IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 59, NO. 7, JULY 2011
2619
Parametric Design of Compact Dual-Frequency Antennas for Wireless Sensor Networks Simone Genovesi, Member, IEEE, Sergio Saponara, and Agostino Monorchio, Senior Member, IEEE
Abstract—A parametric study for the design of a planar compact dual-frequency antenna is presented. The proposed geometry and the suggested tailoring procedure provide a useful template for generating a single antenna able to serve as bidirectional node for a generic Wireless Sensor Network within the UHF and microwave frequency bands. The constructive parameters can be set to adapt the radiating device for different unlicensed bands and to comply with the international radiation power regulations. A careful design procedure will be described to tune the antenna template for working at the required frequencies. In order to prove the effectiveness of the suggested approach as well as the reliability of the adopted technique, comparisons between simulations and measurements of realized prototypes will be reported. Index Terms—Dual-frequency antenna, loop antenna, printed antenna, wireless sensor network.
I. INTRODUCTION
W
IRELESS SENSOR networks (WSN) are receiving a growing interest for several applications such as logistic, home automation, healthcare, structural monitoring, security/safety systems, intelligent transport systems [1]–[12]. They are typically [6] constituted by several client modules, distributed in the environment to be monitored. They are realized as compact printed circuit board (PCB) hosting one or more sensors, a battery or energy harvesting unit, a programmable IC for signal conditioning [12], [13] and a RF transmitter for communicating towards a server module. In many cases, the client’s RF unit can be configured so to support different modulations schemes (ASK, OOK, BPSK, ), frequencies (usually ISM unlicensed bands in the UHF range) and transmitted power. The configurability of the RF part is important also to cope with different national regulations. As an example, we mention tire pressure monitoring systems (TPMS) [1]–[3] where a wireless client hosting pressure and temperature sensors is mounted on each wheel’s rim transmitting tire status to a centralized receiver mounted on the car chassis. Usually, the network covers a local range and the maximum client RF transmitted power amounts to few tens of mW. A key issue for sensor networks is the availability of compact antennas at the wireless client side, possibly requiring only conventional PCB technologies for their production. Manuscript received November 15, 2009; revised November 12, 2010; accepted December 02, 2010. Date of publication May 10, 2011; date of current version July 07, 2011. The authors are with the Dipartimento di Ingegneria dell’Informazione, University of Pisa, 56122 Pisa, Italy (e-mail:
[email protected]; sergio.
[email protected];
[email protected]). 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/TAP.2011.2152313
In WSN a bidirectional communication scheme has to be preferred between client and server units. In addition to the wireless transmission of data from a client sensor to a server unit, a wireless link is needed from the server towards the clients for a wide range of tasks. For example, it can be necessary to acknowledge the safe receipt of data or to request their re-transmission in case of errors. Moreover, it may be useful to wake-up the clients and enable data transmission in wireless networks where the clients are normally in idle mode for power optimization or to configure the clients. Finally, two-way communication can be requested to exchange security codes for identification or for encrypted data transfer. To this aim, most systems known in literature adopt two separate antennas at the client side [1]–[3], [14] operating at distinct frequencies, thus increasing module size and cost. Typical solutions adopt a printed antenna at UHF together with a LF coil and its relevant demodulator (this solution is adopted for the TPMS in [3] or for safe entry system in [11]) or two separate antennas printed on opposite sides of the board and operating at distinct UHF frequencies (e.g., 433 MHz and 868 MHz for the TPMS in [2]). To exploit all added features of bidirectional communication schemes in WSN while avoiding the use of two separate antennas, this work proposes the design of a novel and compact double-loop antenna, patent filed [15], resonating at multiple frequencies in the UHF range, whose realization complies with standard low-cost PCB technologies. Another distinct characteristic of this work is that the antenna design is parametric; while the external loop is fixed to limit the overall size, the other elements (inner loop, tuner) are parameters that can be configured to allow the synthesis of different antennas. Therefore the antenna design is not fixed and customized only for a specific application, as in [7], [8]. Different antennas can be easily generated from the same architectural template having the same topology and the same general characteristics but resonating at different specific frequencies that can be tuned to meet the typical requirements of different national authorities. In fact, it is worth noting that 433 MHz and 868 MHz are the typical frequencies used in Western Europe, 315 MHz, 450 MHz and 915 MHz in US, 315 MHz and 915 MHz in South America, 433 MHz and 915 MHz in Australia [16]–[18]. Hereafter in Section II the parametric antenna design is proposed and the effects of the sizing of its building elements (the tuning element, the inner and outer loops, the matching network) on the resonating frequencies are discussed. Some tuning heuristic rules are derived in Section III allowing to generate and realize the desired antenna configuration for a given target of dual-band working frequencies, starting from the parametric antenna template. The antenna design and the configuration process are described in Section III, while Section IV
0018-926X/$26.00 © 2011 IEEE
2620
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 59, NO. 7, JULY 2011
Fig. 1. Top view of the antenna template.
Fig. 2. Comparison between the antenna input impedance without the inner loop (continuous and dotted lines with triangles) and with the inner loop (continuous and dotted lines). The real part is a dotted line while the imaginary part is continuous.
illustrates the generation, realization and measurements of distinct multi-frequencies antennas presented as case studies. Section V proves that the achieved performances of the novel proposed antennas are useful to realize wireless networking systems compliant with the radiation power regulations (US FCC [19] and ETSI [20]). Finally, conclusions are drawn in Section 6. II. ANTENNA TEMPLATE DESCRIPTION The configuration of the antenna comprises an outer loop, an inner loop and a tuning element (Fig. 1). The outer loop is connected to the source, or the electronic circuitry, by using a simple matching network. In order to show the effect of each single basic element of the antenna template, we consider the investigated radiating device printed on a commercially availmm) and able dielectric substrate ( a trace width equal to 1.25 mm. It is important to point out that we are interested in the impedance of the antenna determined at the open ends of the outer loop. This feeding configuration allows the antenna to be easily connected to electronic circuitry and RFIC input/output and, at the same time, prevents the use of balun or via-hole. The width of the trace equal to 1.25 mm has been chosen according to the solutions commonly adopted in commercial devices but a slightly different width can be employed to perform the design process. As an example, an outer cm has been loop of fixed dimension chosen to perform our parametric analysis but different sizes can be adopted depending on the available space on the PCB. These dimensions refer to the solution proposed in [3]; however, as additional advantage, we now propose an antenna design that occupies the same area but supports multiple frequencies. It is important to highlight that the perimeter of the outer loop is related to the value of the higher resonance, as it will be deeply discussed in the following. Therefore, a good starting guess for the size is represented by an outer loop with a perimeter close to the free-space wavelength at the higher frequency (we can find the same resonance frequency by using, for instance, dimensions of 4.2 cm 4.2 cm instead of 5.3 cm 3.2 cm since
Fig. 3. The increase of the inner loop length lowers both the resonance frequencies F (continuous line) and F (dashed line). The ratio (dotted line) between F and F reveals a faster decrease of F with the increase of the inner loop length.
the perimeter is almost the same). Then, if the antenna dimension does not match the imposed space requirement, a proper value of the matching network and the tuner element allows to shrink it providing a device with the same resonance frequency but with a smaller size. A. Effect of the Inner Loop The most important effect of the presence of the inner element , while is the occurrence of the lower resonance frequency is only slightly changed. In the higher resonance frequency Fig. 2 we show the input impedance of the same antenna with and without the inner loop. If the length of the investigated element is increased by and decreases while folding the inner loop, the value of increases (Fig. 3). Therefore, the inner loop the ratio while its length determines the first resonance frequency has an effect on both the resonance frequencies, although decreases more rapidly than .
GENOVESI et al.: PARAMETRIC DESIGN OF COMPACT DUAL-FREQUENCY ANTENNAS FOR WIRELESS SENSOR NETWORKS
Fig. 4. Changes in the inner loop position (a), determine an effect of the inner loop position on both resonance frequencies (b).
Fig. 5. Comparison between the antenna input impedance without the tuning element (continuous and dotted lines with triangles) and with the inner loop (continuous and dotted lines). The real part is a dotted line while the imaginary part is continuous.
2621
Fig. 6. The increase of the tuner length slightly changes F (continuous line) while F (dashed line) shifts toward lower values. The ratio between F and F (dotted line) decreases in the investigated range (no tuner-4.6 cm).
Fig. 7. Higher values of the inductance in the matching network determine smaller values of both the resonance frequencies although F (dashed line) decrease more rapidly than F (continuous line), as the ratio between F and F (dotted line) highlights.
C. Effect of the Matching Network The position of the inner loop influences the resonance frequencies. In Fig. 4 it is possible to observe that increasing the offset (Fig. 4(a)) between the tuner and the inner loop raises F2 while F1 slightly decreases (Fig. 4(b)). In particular, the dB bandwidth for the first resonance remains value of the stable around 1.0% whereas for the second one we obtain 1.5%. These values are typical for wireless sensor network applications where the amount of information to be transferred requires low bit-rates and they also agree with the ones presented in [9]. Therefore, although the resonance frequencies shift as shown in Fig. 4, the impedance bandwidth is not affected and remains stable for all offsets. B. Effect of the Tuning Element The length of the tuning element placed between the outer and produces a small and the inner loop affects the value of (Fig. 5). variation of More in detail, if we increase the length of the tuning element, decreases as well as the ratio (Fig. 6).
The matching network considered for the study comprises only an inductance connected to the outer loop. The considered inductance is in the range of tens of nH. For these values surface mount inductors exist that can be easily soldered to the PCB trace just before the source or the electronic circuitry. As shown in Fig. 7, the increase of the inductance value determines a shift toward lower resonance frequencies whereas the real part of the input impedance at the resonance varies within a set of reasonable values for the realization of a standard matching network (Fig. 8). III. TUNING OF THE ANTENNA TEMPLATE Each element of the antenna template has to be carefully dimensioned and tuned in order to obtain the desired resonance frequencies. In particular, as pointed out in the previous section, even a single change in one element affects both the resonance and and their ratio . This section will frequencies provide guidelines to exploit the antenna framework for the design of an antenna with two resonances within the ISM band at and . the desired frequencies
2622
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 59, NO. 7, JULY 2011
Fig. 8. The real part of the input impedance is within an acceptable range of values for both the frequencies F (continuous line) and F (dashed line).
As well known, the design can be subject to a series of limitations such as the allowed space for the radiating device or the discrete values assumed by the lumped loads of the matching network. As a case study, we consider again the dimensions of the outer loop equal to 5.3 cm 3.2 cm in order to design an antenna with two resonance frequencies satisfying the Australia MHz and MHz). requirements ( The first step is to obtain an antenna with a resonance and, at the same time, a resonance . Basically, this means that we have to tune the framework in order to achieve the first resonance by choosing a proper matching network and the inner loop length. To accomplish this goal, we set the tuner element as long as possible at this stage since its function will in order to make the antenna resbe to increase the value of onating also at . The aforementioned rules will suggest the changes that have to be made to the length or the position of the elements. For example, in order to increase the first resonance we have to decrease the inner loop size or the value of the inductance of the matching network. Suppose we start our tuning procedure with a matching network comprising an inductance of 33 nH and an inner loop with a length of 25 cm. In this case, we have a value of around 380 MHz and a second resonance frequency around 870 MHz. To allow our framework to satisfy the requirements of the first step we need to decrease the inner loop until its length is equal to with MHz. Next, since 19.5 cm. and hence , we have still to tune the second resonance frequency and therefore we reduce the tuner element from 2.3 cm to 2.15 . All the presented results have been valcm to obtain idated by using Ansoft HFSS [21] and are reported in Fig. 9. IV. MEASUREMENTS The procedure described in the previous section has been successfully applied for different sets of resonance frequencies, also including all the resonance frequencies mentioned in Section I. Moreover, in order to validate the simulated results, two prototypes have been realized and measured. The former is an antenna with two resonances in accordance with the
Fig. 9. Tuning of the antenna framework. The dotted lines are the real part of input impedances while the continuous lines illustrate the imaginary parts. The MHz and lines with triangles refers to the starting configuration with F MHz, the lines with circles to the antenna with F F and the F remaining lines to the final antenna satisfying the imposed requirements.
= 870
= 380 =
Fig. 10. Front view of the prototype for the double loop antenna satisfying the requirements of Australia resonance frequencies.
Australia requirements whereas the latter complies with the South America standard. An half-loop version of both of the antennas was fabricated (Figs. 10 and 11) and its measured input impedance was compared to the simulated results. We recurred to an half-loop version for the fabrication because it allows to validate the antenna performance while preventing the need to design and build a balun for the antenna. The dimensions of the half-loop antennas were determined by using HFSS, following the aforementioned rules for the tuning. More in detail, the tuner length is equal to 4 cm and the inner element is 21.5 cm long for the Australia configuration while the tuner is 4.6 cm and the inner element is 39 cm length for the South America antenna. Each device was printed on a commercially available dielectric substrate mm) with a trace width equals to ( 1.25 mm. The antenna is excited via a SMA connector with the outer conductor connected directly to the ground plane and the center conductor connected to the matching network of outer conductor of the half-loop. According to images’ theorem, the impedance of an half-loop above an infinite ground plane is
GENOVESI et al.: PARAMETRIC DESIGN OF COMPACT DUAL-FREQUENCY ANTENNAS FOR WIRELESS SENSOR NETWORKS
2623
Fig. 11. Front view of the prototype for the double loop antenna satisfying the requirements of South America resonance frequencies.
Fig. 13. Comparison between simulation (lines) and measurements (dots) of the input impedance related to the second resonance (915 MHz) of Australia configuration: real part (continuous line) and imaginary part (dashed line) are shown.
Fig. 12. Comparison between simulation (lines) and measurements (dots) of the input impedance related to the first resonance (433 MHz) of Australia configuration: real part (continuous line) and imaginary part (dashed line) are shown.
equal to one-half of the corresponding full loop. If the ground plane is sufficiently large, this statement still holds and we can use it to estimate the impedance of the half-loop. The input impedances of the fabricated half-loops were used to estimate the performance of the radiating device and comparisons with the numerical results have been reported in order to assess the reliability of the proposed design procedure. A Wiltron 37311A network analyzer was used to measure the input impedance versus frequency of the fabricated prototypes. In Fig. 12 we compare the real and imaginary part of the simulated and measured input impedance at the first and second resonance of the Australia configuration, respectively. The real part of the measured input impedance for the first resonance is in good agreement with the simulated one and also the measured imaginary part is close to the computed one since there is only a 4% off-set discrepancy with respect to the expected resonance frequency of 433 MHz (mainly due to technology and component spreading for the adopted manufacturing process). The real part of the measured input impedance for the second resonance (Fig. 13) exhibits a behavior similar to the simulated curve whereas the imaginary part is quite close to the expected one of 915 MHz (0.14% of discrepancy). The comparison between the real and imaginary part of the input impedance for the South America configuration is shown in Fig. 14 for the first resonance. The measured real part of the
Fig. 14. Comparison between simulation (lines) and measurements (dots) of the input impedance related to the first resonance of South America configuration: real part (continuous line) and imaginary part (dashed line) are shown.
impedance at the first resonance is in accordance with the simulated results and also the imaginary part of the prototype is close to the expected 315 MHz (4% of discrepancy). At the second resonance, the real part of the input impedance (Fig. 15) has a similar trend to the simulated one and assumes acceptable values whereas the imaginary part is quite close to the estimated one (2.29% of discrepancy). The simulated results are in good agreement with the measured one, therefore the proposed procedure for the design can be considered reliable. The small discrepancy (always below 4%) could be determined by a series of different causes such as the tolerance of the components (5% for the inductance of the matching network), the effect of the welding in the half-loop prototypes, a non ideal insertion of the SMA connector to the antenna and the ground plane. However, a further tuning can be easily done by trimming the printed antenna. V. RADIATION PATTERN AND LINK BUDGET In order to characterize the performance of the proposed antenna configuration, radiation patterns for the Australian and
2624
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 59, NO. 7, JULY 2011
Fig. 15. Comparison between simulation (lines) and measurements (dots) of the input impedance related to the second resonance of South America configuration: real part (continuous line) and imaginary part (dashed line) are shown.
South American configuration are reported in Figs. 16 and 17, respectively. As it can be noticed, the pattern retains an omnidirectional behavior on the horizontal plane for both resonance frequencies for the Australian as well as the South American configuration, thus providing a good connection among the nodes of the sensing network. The gain values of the Australia configuration are equal to dB and dB while those of the South America condB and dB respectively at the first figuration are and second resonance. The values of the gain are comparable to those find in [9] and [10] which are recent works in literature targeting similar applications of that considered in our paper. Since the proposed antenna can be employed in the sensing node of a wireless system, it has to be compliant with the radiation power regulations (US FCC and ETSI). Let us consider our antenna on each sensing node and a standard patch antenna (7.0 dB of gain), working at the same frequencies, at the master node. The MCU (Micro Controller Unit) in the master node generally hosts an RF transceiver able to receive signals with a sensitivity dBm, therefore the sensing node has to deliver a below very low power level. At this purpose, currents of a few mA are sufficient [2], [3] for the sensing node to transmit an intelligible signal to the master node. On the other side, particularly in passively-powered [22], [23] or power-optimized [2] sensor networks, the sensing node has an RF rectifier, which delivers outputs compliant with standard microcontroller voltage levels. The RF rectifiers can work properly if the received power at the dBm to dBm, depending on antenna is in the range of the implementation technology [22], [23]. For the case of Australian configuration, the sensing unit communicates with the master node at 433 MHz and the master node uses an RF link at 915 MHz to send data to the sensing node. By using the well known Friis formula in the maximum radiation condition and by supposing a distance between transmitter and receiver equals to 3.0 m, the required power for the 915 MHz transmitter master node ranges from tens of mW up to 105 mW (Fig. 18), well below the allowed transmitter power
Fig. 16. Normalized radiation pattern for Australia configuration at (a) first resonance of 433 MHz; (b) second resonance 915 MHz. The line with dot markers cut, the triangular markers to the cut and the refers to the square markers to plane.
=0
= 90
= 90
limit (500 mW). The aforementioned sensitivity of the MCU RF transceiver guarantees the link at 433 MHZ to work properly with much less power. In order to assess the maximum possible communication distance and hence the covered range of the system, we have studied the necessary level of transmitted power with the increment of the distance between the RF transceiver of the master node and the receiver at the sensing node. As already mentioned, if we consider a distance of 3.0 m we need to transmit few tens of mWs if we employ a performing RF dBm) or around 100 mW if we use an ordinary rectifier (
GENOVESI et al.: PARAMETRIC DESIGN OF COMPACT DUAL-FREQUENCY ANTENNAS FOR WIRELESS SENSOR NETWORKS
2625
Fig. 18. Relation between transmitted and received power at distance equal to 3.0 m between transmitting and receiving antennas for the case of Australian configuration. The dotted line refers to the received power at 433 MHz whereas the continuous one to the received power at 915 MHz.
Fig. 17. Normalized radiation pattern for South America configuration at (a) first resonance of 315 MHz; (b) second resonance 915 MHz. The line with dot cut, the triangular markers to the cut and markers refers to the the square markers to plane.
=0 = 90
Fig. 19. Relation between transmitted and received power at 915 MHz for three different distances antennas for the case of Australian configuration. The continuous line refers to the received power at a distance equal to 3.0 m, the dashed line to 6.0 m and the dotted line to 10.0 m. The grey zone indicates the interval of received power necessary to a correct behavior of the RF rectifier.
= 90
one ( dBm). As shown in Fig. 19, if we double the distance (6.0 m), we have to transmit slightly less than 100 mW if we exploit a good RF rectifier whereas it is necessary to increase the power to 400 mW for the less sensitive one. Even in the worst case we are still under the limit of 500 mW of transmitted power. If we go farther and reach 10.0 m from the transmitter, we are still able to have a reliable communication between the master node and the sensing node if we use an RF rectifier with dBm (Fig. 19). a good sensitivity, at least For the case of South America configuration, the sensing unit communicates with the master MCU at 315 MHz and the master
MCU uses an RF link at 915 MHz to send data to the node. If we set again the distance between transmitter and receiver to 3.0 m, the maximum necessary level of transmitted power is required for the MCU working at 915 MHz, ranging from tens of milliwatts up to 125 mW (Fig. 20). From the analysis of the link budget (minimum transmitted power versus the distance between the nodes), a result similar to the previous case emerges. In fact as shown in Fig. 21, the communication is possible for a distance equal to 6.0 m with the respect of the 500 mW limit even with an ordinary RF rectifier. If the distance is set equal to 10 m, the link is possible only by employing an RF rectifier dBm. with a sensitivity at least equal to
2626
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 59, NO. 7, JULY 2011
allowed frequencies by following the described tailoring procedure. Thus, it is possible to realize a bidirectional node for a generic Wireless Sensor Network simply by using a single low-cost antenna. Moreover, careful attention has been paid to the fulfillment of the international radiation power norms but, at the same time, a communication range up to 10.0 m has been guaranteed. The good agreement between simulations and measurements on the realized prototypes has proved the reliability of the design approach.
REFERENCES
Fig. 20. Relation between transmitted and received power at distance equal to 3.0 m between transmitting and receiving antennas for the case of South America configuration. The dotted black line refers to the received power at 315 MHz whereas the continuous black one to the received power at 915 MHz.
Fig. 21. Relation between transmitted and received power at 915 MHz for three different distances antennas for the case of South America configuration. The grey continuous line refers to the received power at a distance equal to 3.0 m, the black dotted one to 6.0 m and the continuous black one to 10.0 m. The grey zone indicates the interval of received power necessary to a correct behavior of the RF rectifier.
As a result of the examples shown, we can affirm that we are able to successfully employ the proposed antenna in a wireless sensor network, also with passively-powered sensing nodes, which complies with the US FCC and ETSI regulations and, at the same time, guarantees a communication range of 6.0 m or even 10.0 m with a performing RF rectifier.
VI. CONCLUSION A configuration of a dual-frequency planar compact antenna has been proposed. This antenna template provides a radiating device with two resonance frequencies within the UHF and microwave band, which can be adapted to the different worldwide
[1] S. Ergen, A. Sangiovanni-Vincentelli, X. Sun, R. Tebano, S. Alalusi, G. Audisio, and M. Sabatini, “The tire as an intelligent sensor,” IEEE Trans. Comput.-Aided Design Integrat. Circuits Syst., vol. 28, no. 7, pp. 941–955, Jul. 2009. [2] F. Iacopetti, S. Saponara, and L. Fanucci, “Improving power efficiency and reliability in RF tire pressure monitoring modules,” in Proc. IEEE ICECS, 2007, pp. 878–881. [3] J. Burgess, “Freescale Tire Pressure Monitor System Demo,” AN1951, Jun. 2005, rev. 2. [4] A. Sabata and S. Brossia, “Remote Monitoring of Pipelines Using Wireless Sensor Network,” U.S. patent 7526944, May 2009. [5] G. Marrocco, “The art of UHF RFID antenna design: Impedancematching and size-reduction techniques,” IEEE Antennas Propag. Maga., vol. 50, no. 1, pp. 66–79, Feb. 2008. [6] J. P. Lynch and K. J. Loh, “Summary review of wireless sensors and sensor networks for structural health monitoring,” Shock Vibr. Digest, pp. 38–91, 2006. [7] S. Genovesi, A. Monorchio, and S. Saponara, “Double-loop antenna for wireless tyre pressure monitoring,” Electron. Lett., vol. 44, no. 24, pp. 1385–1386, Nov. 2008. [8] K. S. Leong, M. L. Ng, and P. H. Cole, “Dual-frequency antenna design for RFID application,” presented at the 21st Int. Technical Conf. on Circuits/Systems, Computers and Communications (ITC-CSCC 2006), Chiang Mai, Thailand, Jul. 2006. [9] K. Tanoshita, K. Nakatani, and Y. Yamada, “Electric field simulations around a car of the tire pressure monitoring system,” IEICE Trans. Commun., vol. E90-B, no. 9, pp. 2416–2421, Sept. 2007. [10] M. Brzeska, J. Pontes, G.-A. Chakam, and W. Wiesbeck, “RF-design characterization and modelling of tire pressure sensors,” in Proc. EuCAP, 2007, pp. 1–5. [11] M. Brzeska and G.-A. Chakam, “Modelling of the coverage range for modern vehicle access systems at low frequencies,” in Proc. 37th Eur. Microwave Conf., Oct. 2007, pp. 771–774. [12] S. Saponara, E. Petri, L. Fanucci, and P. Terreni, “Sensor modeling, low-complexity fusion algorithms and mixed-signal IC prototyping for gas measures in low-emission vehicles,” IEEE Trans. Instr. Meas., vol. 60, no. 2, 2011. [13] S. Saponara, L. Fanucci, and P. Terreni, “Architectural-level power optimization of microcontroller cores in embedded systems,” IEEE Trans. Ind. Electron., vol. 54, no. 1, pp. 680–683, 2007. [14] ATMEL, “LF—Wake Up Demonstrator ATAK5276,” 2007. [15] “Antenna Multirisonante Compatta a Doppia Spira Con Elemento Parassita (“Compact Multiresonant Double Loop Antenna With Parasitic Element”),” Italian Patent P.I. 2009-A000131, Oct. 22, 2009. [16] [Online]. Available: http://www.atmel.com/products/SmartRF/ [17] A. Harney, “Design, simulate, and document proprietary wireless systems,” Analog Dialogue, vol. 42, no. 10, pp. 15–17, Oct. 2008. [18] Freescale Semiconductor, “MC33596 data sheet,” Mar. 2009, rev. 4. [19] FCC Codes of Regulation, pt.15. Available online: [Online]. Available: http://www.access.gpo.gov/nara/cfr/waisidx_08/47cfr15_08.html [20] ETSI EN 301 489-3,1.4.1 (2002–08) [Online]. Available: http://www. etsi.org/WebSite/Technologies/ShortRangeDevices.aspx [21] [Online]. Available: http://www.ansoft.com/products/hf/hfss/ [22] T. Le, K. Mayaram, and T. Fiez, “Efficient far-field radio frequency energy harvesting for passively powered sensor networks,” IEEE J. Solid State Circuits, vol. 43, no. 5, pp. 1287–302, May 2008. [23] G. De Vita and G. Iannaccone, “Design criteria for the RF section of UHF and microwave passive RFID transponders,” IEEE Trans. Microw. Theory Tech., vol. 53, no. 9, pp. 2978–2990, Sep. 2005.
GENOVESI et al.: PARAMETRIC DESIGN OF COMPACT DUAL-FREQUENCY ANTENNAS FOR WIRELESS SENSOR NETWORKS
Simone Genovesi (S’99–M’07) received the Laurea degree in telecommunication engineering and the Ph.D. degree in information engineering from the University of Pisa, Pisa, Italy, in 2003 and 2007, respectively. Since 2003, he has been collaborating with the Electromagnetic Communication Laboratory, Pennsylvania State University, University Park. From 2004 to 2006, he was a Research Associate at the ISTI institute of the National Research Council of Italy (ISTI-CNR), Pisa. He is currently a Research Associate at the University of Pisa. His research is focused on metamaterials, antenna optimization and evolutionary algorithms.
Sergio Saponara received the Laurea degree (cum laude) and the Ph.D. degree in electronic engineering from the University of Pisa, Pisa, Italy, in 1999 and 2003, respectively. In 2002, he was with IMEC, Leuven (B), Belgium, as a Marie Curie Research Fellow. Since 2001, he has been collaborating with the Consorzio Pisa Ricerche, Pisa. He holds the Chair of Electronic Systems for Automotive and Automation at the Faculty of Engineering, University of Pisa, where he is a Senior Researcher at in the field of electronic circuits and systems for telecom, multimedia, space and automotive applications. He coauthored more than 100 scientific publications and holds four patents. Dr. Saponara is also Research Associate of CNIT and INFN and served as Guest Editor of special issues on international journals and as a program committee member of international conferences.
2627
Agostino Monorchio (S’89–M’96–SM’04) received the Laurea degree in electronics engineering and the Ph.D. degree in methods and technologies for environmental monitoring from the University of Pisa, Pisa, Italy, in 1991 and 1994, respectively. During 1995, he joined the Radio Astronomy Group, Arcetri Astrophysical Observatory, Florence, Italy, as a Postdoctoral Research Fellow, in the area of antennas and microwave systems. He has been collaborating with the Electromagnetic Communication Laboratory, Pennsylvania State University (Penn State), University Park, and he is an Affiliate of the Computational Electromagnetics and Antennas Research Laboratory. He has been a Visiting Scientist at the University of Granada, Spain, and at the Communication University of China in Beijing. He is currently an Associate Professor in the School of Engineering, University of Pisa, and Adjunct Professor at the Italian Naval Academy of Livorno. He is also an Adjunct Professor in the Department of Electrical Engineering, Penn State. He is on the Teaching Board of the Ph.D. course in “Remote Sensing” and on the council of the Ph.D. School of Engineering “Leonardo da Vinci” at the University of Pisa. His research interests include the development of novel numerical and asymptotic methods in applied electromagnetics, both in frequency and time domains, with applications to the design of antennas, microwave systems and RCS calculation, the analysis and design of frequency-selective surfaces and novel materials, and the definition of electromagnetic scattering models from complex objects and random surfaces for remote sensing applications. He has been a reviewer for many scientific journals and he has been supervising various research projects related to applied electromagnetic, commissioned and supported by national companies and public institutions. Dr. Monorchio has served as Associate Editor of the IEEE ANTENNAS AND WIRELESS PROPAGATION LETTERS. He received a Summa Foundation Fellowship and a NATO Senior Fellowship.
