis, and A. Collado, “Solar/ 2018, http://www.ti.com/lit/ds/symlink/bq25570.pdf. EM energy harvester for autonomous operation of a moni- [11] Texas Instruments, “bq25504 ultra low-power boost converter toring sensor platform,” Wireless Power Transfer, vol. 1, no. 1, with battery management for energy harvester applications,” pp. 44–50, 2014. 2019, http://www.ti.com/lit/ds/symlink/bq25504.pdf. [30] J. Bito, R. Bahr, J. G. Hester, S. A. Nauroze, A. Georgiadis, and [12] Taiyo Yuden, “Multi layer ceramic capacitor,” 2020, https://ds. M. M. Tentzeris, “A novel solar and electromagnetic energy yuden.co.jp/TYCOMPAS/or/specSheet? harvesting system with a 3-D printed package for energy pn=AMK432BJ477MM-T. eﬃcient internet-of-things wireless sensors,” IEEE Transac- [13] SCM, “Series-connected super capacitor modules,” 2020, tions on Microwave :eory and Techniques, vol. 65, no. 5, http://datasheets.avx.com/AVX-SCM.pdf. pp. 1831–1842, 2017. [14] PIC16LF1554/1559, “20-Pin ﬂash, 8-bit microcontrollers with [31] W. S. Wang, T. O’Donnell, L. Ribetto, B. O’Flynn, M. Hayes, XLP technology,” 2016, http://ww1.microchip.com/ and C. O’Mathuna, “Energy harvesting embedded wireless downloads/en/DeviceDoc/40001761E.pdf. sensor system for building environment applications,” in [15] FeRAM MR45V100A, “Lapis semiconductor,” 2017, http:// Proceedings of the 2009 1st International Conference on www.lapis-semi.com/en/data/datasheet-ﬁle_db/Memory/ Wireless Communication, Vehicular Technology, Information FEDR45V100A-01.pdf. :eory and Aerospace & Electronic Systems Technology, [16] “TS881 rail-to-rail 0.9 V nanopower comparator,” 2013, pp. 36–41, Aalborg, Denmark, 2009. https://www.st.com/resource/en/datasheet/ts881.pdf. [32] S. J. Marinkovic and E. M. Popovici, “Nano-power wireless [17] SPIRIT1, “Low data rate, low power sub-1 GHz transceiver,” wake-up receiver with serial peripheral interface,” IEEE 2016, https://www.st.com/resource/en/datasheet/spirit1.pdf. Journal on Selected Areas in Communications, vol. 29, no. 8, [18] D. Karolak, Syst`eme de Radiocommunication T´el´e-Aliment´e pp. 1641–1647, 2011. par voie Radiofr´equence a` 2.45 GHz, Universite´ de Bordeaux, [33] H. Medu, “Energy comparison of cypress F-ram and Bordeaux, France, 2015. EEPROM, cypress semiconductor,” 2017, https://www. [19] T. Taris, V. Vigneras, and L. Fadel, “A 900 MHz RF energy cypress.com/ﬁle/46746/download. harvesting module,” in Proceedings of the 10th IEEE Inter- [34] R. Usami, T. Komiyama, Y. Chonan, H. Yamaguchi, and national NEWCAS Conference, pp. 445–448, Montreal, K. Kotani, “Photovoltaic-assisted self-vth-cancellation CMOS Canada, 2012. RF rectiﬁer for wide power range operation,” in Proceedings of Wireless Power Transfer 25 the 2018 IEEE Asian Solid-State Circuits Conference (A-SSCC), pp. 115–118, Tainan, Taiwan, 2018. [35] A. P. Sample, J. Braun, A. Parks, and J. R. Smith, “Photovoltaic enhanced UHF RFID tag antennas for dual purpose energy harvesting,” in Proceedings of the 2011 IEEE International Conference on RFID, pp. 146–153, Orlando, FL, USA, 2011. [36] J. Bito, J. G. Hester, and M. M. Tentzeris, “A fully autonomous ultralow power hybrid RF/photovoltaic energy harvesting system with −25 dBm sensitivity,” in Proceedings of the 2017 IEEE Wireless Power Transfer Conference (WPTC), Taipei, Taiwan, 2017. [37] S. Oh, N. E. Roberts, and D. D. Wentzloﬀ, “A 116 nW multi- band wake-up receiver with 31-bit correlator and interference rejection,” in Proceedings of the IEEE 2013 Custom Integrated Circuits Conference, pp. 1–4, San Jose, CA, USA, 2013. [38] A. Molnar, B. Lu, S. Lanzisera, B. W. Cook, and K. S. J. Pister, “An ultra-low power 900 MHz RF transceiver for wireless sensor networks,” in Proceedings of the IEEE 2004 Custom Integrated Circuits Conference, pp. 401–404, Orlando, FL, USA, 2004. [39] T. Paing, E. Falkenstein, R. Zane, and Z. Popovic, “Custom IC for ultra-low power RF energy harvesting,” in Proceedings of the 2009 24th Annual IEEE Applied Power Electronics Con- ference and Exposition, pp. 1239–1245, Washington, DC, USA, 2009.

Journal

Wireless Power Transfer – Hindawi Publishing Corporation

Published: Aug 3, 2021

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

An Autonomous Wireless Sensor Node Based on Hybrid RF Solar Energy Harvesting

References (40)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies