sensors Article
Passive Sensors for Long Duration Internet of Things Networks Felisberto Pereira *
ID
, Ricardo Correia and Nuno Borges Carvalho
Instituto de Telecomunicações, Departamento de Electrónica Telecomunicações e Informática, Universidade de Aveiro, Aveiro 3810-193, Portugal;
[email protected] (R.C.);
[email protected] (N.B.C.) * Correspondence:
[email protected] Received: 21 August 2017; Accepted: 30 September 2017; Published: 3 October 2017
Abstract: In this work, three different concepts are used to develop a fully passive sensor that is capable of measuring different types of data. The sensor was supplied by Wireless Power Transmission (WPT). Communication between the sensor and reader is established by a backscatter, and to ensure minimum energy consumption, low power techniques are used. In a simplistic way, the process starts by the transmission of two different waves by the reader to the sensor, one of which is used in power transmission and the other of which is used to communicate. Once the sensor is powered, the monitoring process starts. From the monitoring state, results from after processing are used to modulate the incoming wave, which is the information that is sent back from the reader to the tag. This new combination of technologies enables the possibility of using sensors without any cables or batteries to operate 340 cm from the reader. The developed prototype measures acceleration and temperature. However, it is scalable. This system enables a new generation of passive Internet of Things (IoT) devices. Keywords: Wireless Sensor Networks; Radio Frequency Identification; passive sensor; low power sensor; backscatter communication; Wireless Power Transmission
1. Introduction The concept of Wireless Sensor Networks (WSNs) has for some years attracted great academic and industrial attention. The advances in semiconductor, networking, and material science technologies are driving intelligent and low-cost devices to a large-scale paradigm, working autonomously with sensing, computation, and communication capabilities. The applications can range from environmental sensing and wearable biometric monitoring to security and structural monitoring, which perhaps realize the Internet of Things (IoT) concept [1–3]. WSN scalability is dependent on reduced costs and long-term deployment without battery replacement. The urgency of this topic is evident because in the present year of 2017, the predictions point to 28.8 billion IoT devices, a number that is expected to almost double in three years, as seen in Figure 1. To address these constraints, systems have emerged that utilize communication by means of reflected Radio Frequency (RF) power [1]. The most popular system of this technology is Radio Frequency Identification (RFID), which was designed to allow for automatic identification. Its first application was a simple system that could detect the presence or absence of a tag. Although providing effective anti-theft measures continues to be one of the major applications of RFID, many other applications have appeared during the years. Currently, RFID plays a major function in the way people live it is essential in most objects, animals, or people [4,5]. Passive or battery-free RFID tags are an attractive option for WSNs. However, these tags are often application-specific fixed-function devices that can extract all of the power that they need from the carrier wave sent by the reader, as in passive tags used as barcodes alternatives. The devices
Sensors 2017, 17, 2268; doi:10.3390/s17102268
www.mdpi.com/journal/sensors
Sensors 2017, 17, 2268
2 of 12
Sensors 2017, 17, 2268
2 of 12
used in WSNs are configurable, and they have computational power, requiring more power than RFID tags [6].tags Semi-passive tags were created in an attempt to solvetothe power difficulties. Seminormal RFID [6]. Semi-passive tags were created in an attempt solve the power difficulties. passive tags have their their own batteries. However, they they are only usedused to power logiclogic operations andand not Semi-passive tags have own batteries. However, are only to power operations uplink communications [7]. The use of tags powered with batteries allowed enhancements in the not uplink communications [7]. The use of tags powered with batteries allowed enhancements in the Interrogation Zone Zone (IZ): (IZ): the the carrier carrier wave, wave, which which does does not not provide energy to to the the tag, Interrogation provide energy tag, can can be be completely completely reflected. Another method that is used to achieve longer distances between the tag and reader is the the reflected. Another method that is used to achieve longer distances between the tag and reader is separation between carrier emitter and reader. In a monostatic approach, the carrier emitter and separation between carrier emitter and reader. In a monostatic approach, the carrier emitter and reader reader are located in thedevice. same device. To deliver maximum power to the theemitter carrier is emitter is are located in the same To deliver maximum power to the tag, the tag, carrier located located near the tag, meaning that the reader is also close to the emitter. In a bistatic architecture, the near the tag, meaning that the reader is also close to the emitter. In a bistatic architecture, the carrier carrier emitter and are in different locations. The carrier near the can reader emitter and tags aretags in different locations. The carrier emitteremitter is near is the tag,the buttag, thebut reader be can be located at a greater distance. located at a greater distance. 60
Connected devices in billions
50.1 50 42.1 40 30
34.8 28.4
20 10 0 2017
2018
2019
2020
Figure 1. Internet of of Things Things (IoT): (IoT): Number Number of Figure 1. Internet of connected connected devices devices worldwide worldwide [8]. [8].
Reference research involving a battery-powered tag and topology. It presents Reference [9] [9]describes describes research involving a battery-powered tagbistatic and bistatic topology. It the results a distance of 130 mof between the tag and 4 mwith between he carrier presents theofresults of a distance 130 m between the reader, tag andwith reader, 4 m between heemitter carrier and tag.and The was transmitting a Continuous Wave (CW) with a power ofof1313dBm. emitter tag.carrier The carrier was transmitting a Continuous Wave (CW) with a power dBm. The same battery-powered tag and bistatic architecture is proposed in reference [10]. However, coherent detection and channel coding were used to improve the results to 150 m between the tag and reader and 10 10 m mbetween betweenthe thecarrier carrieremitter emitter and tag. Although results obtained by battery-powered and tag. Although thethe results obtained by battery-powered tags tags are excellent in terms of distance, there are limitations: the battery cost is considerable and are excellent in terms of distance, there are limitations: the battery cost is considerable and battery battery replacement can be a problem. The alternatives to battery-powered tags are tags with energy replacement can be a problem. The alternatives to battery-powered tags are tags with harvesting systems. systems. Reference Reference[11] [11]proposed proposeda asystem systemwith withphotovoltaic photovoltaic cells power tags, and cells to to power tags, and in in reference [12], power sourceisismotion. motion.However, However,ininthe thelatter lattercase, case, there there are are ambient power reference [12], thethe power source constraints, both on where one one can can place place sensors sensors and and when when they they can can be be used used [13]. [13]. To overcome overcome the the battery battery and and energy energy harvesting harvesting restrictions, the alternative is to use energy transmission to supply the tag with The first idea on this subject, which concerned the with power. power. The transmission of energy from space down to Earth, was suggested by Peter Glaser (1969) in his project Solar Power Satellite. The The idea idea was was to place solar cells in space, and then send a microwave beam down to Earth. The wave would bebe received by abyrectenna that that would convert the radiation into Direct to Earth. The wave would received a rectenna would convert the radiation into Current (DC) power Although the first idea for Wireless Transmission (WPT) focused on Direct Current (DC) [14]. power [14]. Although the first idea forPower Wireless Power Transmission (WPT) high power, the major developments in this field have beenfield occurring lowoccurring power transmission. focused on high power, the major developments in this have in been in low power In [15], a solution using inductive WPT and Ultra High Frequency (UHF) RFID was presented. transmission. The work proves that combining the inductive WPT withHigh the UHF RFID increments thewas tag presented. sensitivity In [15], a solution using inductive WPT and Ultra Frequency (UHF) RFID The work proves that combining the inductive WPT with the UHF RFID increments the tag sensitivity
Sensors 2017, 17, 2268
3 of 12
in 21 dB, which increases the IZ. However, the use of inductive WPT requires proximity between the tag and the power source. Regarding electromagnetic WPT, a comparison between different rectifier topologies and different stage levels was presented in [16]. The obtained results show a high dependence between the received power and the most efficient topology. A structure with a two-tone signal at 1.8 GHz and 2.4 GHz was shown in [17]. The results present a voltage output 20% higher in average when comparing with a single-tone input. The work in reference [18] begins with a reader that is configured to transmit power in CW. After rectification, the power charges the storage capacitor to 5.5 V. The storage capacitor, once charged, powers the tag that performs sensing and communication, reflecting the carrier wave. This entire process occurs at a distance of 1 m between the reader and tag. The same principle is used in reference [6]. Different from the previous solutions, this work presents a solution using dual band wireless power and data transfer. One frequency is exclusively used to transmit energy to the sensor, and the other is fully dedicated to communication through backscatter. The technology was previously introduced without the use of any sensing nodes in reference [19]. A similar approach, using different frequencies for transmitted energy and communication was presented in [20]. In that work, for each frequency that is used a differentiated circuit for the RF-DC conversion. In the work presented in this manuscript, the entire process is done in a single circuit. The use of one single circuit reduces the number of components, which means lower losses, and allows for a more compact solution, which make a very reliable solution for IoT applications. It is important to refer that this work is a proof of concept regarding the combination of different techniques, upper layers like communication protocols are not taken into account. This document is organized as follows. In the subsequent Sections 2 and 3, the WPT and backscatter techniques are described and the methods used to combine both techniques are demonstrated. Section 4 explores the sensing and processing elements, first by outlining the components used and second by presenting the software iterations. All of the measurements and results are provided in Section 5 where the step-by-step process is explained and some considerations are also presented. Finally, Section 6 discusses the main conclusions of this work. 2. WPT and Backscatter Techniques In recent years, the increasing use of IoT sensors has allowed smart objects to provide major industries with the vital data that they need to track inventory, manage machines, increase efficiency, save costs, and even save lives. In this context, in which billions of connected objects are expected to be placed all over the world, frequent battery maintenance of wireless nodes is undesirable or even impossible. In these scenarios, passive-backscatter radios will play a crucial role due to their low cost, low complexity, and battery-free operation. Wireless power transmission technology also plays an important role in providing continuous power for these backscatter radios. To implement the combination of these two technologies, the approach used in reference [21] was followed. In this approach, two different transmitter frequencies are used, one to supply the sensor and the other to communicate via backscatter. Figure 2 shows a block diagram of the proposed combined technologies. Two main blocks are presented. The first, the RF harvesting circuit, is responsible for converting RF signals into DC energy to supply the microcontroller and modulate the information acquired by the sensors. The modulation is provided by switching the transistor ON and OFF, which is part of the backscatter modulator block. This process is known as Modulated Scattering Technique (MST) [22–25], resulting in this case in an Amplitude Shift Keying (ASK) modulation.
Sensors2017, 2017,17, 17,2268 2268 Sensors Sensors 2017, 17, 2268
12 44ofof12 4 of 12
Figure 2. Block diagram of the implemented system. Figure 2. Block diagram of the implemented system.
3. RF Front End Design and Optimization 3. 3. RF RF Front Front End End Design Design and and Optimization Optimization The RF circuit is shown in Figure 3. It is divided into three main sections: the backscatter The RF circuit circuit is is shown shown in in Figure Figure 3. 3. It is divided divided into into three three main main sections: sections: the the backscatter backscatter The RF It is modulator, which is composed of a switch transistor that modulates the impedance of the antenna; modulator, which is composed of a switch transistor that modulates the impedance of the antenna; modulator, is composed of adesigned switch transistor that modulatesload the modulation impedance of the the matchingwhich network, which was to provide backscatter atthe oneantenna; frequency the matching network, which was designed to provide backscatter load modulation at one frequency matching network, which was at designed to provide backscatter load modulation at multiplier, one frequency and and continuous flow of WPT other frequencies; and, the five-stage Dickson which and continuous flow of at WPT atfrequencies; other frequencies; and, the five-stage Dickson multiplier, which continuous flow of WPT other and, the five-stage Dickson multiplier, which allows for allows for RF-DC conversion to provide sufficient DC power to supply the microcontroller. allows RF-DC conversion provideDC sufficient powerthe to supply the microcontroller. RF-DC for conversion to provide to sufficient power DC to supply microcontroller.
Figure 3. Photograph of implemented system with backscatter modulator combined with Wireless Figure 3. Photograph of implemented system with backscatter modulator combined with Wireless Figure Photograph (WPT). of implemented withL1 backscatter modulator combined with Wireless Power 3. Transmission Element system values are = 21.2 mm, W1 = 1.87 mm, L2 = 15.1 mm, Power Transmission (WPT). Element values are L1 = 21.2 mm, W1 = 1.87 mm, L2 = 15.1 mm, W2 = 1.0 Power Transmission (WPT). Element L1 =mm, 21.2 W4 mm,=W1 1.87 mm, L2 = mm, 15.1 mm, 1.0 W2 = 1.0 mm, L3 = 21.9 mm, W3 = 0.8 values mm, L4are = 11.3 1.87=mm, L5 = 17.1 W5 =W2 1.2 =mm, mm, L3 = 21.9 mm, W3 = 0.8 mm, L4 = 11.3 mm, W4 = 1.87 mm, L5 = 17.1 mm, W5 = 1.2 mm, L6 = 6.7 mm, L3 =mm, 21.9W6 mm, W3mm, = 0.8L7 mm, L4 =mm, 11.3W7 mm, W4mm, = 1.87 17.1C1 mm, W5 1.2 mm, L6 6.7 L6 = 6.7 = 1.1 = 18.6 = 0.7 R1mm, = 50L5 Ω,=and = 47 pF.= Substrate for= the mm, W6 = 1.1 mm, L7 = 18.6 mm, W7 = 0.7 mm, R1 = 50 Ω, and C1 = 47 pF. Substrate for the mm, W6 = 1.1 mm, L7 = MT77, 18.6 mm, W7 = = 0.70.762 mm,mm, R1 ε=r = 503.0, Ω, tan andδ = C10.0017. = 47 pF. Substrate for the transmission lines is Astra thickness transmission lines is Astra MT77, thickness = 0.762 mm, ε = 3.0, tan δ = 0.0017. transmission lines is Astra MT77, thickness = 0.762 mm, ε = 3.0, tan δ = 0.0017.
Themain main differences differences from consist of of thethe design andand the The from the the works workspresented presentedininreferences references[19,26] [19,26] consist design The main differences from the works presented in references [19,26] consist of the design and optimization of the matching network to provide enough DC power to supply the digital processing the optimization of the matching network to provide enough DC power to supply the digital the optimization of the matching network to provide enough DC power to supply the digital unit. Once unit. all ofOnce the lower within the sensor were designed, thedesigned, sleep current processing all offrequency the lowercircuits frequency circuits withinnode the sensor node were the processing unit. Once all of the lower frequency circuits within the sensor node were designed, the consumption that resulted from the sum from of all of quiescent was taken into consideration sleep current consumption that resulted thethe sum of all ofcurrents the quiescent currents was taken intoto sleep current consumption that resulted from the sum of all of the quiescent currents was taken into model the load the energy whichharvester, was designed to mimic the real to load that the the real harvester consideration to for model the loadharvester, for the energy which was designed mimic load consideration to model the load for the energy harvester, which was designed to mimic the real load seesthe while the sensor sleeping. Out of of the circuitry from the output from of the that harvester sees node whileisthe sensor node is all sleeping. Out of alldownstream of the circuitry downstream that the harvester sees while the sensor node is sleeping. Out of all of the circuitry downstream from energy harvester, only the large storage capacitor and Schottky diode precedes it were included the output of the energy harvester, only the large storage capacitor andthat Schottky diode that precedes the output of the energy harvester, only the large storage capacitor and Schottky diode that precedes into the simulation model of the load. All of the other circuits were replaced by an ideal current source it were included into the simulation model of the load. All of the other circuits were replaced by an it were included into the simulation model of the load. All of the other circuits were replaced by an that was placed in parallel with the storage capacitor a value capacitor set equal to the aestimated ideal current source that was placed in parallel withwith the storage with value setmaximum equal to ideal current source that was placed in parallel with the storage capacitor with a value set equal to sleep current consumption (thatcurrent is, 5 µA). A harmonic(that balance was performed Advanced the estimated maximum sleep consumption is, 5simulation μA). A harmonic balanceinsimulation the estimated maximum sleep current consumption (that is, 5 μA). A harmonic balance simulation was performed in Advanced Design System (ADS) to tune the matching network for the maximum was performed in Advanced Design System (ADS) to tune the matching network for the maximum
Sensors 2017, 2017, 17, 17, 2268 2268 Sensors
55 of of 12 12
conversion efficiency and DC output with the ideal current. In references [19,26], simulations were performed for (ADS) a fixed load the value, without any for consideration of conversion the digitalefficiency processing Design System to tune matching network the maximum andunit DC consumption. output with the ideal current. In references [19,26], simulations were performed for a fixed load value, To any provide both backscatter communication and consumption. wireless power transfer, a Large Signal Swithout consideration of the digital processing unit Parameter (LSSP)both simulation was performed with theand objectives of matching the circuit in bothSignal states To provide backscatter communication wireless power transfer, a Large of the transistor at 1.7 GHz, matching the system at 2.4 in oneofstate and mismatching the states other S-Parameter (LSSP) simulation was performed with the GHz objectives matching the circuit inatboth state. change antenna’s impedance at at 2.42.4 GHz provides the backscatter communication. of the This transistor at of 1.7the GHz, matching the system GHz in one state and mismatching at the other theofRF three patch antennas were used. Two of them were connected state.Regarding This change thetransmission, antenna’s impedance at 2.4 GHz provides the backscatter communication. to theRegarding wave generator at 1.7 three GHz and 2.4antennas GHz, respectively, to the the RFmatched transmission, patch were used.and Twothe of third themconnected were connected sensor matched at bothmatched frequencies (dual-band). patchrespectively, antennas were due connected to their high to the wave generator at 1.7 GHz and The 2.4 GHz, andchosen the third to directivity. the sensor matched at both frequencies (dual-band). The patch antennas were chosen due to their high directivity. 4. Sensing and Processing Elements 4. Sensing and Processing Elements The sensing part of the system is composed of a temperature sensor and an accelerometer, both The sensing part of the system is composed of aconsumption temperature sensor and anmeasurements. accelerometer, both of of which were chosen carefully to have low power and reliable which wereInstruments’ chosen carefully to have consumption and was reliable measurements. Texas LM94021 waslow thepower temperature sensor that selected. In addition to its low Texas Instruments’ LM94021 was the temperature sensor that was selected. In addition to its and low power consumption (12 μA) and precise calibration, the sensor has a simple calibration process power consumption (12 µA) has a simple calibration process and temperature range from −50 and °C toprecise 150 °Ccalibration, [27]. Since the thesensor measurements performed during the tests temperature range from −50 ◦ C tosensor 150 ◦ Cwas [27].calibrated Since thetomeasurements performed the tests were at ambient temperatures, the operate between 0 °C andduring 40 °C, allowing ◦ ◦ were at ambient temperatures, for a better output resolution. the sensor was calibrated to operate between 0 C and 40 C, allowing for a The better outputsensor resolution. second was added to prove that the system can work with multiple sensors. The second sensor wasdue added to prove that the system can work interaction with multiple sensors. Accelerometers were chosen to their low power consumption, human factor, and a Accelerometers were chosen due to their low power consumption, human interaction factor, and wide range of possible applications. The accelerometer chosen was Analog Devices’ ADXL362, which a wide range possible Themode accelerometer chosen was Analog ADXL362, which consumes 1.8ofμA at 100 applications. Hz in operation and 10 nA in standby mode Devices’ [28]. consumes 1.8 µA at 100 Hz in operation mode and 10 nA in standby mode [28]. To acquire and process all of the data that come from the sensors, a processing unit is necessary. acquire and process all of the data that come from the sensors, a processing necessary. The To chosen microcontroller was MSP430F2132 from Texas Instruments (Dallas, unit TX, is USA). This The chosen microcontroller was MSP430F2132 fromand Texas Instruments (Dallas, TX, contains USA). This microcontroller only consumes 2 μA in active mode 0.3 μA in sleep mode. It also I/O microcontroller only consumes 2 µA in active mode and 0.3 µA in sleep mode. It also contains ports, an Analog to Digital Converter (ADC), timers, and communications protocols. The only feature I/O ports, to Digitalwith Converter timers, and communications protocols. The only missing foran theAnalog microcontroller regard (ADC), to the final application was a Digital to Analog Converter feature missing for the microcontroller with regard to the final application was a Digital to Analog (DAC) unit. This unit was added to the system as an external component [29]. Converter (DAC) unit. This unit was added to the system as an external component [29]. The sensing and processing elements were supplied by the RF-DC converter. The voltage The from sensing processing elements wereon supplied by thefrom RF-DC delivery delivery theand RF-DC converter depends the distance theconverter. reader as The wellvoltage as many other from the RF-DC converter depends on the distance from the reader as well as many other factors, factors, such as wave reflections. Different levels of voltage create different behaviors insuch the as wave reflections. Different levels ofare voltage create for different behaviors in theactivities. components, components, and stable conditions important performing sensing To and solvestable this conditionsaare important for performing sensing activities. To solve this problem, voltage regulator problem, voltage regulator circuit was placed between the RF-DC circuit aand sensing and circuit was placed between the RF-DC circuit and sensing and processing units. The voltage regulator processing units. The voltage regulator output was 1.8 V. In Figure 4, the PCB with the sensing and output waselements 1.8 V. In Figure 4, the PCB with the sensing and processing elements is shown. processing is shown.
Figure 4. Developed Printed Circuit Board (PCB). Figure 4. Developed Printed Circuit Board (PCB).
As important as choosing low power components is in the software development, they should Asefficient important as choosing low power components in themode software development, theyneeded. should also be to ensure that the components are only inisactive when they are actually also be efficient to ensure that the components are only in active mode when they are actually needed. These techniques allow for a reduction of the system power consumption. The first software
Sensors Sensors 2017, 2017, 17, 17, 2268 2268
66 of of 12 12
operation is the configuration of all of the elements, clocks, I/O ports, timers, ADC, DAC, and sleep These techniques allow for a reduction of the system power consumption. The first software operation modes. After configurations, the software processes the sensing block, reading the temperature and is the configuration of all of the elements, clocks, I/O ports, timers, ADC, DAC, and sleep modes. After acceleration. The measured data need to be processed since misleading measurement errors can configurations, the software processes the sensing block, reading the temperature and acceleration. occur. A message is then produced to be delivered to the DAC. At this point, the DAC is responsible The measured data need to be processed since misleading measurement errors can occur. A message is for controlling the system impedance by varying the voltage. When the message bit is a zero, the then produced to be delivered to the DAC. At this point, the DAC is responsible for controlling the DAC output voltage is 0.0 V, which means that the system is adapted and, consequently, that there system impedance by varying the voltage. When the message bit is a zero, the DAC output voltage is is no reflection. By contrast, when the message bit is one, the DAC output voltage is 0.6 V, which 0.0 V, which means that the system is adapted and, consequently, that there is no reflection. By contrast, means that the impedance of the system is mismatched, which occurs with the reflection of the carrier when the message bit is one, the DAC output voltage is 0.6 V, which means that the impedance of the wave. After sending each bit, the microcontroller enters sleep mode for a short period of time. When system is mismatched, which occurs with the reflection of the carrier wave. After sending each bit, the it sends the complete message, the microcontroller enters a longer period of sleep. Figure 5 illustrates microcontroller enters sleep mode for a short period of time. When it sends the complete message, the the process described. microcontroller enters a longer period of sleep. Figure 5 illustrates the process described.
Figure 5. Code diagram. Figure 5. Code diagram.
All All of of the the message message and and sleep sleep periods periods are are defined defined by by aa simple simple variable variable on on the the code. code. The The same same is is done modify done with with the the output output voltage voltage delivered delivered by by the the DAC. DAC. This This gives gives the the system system the the ability ability to to easily easily modify the ofof a total of of 1111 bits—3 bitsbits to the parameters parametersfor fordifferent differentapplications. applications.The Themessage messageframe frameisiscomposed composed a total bits—3 identify thethe tag tag ID, ID, 2 bits to identify the sensor, 4 bits4for and 2 and bits to detect to identify 2 bits to identify the sensor, bitsthe formeasurement, the measurement, 2 bits to errors detect in the message. The following Table 1Table exemplifies the message frameframe with the coding resulting from errors in the message. The following 1 exemplifies the message with the coding resulting the measured temperature: from the measured temperature: Table 1. Table 1. Corresponding Corresponding code code from from the the measured measured temperature. temperature.
Temperature (Celsius) Temperature (Celsius) 0–2.5 ◦ C °C 0–2.5 ◦ C °C 2.5–5 2.5–5 ◦C 5–7.5 5–7.5 °C ( . . .(…) ) 37.5–40 ◦ C
37.5–40 °C
CodeCode (bits) (bits) 101-00-0001-11 101-00-0001-11 101-00-0010-11 101-00-0010-11 101-00-0011-01 101-00-0011-01 (…)( . . . ) 101-00-1111-01 101-00-1111-01
The power consumption of the tag is is defined defined by by the the bit bit period period and and sleep sleep period. period. Long sleep periods periods result in in low low power power consumption. consumption. The bit period is related to the frequency of the reflected reflected wave, which defines the deviation of the modulated message from the carrier. Having a good deviation wave, which defines the deviation of the modulated message from the carrier. Having a good from the carrier wave is crucial the tag from thetag carrier Tests deviation from the carrier wavetoisisolate crucialthe to wave isolatemodulated the wave by modulated by the fromwave. the carrier were to optimize theto periods. Periods of 366 µs Periods for the bit the sleep wave.conducted Tests were conducted optimize the periods. ofperiod 366 μsand for6.312 the ms bit for period and period were established. The tag average consumption was 59 µA. 6.312 ms for the sleep period were established. The tag average consumption was 59 μA.
Sensors 2017, 2017, 17, 17, 2268 2268 Sensors
of 12 12 77 of
5. Measurements and Results 5. Measurements and Results The first measurements were to ensure the correct operation of the digital logic. Figure 6 shows the DAC output with the code “101-00-1000-11” that corresponds a temperature °C The first measurements were to ensure the correct operation oftothe digital logic.between Figure 617.5 shows andDAC 20 °C.output with the code “101-00-1000-11” that corresponds to a temperature between 17.5 ◦ C the and 20 ◦ C.
Figure 6. Message with tag ID, sensor ID, measured temperature, and correction bits. Figure 6. Message with tag ID, sensor ID, measured temperature, and correction bits.
the goal goal of of this work is to develop a passive tag. However, to perform As discussed previously, the and evaluate measurements, a reader reader is is needed. needed. readerisiscomposed composedofofa aRohde Rohde Schwarz SMW200A Vector Signal Generator to generate The reader && Schwarz SMW200A Vector Signal Generator to generate two two signals at different frequencies, oneband dualpatch bandantenna patch antenna and a& Rohde & FSP Schwarz FSP signals at different frequencies, one dual and a Rohde Schwarz Spectrum Spectrum(Munich, AnalyzerGermany) (Munich, Germany) analyze signal modulated andby reflected thetag tag.has The Analyzer to analyzeto the signal the modulated and reflected the tag.byThe a tag has a dual bandaantenna, a board with circuit, an RF-DC another board with theand processing dual band antenna, board with an RF-DC andcircuit, anotherand board with the processing sensing and sensing elements. The RF-DC circuit andsensing processing sensing aresame not on the same board elements. The RF-DC circuit and processing elements areelements not on the board to allow for to allowtests for specific for each approach. Figure 6 shows the measurements setup. specific for eachtests approach. Figure 6 shows the measurements setup. The first measurement was taken to determine the optimal frequencies for system system operation. operation. The S11 S11parameter parameterofof the antennas and DC output voltage of the rectifier were measured. the antennas and DC output voltage of the rectifier were measured. In FigureIn 7, Figure 7, it is that observed theregood is a very good relationship between the most efficient onand the it is observed there that is a very relationship between the most efficient point on thepoint RF-DC RF-DC and antenna adaptation. optimal frequencies are on the same points. Based on antenna adaptation. The optimalThe frequencies for both arefor on both the same points. Based on the results, results, the frequency chosen for GHz WPT and wasfor1.720 GHz and for performing backscatter the frequency chosen for WPT was 1.720 performing backscatter communication was communication was 2.405 GHz. 2.405 GHz.
Figure 7. 7. Measurements setup. Figure Measurements setup.
Sensors 2017, 17, 2268
8 of 12
Sensors 2017, 17, 2268 12 The following measurements are used to evaluate the input power and output voltage8 of after the Sensors 2017,under 17, 2268 different test conditions. There are three parameters that changed during 8 of 12 tests: RF-DC circuit the The following measurements are used to evaluate the input power and output voltage after the load, RF-DC tag adaptation, anddifferent power delivery. No load means that there is no loadchanged attachedduring to thethe RF-DC under test are conditions. There are Thecircuit following measurements used to evaluate thethree inputparameters power andthat output voltage after the circuit, and a load occurs when the sensing and processing board are attached to the RF-DC circuit. tests: load, tag under adaptation, andtest power delivery.There No load that there isthat no changed load attached to the the RF-DC circuit different conditions. are means three parameters during Backscatter absorbing is the state at which the DAC output voltage is 0.0 V, meaning that the circuit is RF-DC circuit, and a load occurs whendelivery. the sensing attached to the RF-DC tests: load, tag adaptation, and power No and loadprocessing means thatboard thereare is no load attached to the circuit. Backscatter absorbing is the state at sensing which the DAC output reflection, voltage is 0.0 V, DAC meaning the at its RF-DC most adapted point (maximum absorption). In backscatter the is 0.6 V, circuit, and a load occurs when the and processing board are attached to output thethat RF-DC circuit isBackscatter at to itsamost adaptedofpoint (maximum absorption). In backscatter reflection, the DACthat output corresponding mismatch the (maximum reflection). The power delivery circuit. absorbing is thesystem state atimpedance which the DAC output voltage is 0.0 V, meaning the to is 0.6 V, corresponding a point mismatch of the system impedance (maximum reflection). The output power circuit is at its from most adapted absorption). In backscatter reflection, the DAC the system varies −20todBm to 0(maximum dBm. delivery to the system varies from −20 dBm to 0 dBm. is 0.6 V, corresponding to a mismatch of the system impedance (maximum reflection). The power The results are shown in Figure 8. It can be concluded that the frequency that is exclusively used The to results are shown in from Figure 8. dBm It canto be0concluded the frequency that is exclusively used delivery the system −20 dBm.or thethat for WPT is not affected byvaries the backscatter absorption reflection state. The figure also shows that for WPT is not affected by in the backscatter or the state.that Theisfigure also shows The results are shown Figure 8. It canabsorption be concluded thatreflection the frequency exclusively used the wave with 2.405 GHz generates more DC voltage whenwhen the system is adapted than when ititis not. that the wave with 2.405 GHz generates more DC voltage the system is adapted than when for WPT is not affected by the backscatter absorption or the reflection state. The figure also shows Another aspect is that there is a decrease of voltage when a load is present. As is known, maintaining is not. aspect that generates there is a more decrease voltage when load isispresent. is when known, that theAnother wave with 2.405is GHz DC of voltage when theasystem adaptedAs than it the delivered power and increasing the current makes the voltage drop. maintaining the delivered power and increasing the current makes the voltage drop. is not. Another aspect is that there is a decrease of voltage when a load is present. As is known,
maintaining the delivered power and increasing the current makes the voltage drop.
Figure 8. Optimal operational frequency. Figure 8. Optimal operational frequency. Figure 8. Optimal operational frequency.
To better understand the reliability of the WPT technique, the reader and tag were placed at a To better reliability ofofthe technique, the reader tag were distance ofunderstand 300understand cm from the each other. During theWPT test, the power transmitted by and the was placed varied To better the reliability the WPT technique, the reader and tagreader were placed at a at a distance 300 cm from each other. the thepower power transmitted reader was varied fromof0 dBm to 30 dBm. By measuring the power received by thetransmitted tag at both frequencies, itswas respective distance of 300 cm from each other.During During the test, test, the by by thethe reader varied output voltage, the tag state (ONthe or power OFF), is possible obtain of the information about the from from 0 dBm to 30todBm. ByBy measuring received byto the tag at both frequencies, its respective 0 dBm 30and dBm. measuring the poweritreceived by the tag atall both frequencies, its respective WPT process. Figure 9 presents the results. output voltage, and tag state (ON OFF), it it is possible allall of of thethe information about the the output voltage, and thethe tag state (ON ororOFF), possibletotoobtain obtain information about process. Figure 9 presents results. WPT WPT process. Figure 9 presents thethe results.
Figure 9. Input power vs. output voltage. Figure9.9.Input Inputpower power vs. Figure vs.output outputvoltage. voltage.
Sensors 2017, 17, 2268 Sensors 2017, 17, 2268
9 of 12 9 of 12
Figure showsthat thatanan increment in transmitted the transmitted power in growth linear growth of the Figure 99shows increment in the power resultsresults in linear of the received received power for both frequencies until the tag is OFF. The tag turns ON when the transmitted power for both frequencies until the tag is OFF. The tag turns ON when the transmitted power is near power near dBm.the Attag this point, the tag starts performing the backscatter 26 dBm.is At this26point, starts performing the backscatter process, changing process, between changing reflecting between reflecting and absorbing states. The ON state of the tag is reflected on the receiver power, and absorbing states. The ON state of the tag is reflected on the receiver power, at 2.405 GHz, whereata 2.405 GHz, where aondecrease is visible thethat growing decrease is visible the growing line on after point. line after that point. In addition to the previous test for understanding In addition to the previous test for understanding the the characteristics characteristics of of WPT, WPT, another another test test was was conducted to measure the maximum distance at which the reader can supply the tag. The maximum conducted to measure the maximum distance at which the reader can supply the tag. The maximum transmitted and thethe maximum distance of 430 cm between reader and transmitted power, power,30 30dBm, dBm,was wasdefined, defined, and maximum distance of 430 cm between reader tag was measured. In this point it is important to refer that European Telecommunications Standards and tag was measured. In this point it is important to refer that European Telecommunications Institute (ETSI) legislates amountthe ofamount power that can bethat transmitted. For an UHF the RFID legal Standards Institute (ETSI)the legislates of power can be transmitted. ForRFID an UHF limitation is 33 dBmisEffective (ERP). The(ERP). value The of 30value dBm of was laboratory the legal limitation 33 dBmradiated Effectivepower radiated power 30used dBmdue wastoused due to limitations. laboratory limitations. Even Even though though the the maximum maximum distance distance at at which which the the reader reader can can supply supply the the tag tag is is 430 430 cm, cm, this this does does not mean that the complete system can work at that distance. For this system to work, the tag needs not mean that the complete system can work at that distance. For this system to work, the tag needs to power so so that it can be received by the reader. To to modulate modulate the thesignal signaland andreflect reflectit itwith withenough enough power that it can be received by the reader. understand all all of of thethe behaviors, new measurements were To understand behaviors, new measurements weremade. made.The Thereceived receivedpower poweratatthe the tag tag at at both frequencies, power reflected by the tag, and power received by the reader were measured. These both frequencies, power reflected by the tag, and power received by the reader were measured. These measurements measurements were were made made by by varying varying the the transmitted transmitted power power between between 10 10 dBm dBm and and 30 30 dBm, dBm, and and the the distance between reader and tag was recorded. All of the results are presented in Figure 10. distance between reader and tag was recorded. All of the results are presented in Figure 10.
Figure 10. 10. WPT WPT at at distance distance of of 300 300 cm. cm. Figure
The power powerreceived receivedbyby not suffer large variations, natural once the thethe tag tag doesdoes not suffer large variations, which iswhich naturalisonce the minimum minimum received power to turn the tag ON is constant. Since the power received is constant and received power to turn the tag ON is constant. Since the power received is constant and reflection reflection always the same fromit the tag, it that is logical that the power reflected has a value always occurs theoccurs same way fromway the tag, is logical the reflected haspower a value that does that does not have large variations. However, since tag always the same amount of power not have large variations. However, since the tagthe always reflectsreflects the same amount of power and and the distance between reader and tag increases, the power received by the reader decreases. At a the distance between reader and tag increases, the power received by the reader decreases. distance of 340 cm cm between between reader reader and and tag, tag, the the reflected reflected power power that that arrives arrives at at the the tag tag isisapproximately approximately −85 −85dBm. dBm.Values Valuesbelow belowthis thisthreshold thresholddo donot notallow allowthe thereader readerto todemodulate demodulate the the signal, signal, meaning meaning that that communication between between the the tag tagand andreader readerisisnot notpossible possibleatata adistance distance further than 340 cm, as further than 340 cm, as cancan be be seen in Figure seen in Figure 11.11.
Sensors 2017, 17, 2268 Sensors 2017, 17, 2268
10 of 12 10 of 12
Figure Figure11. 11. Transmitted Transmittedpower powervs. vs.maximum maximumdistance. distance.
Considering Consideringthe theresults resultsand andto tobetter betterunderstand understandthem, them,the thefree freespace spaceloss losswas wascalculated calculatedbased based on measurements. At the maximum distance, 340 cm, the sensor is reflecting − 46.67 dBm and on measurements. At the maximum distance, 340 cm, the sensor is reflecting −46.67 dBm and the the reader 85.17 dBm, were measured using a coupler (Marki, Morgan Hill, Hill, CA, readerisisreceiving receiving−−85.17 dBm,these thesevalues values were measured using a coupler (Marki, Morgan USA, CBR16-0012) to differentiate incident from from reflected waves. The total of losses is: is: CA, USA, CBR16-0012) to differentiate incident reflected waves. The amount total amount of losses
= Power + Antennas − Losses Power Power received = Powerreflected + AntennasGain − Losses h=i Losses == −− Power Powerreceived ++Power Antennas dB h=i Losses Power Antennas =46.5 46.5 dB Gain = reflected ++
(1) (1)
Despitethis, this,we wecan cannot notuse usethe theFriis Friisformula formula in in this this scenario, scenario, due due to to multipath multipath effects effectsand andother other Despite imperilments we use it to show that the measured results are not out of scope. Considering the same imperilments we use it to show that the measured results are not out of scope. Considering the same distance obtained in the experiments, the theoretical losses were calculated: distance obtained in the experiments, the theoretical losses were calculated: λ (2) Losses = λ 2 = 42.68 dB 4πR = 42.68 dB (2) Losses = 4πR This validates the measured results. This validates the measured results. 6. Conclusions 6. Conclusions In this paper, a passive tag with sensing and processing elements was presented. The developed In paper, a passive tag with sensing andand processing elements was presented.toThe developed tag hasthis two sensing elements: temperature acceleration. It is important note that the tag has two sensing elements: temperature and acceleration. It is important to note that The the microcontroller has the capability to incorporate a greater number and different types of sensors. microcontroller hastag theincapability to incorporate a greater numberthat anddodifferent types sensors. aim is to use the multiple scenarios, especially situations not allow for of a physical The aim is to use the tag in multiple scenarios, especially situations that do not allow for a physical connection, maintenance and access. Agriculture, space, and smart house applications are good connection, and access. Agriculture, and smart house are good examples formaintenance which this technology could be used. Inspace, agriculture, sensors needapplications to be underground to examples for which this technology could be used. In agriculture, sensors need to be underground to measure soil conditions. This scenario is not favorable for battery replacement. In space applications, measure soil conditions. scenario is not by favorable battery In space applications, the monitoring system isThis usually connected a large for amount of replacement. cables. Removing the cables would the monitoring system is usually connected by afuel large amount of cables. Removing cables would decrease the ship weight and consequently consumption. In smart house the applications, the decrease the ship weight and consequently fuel consumption. In smart house applications, the sensors sensors could be embedded at different locations, making different types of measurements. couldOne be embedded at differentoflocations, making different typesand of measurements. of the contributions this paper is the measurement testing of the technology, which One ofWPT the contributions of this paper is the measurement testingthe of the which combines and backscatter techniques. This technologyand enables use technology, of one frequency combines WPT and backscatter techniques. This technology enables the use of one frequency exclusively for power transmission and another for communication, which allows for greater exclusively for power andDevelopment another for communication, which allows forelements greater distances distances between thetransmission reader and tag. of the sensing and processing to ensure between the reader and tag. Development of the sensing and processing elements to ensure maximum maximum power efficiency is an important addition. The obtained results show a very good distance power efficiency is an important addition. The obtainedits results show a veryafter good distance at whichbyit its is at which it is possible for the sensor to communicate measurements they are handled possible for unit. the sensor to communicate its measurements after they are handled by its processing unit. processing Acknowledgments: This work is funded project (ref. POCI-010145-FEDER-016426), SAICTPAC/0011/2015 MobiWise.
Sensors 2017, 17, 2268
11 of 12
Acknowledgments: This work is funded project (ref. POCI-010145-FEDER-016426), SAICTPAC/0011/2015 MobiWise. Author Contributions: All the authors have contributed to this paper. Felisberto Pereira has designed the digital circuit, performed the experiments results and wrote the paper. Ricardo Correia has designed the RF front-end and supervised the entire research. Nuno Borges Carvalho has supervised the entire research, the used approach, the results analysis and discussion, and provided the guidance for writing the paper. Coflicts of Interest: The authors declare no conflict of interest.
References 1.
2.
3. 4. 5. 6. 7. 8. 9. 10.
11.
12. 13. 14.
15. 16. 17.
18.
Kampianakis, E.; Kimionis, J.; Tountas, K.; Konstantopoulos, C.; Koutroulis, E.; Bletsas, A. Wireless Environmental Sensor Networking with Analog Scatter Radio and Timer Principles. IEEE Sens. J. 2014, 14, 3365–3376. [CrossRef] Sarwesh, P.; Shet, N.S.V.; Chandrasekaran, K. Energy efficient network architecture for IoT applications. In Proceedings of the 2015 International Conference on Green Computing and Internet of Things (ICGCIoT), Noida, Delhi, India, 8–10 October 2015; pp. 784–789. Silicon Labs. The Evolution of Wireless Sensor Networks; White Papers; Silicon Laboratories, Inc.: Austin, TX, USA, 2013. Landt, J. The history of RFID. IEEE Potentials 2005, 24, 8–11. [CrossRef] Rida, A.; Yang, L.; Tentzeris, M. RFID-Enabled Sensor Design and Applications; Artech House: Norwood, MA, USA, 2010. Sample, A.P.; Yeager, D.J.; Powledge, P.S.; Mamishev, A.V.; Smith, J.R. Design of an RFID-Based Battery-Free Programmable Sensing Platform. IEEE Trans. Instrum. Meas. 2008, 57, 2608–2615. [CrossRef] Roy, S.; Jandhyala, V.; Smith, J.R.; Wetherall, D.J.; Otis, B.P.; Chakraborty, R.; Buettner, M.; Yeager, D.J.; Ko, Y.C.; Sample, A.P. RFID: From Supply Chains to Sensor Nets. Proc. IEEE 2010, 98, 1583–1592. [CrossRef] Statista. The Statistics Portal for Market Data, Market Research and Market Studies. Available online: https://www.statista.com (accessed on 5 July 2017). Kimionis, J.; Bletsas, A.; Sahalos, J.N. Increased Range Bistatic Scatter Radio. IEEE Trans. Commun. 2014, 62, 1091–1104. [CrossRef] Fasarakis-Hilliard, N.; Alevizos, P.N.; Bletsas, A. Coherent detection and channel coding for bistatic scatter radio sensor networking. In Proceedings of the 2015 IEEE International Conference on Communications (ICC), London, UK, 8–12 June 2015; pp. 4895–4900. Sample, A.P.; Braun, J.; Parks, A.; Smith, J.R. Photovoltaic enhanced UHF RFID tag antennas for dual purpose energy harvesting. In Proceedings of the 2011 IEEE International Conference on RFID, Orlando, FL, USA, 12–14 April 2011; pp. 146–153. Roundy, S.; Wright, P.K.; Rabaey, J. A study of low level vibrations as a power source for wireless sensor nodes. Comput. Commun. 2003, 26, 1131–1144. [CrossRef] Philipose, M.; Smith, J.R.; Jiang, B.; Mamishev, A.; Roy, S.; Sundara-Rajan, K. Battery-free Wireless Identification and Sensing. IEEE Pervasive Comput. 2005, 4, 37–45. [CrossRef] Franceschetti, G.; Gervasio, V. Wireless Power Transmission. A new science is borne. In Proceedings of the 2012 IEEE MTT-S International Microwave Workshop Series on Innovative Wireless Power Transmission: Technologies, Systems, and Applications, Kyoto, Japan, 10–11 May 2012; pp. 21–23. Kracek, J.; Švanda, M.; Mazanek, M.; Machac, J. Implantable Semi-Active UHF RFID Tag with Inductive Wireless Power Transfer. IEEE Antennas Wirel. Propag. Lett. 2016, 15, 1657–1660. [CrossRef] Borges, L.M.; Chávez-Santiago, R.; Barroca, N.; Velez, F.J.; Balasingham, I. Radio-frequency energy harvesting for wearable sensors. Healthc. Technol. Lett. 2015, 2, 22–27. [CrossRef] [PubMed] Kuhn, V.; Seguin, F.; Lahuec, C.; Person, C. A multi-tone RF energy harvester in body sensor area network context. In Proceedings of the 2013 Loughborough Antennas & Propagation Conference (LAPC), Loughborough, UK, 11–12 November 2013; pp. 238–241. Yeager, D.J.; Holleman, J.; Prasad, R.; Smith, J.R.; Otis, B.P. NeuralWISP: A Wirelessly Powered Neural Interface with 1-m Range. IEEE Trans. Biomed. Circuits Syst. 2009, 3, 379–387. [CrossRef] [PubMed]
Sensors 2017, 17, 2268
19.
20. 21. 22. 23. 24. 25. 26.
27. 28. 29.
12 of 12
Correia, R.; de Carvalho, N.B.; Fukuday, G.; Miyaji, A.; Kawasaki, S. Backscatter wireless sensor network with WPT capabilities. In Proceedings of the 2015 IEEE MTT-S International Microwave Symposium, Phoenix, AZ, USA, 17–22 May 2015; pp. 1–4. Jankowski-Mihułowicz, P.; W˛eglarski, M.; Pitera, G.; Kawalec, D.; Lichon, ´ W. Development board of the autonomous semi-passive RFID transponder. Bull. Pol. Acad. Sci. Tech. Sci. 2016, 64, 647–654. [CrossRef] Correia, R.; Carvalho, N.B.; Kawasaki, S. Continuously Power Delivering for Passive Backscatter Wireless Sensor Networks. IEEE Trans. Microw. Theory Tech. 2016, 64, 3723–3731. [CrossRef] Hu, M. On Measurements of Microwave E and H Field Distributions by Using Modulated Scattering Methods. IRE Trans. Microw. Theory Tech. 1960, 8, 295–300. Harrington, R.F. Small Resonant Scatterers and Their Use for Field Measurements. IRE Trans. Microw. Theory Tech. 1962, 10, 165–174. [CrossRef] Harrington, R. Electromagnetic scattering by antennas. IEEE Trans. Antennas Propag. 1963, 11, 595–596. [CrossRef] Harrington, R.F. Theory of loaded scatterers. Electr. Eng. Proc. Inst. 1964, 111, 617–623. [CrossRef] Belo, D.; Correia, R.; Pereira, F.; De Carvalho, N.B. Dual band wireless power and data transfer for space-based sensors. In Proceedings of the 2017 Topical Workshop on Internet of Space (TWIOS), Phoenix, AZ, USA, 15–18 January 2017; pp. 1–4. Texas Instruments. LM94021/LM94021Q Multi-Gain Analog Temperature Sensor; Texas Instruments: Dallas, TX, USA, 2013. Analog Devices. Micropower, 3-Axis, 2 g/4 g/8 g Digital Output MEMS Accelerometer; Analog Devices: Norwood, MA, USA, 2016. Texas Instruments. Mixed Signal Microcontroller MSP430F2132-E; Texas Instruments: Dallas, TX, USA, 2013. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).