IEEE JOURNAL OF ELECTROMAGNETICS, RF, AND MICROWAVES IN MEDICINE AND BIOLOGY, VOL. 2, NO. 2, JUNE 2018 109 Compact Multiband Wireless Energy Harvesting Based Battery-Free Body Area Networks Sensor for Mobile Healthcare Liu Yang, Yong Jin Zhou , Member, IEEE, Chao Zhang, Xin Mi Yang, Xue-Xia Yang , Senior Member, IEEE, and Chong Tan Abstract—This paper demonstrates a prototype of a selfsustained body area networks (BAN) sensor, which consists of the electrically small triple-band rectenna, the direct current (dc) energy management and storage module, the microcontroller, and the sensing and communication module. The proposed antenna is composed of corrugated metal-insulator-metal plasmonic structures, which covers triple frequency bands, including GSM-900, UTMS2100, and TD-LTE bands. Its electrical size is only 0.21 λ × 0.2 λ at 900 MHz. The gains reach 1 dBi, 2.64 dBi, and −0.19 dBi at 0.9 GHz, 2.025 GHz, and 2.36 GHz, respectively. A triple-band rectifier for low power application is designed to convert the harvested radio frequency (RF) power into dc power. The maximum RF to dc conversion efficiency of the rectifier reaches 59% when the input power is −10 dBm. The proposed compact BAN sensor based on multiband wireless energy harvesting is suitable for human body self-monitoring and mobile healthcare. Index Terms—Energy harvesting, body sensor networks, rectenna, rectifiers, plasmons. I. INTRODUCTION OBILE healthcare is getting more and more attention for prevention and better management of chronic diseases, nursing care of the aging society and saving medical expenses . The key technologies for mobile healthcare or remote medical care are body area networks (BANs) for the M Manuscript received October 12, 2017; revised December 22, 2017 and February 6, 2018; accepted March 14, 2018. Date of publication April 12, 2018; date of current version May 19, 2018. This work was supported in part by the National High-Tech Research Development Plan (863 plan) under Grant 2015AA016201 and in part by the Science and Technology Commission Shanghai Municipality under Grant SKLSFO2017-05 and Grant 13ZR1454500. (Corresponding author: Yong Jin Zhou.) L. Yang, C. Zhang, and X.-X. Yang are with the School of Communication and Information Engineering, Shanghai University, Shanghai 200444, China (e-mail:, firstname.lastname@example.org; email@example.com; yang.xx@shu. edu.cn). Y. J. Zhou is with the Key Laboratory of Specialty Fiber Optics and Optical Access Networks, Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication, Shanghai Institute for Advanced Communication and Data Science, Shanghai University, Shanghai 20044, China, and also with the State Key Laboratory of Transducer Technology, Chinese Academy of Sciences, Shanghai 200050, China (e-mail:,firstname.lastname@example.org). X. M. Yang is with the Soochow University, Suzhou 215006, China (e-mail:, email@example.com). C. Tan is with the Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai 200050, China (e-mail:, firstname.lastname@example.org). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JERM.2018.2817364 real-time monitoring of various physiological signals of human body . Nevertheless, the BAN must have smaller nodes relative to a conventional wireless sensor network (WSN) for comfortable user experience . Smaller nodes imply smaller batteries, creating strict tradeoffs between the consumed energy and the performance. The capacity of the high energy density of lithium-based batteries is limited in diminutive BAN enclosures. The need to replace or recharge batteries frequently makes BAN less desirable, which will hamper the widespread adoption of these devices in daily healthcare . The self-sustainable BAN has the potential to pave the road towards the massive utilization of wireless wearable sensors , . In this context, energy harvesting (EH) technologies, which take energy from ambient sources (such as mechanical, thermal, and electromagnetic (EM) sources), are used to power autonomous wireless systems . Wireless energy harvesting (WEH) harvests surrounding EM energy to supply continuous power to the selfsustainable standalone devices , which provides a solution to replace the battery or save maintenance cost. Although the main drawback of WEH is the low power density , the available ambient wireless sources keep increasing due to the ever expanding wireless communication and broadcasting infrastructure. Moreover, ambient EM energies are available at all day and night. Hence, EM energy becomes a relatively reliable and steady ambient energy source, which has been applied to wireless structural health monitoring sensors , continuous health monitoring sensors , biotelemetry communication , etc. Dedicated radio frequency (RF) sources are necessary for the wirelessly powered wearable sensors in –. Here, we focus on harvesting ambient RF energy from the air to power a BAN battery-free sensor node which may continuously communicate with personal monitor/remote medical staff through BT/WiFi or Internet [see Fig. 1(a)]. Fig. 1(b) shows the block diagram of a typical BAN batteryfree sensor node, which is composed of a rectifying antenna, a DC energy management and storage module, a microcontroller, a sensing module, and a communication module. The rectifyingantenna (rectenna) that converts the incident EM power into direct current (DC) power is the most vital device for the WEH system. Ambient EM signals are normally distributed over various frequency bands . A multiband or broadband rectenna is desirable to effectively capture the free energy from these frequency bands simultaneously. To design a multiband or broad- 2469-7249 © 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications standards/publications/rights/index.html for more information.
110 IEEE JOURNAL OF ELECTROMAGNETICS, RF, AND MICROWAVES IN MEDICINE AND BIOLOGY, VOL. 2, NO. 2, JUNE 2018 Fig. 1. (a) Future application of WEH battery-free sensor networks for daily healthcare. (b) Block diagram of a BAN battery-free node. band rectenna is very challenging, since the input impedance of the rectifier circuit varies as a function of the operating frequency, input power level, and load impedance. High efficiency multiband and broadband rectennas have been proposed –, however, there are still various limitations. First, due to BAN node placement variability and uncertainty about the user’s exposure to ambient energy, the antenna for WEH should be of wide half power beamwidth to harvest ambient energy incident from random angles. Second, the BAN nodes are generally smaller (less area covered) and have fewer opportunities for redundancy . Hence, compared to the rectifier circuit, power management system, or sensor module, the volume of the antenna for WEH is too large ,  and it should be electrically small. A wideband rectenna operating within 900–2450 MHz for wearable sensors in outdoor environments was introduced in . However, the rectenna is not compact and its size is large. A fully-autonomous integrated RF energy harvesting system for wearable applications was described, which harvests RF energy from GSM 900/1800 and wireless fidelity in 2.4 GHz on user request . However, the harvesting antenna is still a little large (150 mm × 150 mm). In this paper, a novel electrically small triple band rectenna has been proposed, which works at GSM-900, UTMS-2100, and TD-LTE bands. The antenna is composed of corrugated metalinsulator-metal (MIM) plasmonic structures and its electrical size is only 0.21 λ × 0.2 λ (66 mm × 70 mm) at 900 MHz. A triple-band low-power rectifier has also been proposed and combined with the antenna. The simulated peak efficiencies are 59%, 49%, 48% at the frequencies of 0.9 GHz, 2.025 GHz, 2.36 GHz, respectively, when the input power is −10 dBm. The measured RF to DC conversion efficiencies at 2.025 GHz agree well with the simulation results. When the input power is decreased to −11.1 dBm, the efficiency remains 47%. A prototype of a self-sustained body area sensor networks node has been demonstrated, where the electrically small triple-band rectenna has successfully supplied power for the node consisting of the DC energy management and storage module, the microcontroller, and the sensing and communication module. The proposed compact multiband WEH system is suitable for human body self-monitoring and daily healthcare. II. ANTENNA DESIGN Metamaterials are composed of sub-wavelength particles that can achieve parameters not possible within naturally occurring Fig. 2. (a) Front view and side view of the antenna. Dimensions: l = 66 mm, w = 70 mm, R = 16 mm, h 1 = 10 mm, h 2 = 12 mm, w 1 = 2.2 mm, l1 = 5 mm, g = 1 mm, r = 2.85 mm, w s = 2.33 mm, t = 0.018 mm, d = 1.016 mm. (b) Fabricated antenna. materials. Recently, metamaterials are also used to harvest EM energy in the microwaves regime, including a flower-like structure composed of four electrically small split-ring resonators (SRRs) , a circular slotted truncated corner square patch radiator placed on reactive impedance surface (RIS) , a parallel connection of five SRRs loaded with embedded devices , etc. However, the previous design was mainly for a single narrow frequency band and required a relatively high input power level. Loop antennas over artificial magnetic conductor surface for dual-band energy harvesting have been proposed , but the size of the antenna is not compact. Since spoof surface plasmons (SPs) are not constrained by the diffraction limit and can achieve subwavelength confinements to the EM waves , they have important potential applications in the miniaturization of spoof plasmonic circuits. Many antennas based on spoof SPPs have been demonstrated , . While the near field characteristics of spoof LSPs have been fully investigated , , their far field behaviors are still unknown. To our knowledge, only a sub-wavelength unidirectional antenna was designed by combining two spoof LSPs resonators . Here, we propose a multiband antenna based on spoof LSPs resonator, as shown in Fig. 2(a), which is composed of an annular ring slot and periodic array of T-shaped grooves. The corrugated slot line ring is printed on the substrate (Rogers RO4350), with relative dielectric constant of 3.48 and loss tangent of 0.004. It is fed by a 50-Ω microstrip line. The metal disk at the end of the microstrip conductor is used to increase the coupling degree of electromagnetic energy. The fabricated multiband antenna based on spoof LSPs resonator is shown in Fig. 2(b). In order to illustrate the influences of the grooves with different shapes, three different structures are shown in Fig. 3. Structure I is the conventional slot line ring. Structure II and III are the corrugated slot line ring with periodic array of rectangle grooves and T-shaped grooves, respectively. The dispersion curves for these three kinds of waveguides are calculated by use of the eigenmode solver of the commercial software, CST Microwave Studio. They are plotted in Fig. 4(a). It can be seen that the dispersion curve is lowered when the slot is corrugated. That means that the operating frequency of structure III will be
YANG et al.: COMPACT MULTIBAND WIRELESS ENERGY HARVESTING BASED BATTERY-FREE BAN SENSOR FOR MOBILE HEALTHCARE Fig. 3. 111 Different structures based on the slot ring. Fig. 4. (a) Dispersion curves of the three structures. (b) Simulated reflection coefficients of the three structures I-III. Fig. 5. (a) Simulated reflection coefficients for different h2 . (b) Simulated and measured reflection coefficients. the lowest when the wave vector β is fixed, since its dispersion curve is the lowest. It is known that the slot ring resonator antenna can form multimode resonances. For the above three different structures based on the slot ring, the circumference L of the slot ring and the multimode guided wavelength λn satisfy L ≈ n × λn (n = 1, 2, 3...), where n is a positive integer (corresponding to the resonant modes). For a fixed L (R = 16 mm), the guided wavelength λn at the multiple resonance modes can be calculated from the above equation. Then we can get βn by βn = 2π/λn and the operating resonant frequency can be obtained from the dispersion curves in Fig. 4(a). The simulated reflection coefficients using CST time domain solver are presented in Fig. 4(b). We can see that the resonant frequency of the first mode for the structure III is indeed the lowest, much lower than that of the first mode for the structure I. This indicates that more compact resonator antennas can be achieved based on such spoof plasmonic structures. The reflection coefficients for different h2 are plotted in Fig. 5(a). It can be seen that the resonant frequencies red shift when h2 is gradually increased from 10 mm to 12 mm. However, the offset is different for different resonant modes, which Fig. 6. Simulated and measured radiation patterns at (a) 0.9 GHz, (b) 1.575 GHz, (c) 2.025 GHz, and (d) 2.36 GHz. indicates that we can control the resonant frequencies by tuning the geometrical parameters. The reflection coefficients of the multiband antenna are measured by a vector network analyzer (Agilent N5227A), which are shown in Fig. 5(b). It can be observed that the measured result agrees well with the simulated one. The result shows that there are four resonant modes marked as m1 –m4 . The corresponding resonant frequencies are 0.9 GHz, 1.575 GHz, 2.025 GHz, and 2.36 GHz, which include the GSM900, UTMS-2100 and TD-LTE bands. Hence, the electrical size of the antenna is only 0.21 λ × 0.2 λ at 0.9 GHz. The radiation patterns of E-plane at the four resonant frequencies are simulated and measured, as shown in Fig. 6. The proposed multiband antenna exhibits radiation characteristics similar to the dipole antenna in all frequency bands, with broad half power beamwidth, since the electrical size are 0.2 λ, 0.35 λ, 0.45 λ, and 0.52 λ at 900 MHz, 1.575 GHz, 2.025 GHz, and 2.36 GHz, respectively. The measured antenna gains are 1 dBi, 1.6 dBi, 2.64 dBi and −0.19 dBi at 0.9 GHz, 1.575 GHz, 2.025 GHz, and 2.36 GHz, respectively. The measured radiation efficiencies are 42.2%, 61.3%, 72.6%, 32.8% at 0.9 GHz, 1.575 GHz, 2.025 GHz, and 2.36 GHz, respectively. The gain is expected to increase when the operating frequency is increased, due to the increased electrical size. In order to explain why the gain and radiation efficiency at 2.36 GHz are lower, we have conducted more simulations and found that the high-order mode (2.36 GHz) is more sensitive to the metal loss and dielectric loss than the other modes. First, when the metal is changed from perfect conductor (PEC) to copper, the radiation efficiencies at m2 and m3 decrease 1.1% and 4.9%, respectively, while the efficiency at m4 decreases 7.6%. When the loss tangent of the dielectric substrate is changed from 0.004 to 0.01 (the metal is still copper), the efficiencies at m2 and m3 decrease 3.2% and 9.7%, respectively, while the efficiency at m4 decreases 15.8%. Hence, the m4 mode (2.36 GHz)
112 IEEE JOURNAL OF ELECTROMAGNETICS, RF, AND MICROWAVES IN MEDICINE AND BIOLOGY, VOL. 2, NO. 2, JUNE 2018 TABLE I COMPARISON BETWEEN OUR DESIGN AND OTHERS’ Ref. 15 17 18 26 This work Frequency (GHz) Size (mm2 ) Gain (dBi) 1.8–2.5 0.9, 1.75, 2.15, 2.45 0.55–2.5 0.5, 0.875 0.9, 2.025, 2.36 70 × 70 155 × 155 160 × 160 500 × 500 66 × 70 2.5, 3.5, 3.8 9.8, 5, 4.8, 9.6 2.5∼5 8.2, 8.5 1, 2.64, −0.19 Fig. 8. (a) Simulated and measured S 1 1 of the rectifier. (b) The simulated conversion efficiency versus frequency at the input power of −10 dBm. Fig. 7. Layout and photograph of the rectifier. The parameters: w 0 2.64 mm, w 1 = 2.52 mm, w 2 = 1.92 mm, w 3 = 1 mm, w 4 = 1 mm, w 5 2.2 mm, w 6 = 2 mm, w 7 = 1.4 mm, l0 = 7.5 mm, l1 = 21.2 mm, l2 32.7 mm, l3 = 3.05 mm, l4 = 30.5 mm, l5 = 3 mm, l6 = 8 mm, l7 33.8 mm, θ1 = θ2 = 120, r1 = 11.5 mm, and r2 = 12.1 mm. = = = = is more sensitive to the metal loss and dielectric loss, which may lead to the decreased radiation efficiency and gain. The comparison between this work and some typical designs is given in Table I. The goal is to harvest energy from GSM900, UTMS-2100 and TD-LTE bands simultaneously. Hence 1.575 GHz is not listed. We can see that for the multiband antenna, our design is very compact and the electrical size is only 0.21 λ × 0.2 λ at 0.9 GHz. III. RECTIFIER DESIGN AND MEASUREMENT Although rectifiers based on CMOS technology are compact and can work at very low input power, they operate only at single band , . The rectifier using PCB technology is easy to be integrated with printed circuits and antennas can be designed to operate at multiple frequency bands . Here, a simple triple-band rectifier based on PCB technology is demonstrated. The substrate is F4B-2 with relative dielectric constant of 2.65 and loss tangle of 0.001, whose thickness is 0.8 mm. The layout of the tripe-band rectifier is shown in Fig. 7 and the inset gives the photograph of the fabricated rectifier. The rectifier is composed of an impedance match network (IMN), a Schottky diode, a DC-pass filter and a resistive load. The diode HSMS2850 is suitable for the low power application, whose threshold and breakdown voltages are 0.15 V and 3.8 V. However, its impedance is complex and frequency dependent. A closed-form solution to match a frequency dependent complex impedance load to a real impedance source at two arbitrary frequencies by using T-shaped transmission lines has been introduced in . With the T-shaped structure (TL0 , TL1 , and TL6 in Fig. 7), the impedance match at the first resonant frequency (around 0.9 GHz) and the last resonant frequency (around 2.36 GHz) Fig. 9. (a) Measurement system. (b) The simulated and measured conversion efficiency versus input power at 2.025 GHz. is achieved. However, this approach can match at only two frequency points. Therefore, to match the diode in triple-band, another T-shaped structure (TL1 , TL2 , and TL7 in Fig. 7) is adopted, which mainly affects the impedance match at the other resonant frequency (around 2.025 GHz). Hence, the final IMN consists of two T-shape transmission lines and the final optimized parameters of the rectifier by use of HFSS are provided in the caption of Fig. 7. The DC-pass filter is composed of two cascaded radial open stubs, which is used to block the fundamental wave and the harmonics and to smooth the output DC power. The resistive load of 2 kΩ is attached to the output port for collecting the DC power. The simulations were carried out by use of the software Advanced Design System (ADS). The simulated and measured reflection coefficients of the rectifier are shown in Fig. 8(a). Compared with the simulated result, the measured one has a slight shift, which may be caused by the manual welding process. Fig. 8(b) shows the simulated conversion efficiency versus frequency at the input power of −10 dBm. The simulated peak efficiencies are 59%, 49%, and 48% at the frequencies of 0.9 GHz, 2.025 GHz, and 2.36 GHz, respectively. It can be seen that the conversion efficiency of the rectifier is good. However, the dimension of the rectifier is still big, compared to the electrically small antenna. In the next work, the dimension of the rectifier can be miniaturized by use of metamaterials , lumped elements , stacked multilayer structure , and so on. IV. RECTENNA MEASUREMENT By combing the antenna with the rectifier, a multiband rectenna can be achieved. The experiment was carried out in an anechoic chamber. The measurement system is illustrated in Fig. 9(a). The distance between the standard horn antenna and the proposed antenna meets the far field requirement. The re-