A NOVEL 10 GHZ SUPER-HETERODYNE BIO-RADAR SYSTEM BASED ON A FREQUENCY MULTIPLIER AND PHASE-LOCKED LOOP

Transcription

1 Progress In Eletromagnetis Researh C, Vol. 19, , 2011 A NOVEL 10 GHZ SUPER-HETERODYNE BIO-RADAR SYSTEM BASED ON A FREQUENCY MULTIPLIER AND PHASE-LOCKED LOOP S.-S. Myoung, Y.-J. An, and J.-G. Yook Yonsei University 262 Seongsanno, Seodaemun-gu, Seoul , Korea B.-J. Jang Kookmin University Jeongneung-dong, Seongbuk-gu, Seoul , Korea J.-H. Moon PhilTeh Co., Ltd Sangdaewon-dong, Jungwon-gu, Seongnam-si, Gyeonggi-do , Korea Abstrat This paper presents a novel 10 GHz bio-radar system based on a frequeny multiplier and phase-loked loop (PLL) for non-ontat measurement of heartbeat and respiration rates. In this paper, a 2.5 GHz voltage ontrolled osillator (VCO) with PLL is employed as a frequeny synthesizer, and 10 GHz ontinuous wave (CW) signal is generated by using frequeny multiplier from 2.5 GHz signal. This paper also presents the noise harateristis of the proposed system, and the analysis result shows that the same signal-to-noise-ratio (SNR) performane an be ahieved with the proposed system based on the frequeny multiplier ompared with the onventional system with idential arrier frequeny. The experimental results shows exellent vital-signal measurement up to 100 m without any additional digital signal proessing (DSP), thus proving the validity of the proposed system. Reeived 13 November 2010, Aepted 12 January 2011, Sheduled 27 January 2011 Corresponding author: Seong-Sik Myoung (myoungss@yonsei.a.kr).

2 150 Myoung et al. 1. INTRODUCTION The bio-radar based on Doppler effet has inspired a great deal of interests in measuring vital signals, suh as heartbeat and respiration, due to its simpliity as well as non-invasive non-ontat measurement sheme. Droitour showed that the phase noise of loal osillator (LO) signal soure is greatly redued with the range orrelation effet based on Budge s study [1 3]. As a result, it has been onsidered that the phase noise from the VCO is not an important parameter in bioradar system, even though the heartbeat and respiration signals are in very low frequeny regime ranging from 1 Hz to 30 Hz and the phase noise at that offset frequeny of a onventional VCO is very high. So, diret onversion transeiver arhiteture has been generally employed with a shared LO for up- and down-onversion mixers, and most of reently reported bio-radar systems are based on free running VCO [4 7]. However, the team of authors arefully analyzed the phase noise effet with onventional parameters of ommerial omponents with respet to antenna as well as mixer leakage effets, and the analysis result shows that the phase noise of LO is still dominant fator of the SNR degradation in a bio-radar system due to the low isolation between T x and Rx bloks [8]. Moreover, the bio-radar system with PLL has been proposed, and the experimental results showed the validity of the bio-radar system with improved LO soure [9]. In this paper, performane of a 10 GHz Doppler-effet-based bio-radar system with a PLL is presented based on 10 GHz super-heterodyne arhiteture with PLL and frequeny multiplier for the new frequeny band a primary basis worldwide [10] from to 10.5 GHz. Moreover, the noise mehanism in the proposed super-heterodyne arhiteture is analyzed, and the validity of the proposed bio-radar system is presented with the noise analysis and experimental results GHZ SUPER-HETERODYNE BIO-RADAR 2.1. Bio-radar System Topology Figure 1 shows the blok diagram of the proposed 10 GHz superheterodyne bio-radar system. As presented in [9], the performane of bio-radar system an be greatly improved with PLL employment. The frequeny synthesizer in 2.5 GHz is designed with PLL, and 10 GHz soure signal is generated by using a frequeny multiplier (quadrupler). So, it is possible to generate very stable CW signal with extremely low phase noise harateristis. The reeiver onsists of the IQdemodulator and down-onversion mixer, where the IQ-demodulator is driven by 2.5 GHz signal generated by the LO, while the down-

3 Progress In Eletromagnetis Researh C, Vol. 19, GHz Heterodyne Bio-radar System Tx Antenna (2 1 array) 3 4 Power divider BPF Tx Amp. T(t) 0 dbm BB Proessing L(t) B(t) IQ Demod. R IF(t) LPF BPF Rx Antenna (2 1 array) LO Amp. M(t) R(t) Down-onversion mixer BPF LNA Figure 1. system. The blok diagram of the proposed 10 GHz bio-radar onversion mixer employs the 7.5 GHz signal generated by another frequeny multiplier (tripler). This reeiver topology is a superheterodyne arhiteture with 2.5 GHz intermediate frequeny (IF), whih is the PLL operating frequeny or the fundamental frequeny of the frequeny synthesizer. With the proposed topology, the IQdemodulator requirement an be less strit, and IF band filtering with relatively sharp skirt harateristis is possible. These advantages are the ommon harateristis of a super-heterodyne arhiteture. By using the frequeny multipliers, tripler as well as quadrupler, only one signal synthesizer with PLL is neessary. In the proposed super-heterodyne bio-radar system, the output of the signal synthesizer (L(t)) as well as the multiplied signals (T (t) and M(t)) in Figure 1 an be expressed as followings: L(t) = os(2πf t + φ(t)) (1) T (t) = os(8πf t + 4φ(t)) (2) M(t) = os(6πf t + 3φ(t)) (3) where, f and φ(t) are the fundamental frequeny and phase noise of the frequeny synthesizer, respetively. The reeived signal, refleted and modulated by human body, an be expressed as a funtion of the time-varying distane between the bio-radar and target human body (d(t)) as following equations: ( 2d R(t)=A 1 os 8πf t t d(t) ) ( 2d +4φ t t d(t) ) (4)

4 152 Myoung et al. where, is the speed of light. The distane (d(t)) an be deomposed with the onstant distane from the bio-radar system to the human body (d 0 ) and very small variation of hest (x(t)) due to heartbeat and respiration. With the small angle approximation, the down-onverted baseband signal (B(t)) in Figure 1 an be alulated as followings: [ 16πd0 B(t) = A 2 os + 16πx(t) 4φ λ λ ] +3φ(t τ m ) + φ(t) + θ 1 + θ 2 ( t 2d ) 0 τ m where, θ 1 and θ 2 are the phase delays from LO to down-onversion mixer and to IQ-demodulator, respetively, and τ m is the time delay from RF to IF due to the down-onversion mixer. The phase term of the above baseband signal, B(t), onsists of three parts, whih are followings: 16πd 0 + θ 1 + θ 2 (6) λ 16πx(t) (7) and ( 4φ t 2d 0 τ m (5) λ ) + 3φ(t τ m ) + φ(t) (8) The first term is onstant, and it does not affet the bio-radar system operation. The seond and third terms are the measured biosignal and down-onverted phase noise in baseband, respetively. The measured signal amplitude by the proposed bio-radar based on the frequeny multiplier is inreased by four times ompared with the signal measured by the onventional bio-radar system due to four times higher arrier frequeny. The phase noise is also inreased due to frequeny multiplier, but it is onsidered that the degree is not serious. For example, if the time delay due to the down-onversion mixer is negligibly small ompared with the time delay of the signal propagation from T x to Rx, the phase noise in (8) and baseband signal in (5) an be approximated as followings: B(t) os ( 4φ t 2d 0 ( θ + 16πx(t) 4φ λ ) + 4φ(t) (9) ( t 2d ) ) 0 + 4φ(t) (10) The phase noise in (8) is exatly four times of the phase noise in the onventional diret onversion bio-radar system employing the

5 Progress In Eletromagnetis Researh C, Vol. 19, fundamental signal of the frequeny synthesizer. Finally, there is not any SNR degradation due to frequeny multiplier of the proposed super-heterodyne bio-radar system due to the fat that the phase noise power and the signal power are inreased by the same ratio, four times in this ase Noise Analysis In this paper, the SNR analysis is foused on the respiration signal, sine the respiration signal is generally the main issue of bio-radar system design. The reeived signal power in the I-hannel (S I ) of the onventional bio-radar system based on diret onversion arhiteture an be alulated as following [8]: ( ) 2P T x G T G R G Rx λ 2 σ h L h os 2 θ h + 4πx(t) S I = (4π) 3 d 4 0 P T xg T G R G Rx σ h L h x 2 (t) 2πd 4 (11) 0 The reeived signal power in the I-hannel (S I ) of the proposed bio-radar system based on the frequeny multiplier, quadrupler, an also be derived by using (11), and it is the same as that of the onventional system as shown in the following equation: S I P T xg T G R G Rx σ h L h x 2 (t) 2πd 4 (12) 0 To analyze the noise mehanism of the prosed bio-radar system, the residual phase noise of the LO in baseband is the most important parameter. In this paper, the power spetral density of the residual phase noise in baseband is alulated by using the autoorrelation of the residual phase noise in time-domain as well as Fourier transform of the alulated autoorrelation. The residual phase noise (R φ (τ)) and its power spetral density (S φ (f)) in the onventional bio-radar system are [1, 2]: where and R φ (τ) = E{[φ(t + τ τ d ) φ(t + τ)][φ(t τ d ) φ(t)]} = 2R φ (τ) R φ (τ τ d ) R φ (τ + τ d ) (13) S φ (f) = F{R φ (τ)} = S φ (f)(4 sin 2 πfτ d ) (14) S φ (f) = F{φ(t)} (15) τ d = 2d 0 λ (16)

6 154 Myoung et al. and it has to be notied that f in (14) and (15) is not the arrier frequeny but the offset frequeny. The autoorrelation (R φ (τ)) of the residual phase noise in the baseband of the proposed bio-radar system summarized in (8) and its power spetral density (S φ (τ)) an be alulated by using (8) by the same proedure with the onventional system as following: R φ (τ) = 26R φ(τ) 12R φ (τ τ d ) 12R φ (τ + τ d ) 4R φ (τ τ d τ m ) 4R φ (τ + τ d + τ m ) +3R φ (τ τ m ) 3R φ (τ + τ m ) (17) S φ (τ) = F{R φ (τ)} = S φ (f)(48 sin 2 πfτ d + 4 sin 2 πf(τ d + τ m ) +12 sin πfτ d sin πf(τ d + 2τ m )) (18) If τ m τ d, then the above equation an be further simplified as follows: S φ (f) 64 sin2 πfτ d (19) The analyzed residual phase noise spetral density in baseband of the proposed super-heterodyne bio-radar system in (19) is greatly redued from the LO phase noise (S φ (f)) with very small delay time (τ d and τ m ) in nano-seond order. It means that the range orrelation is still available and stable vital signal aquisition is possible with the proposed system. The main noise soures in the Doppler effet based bio-radar system are the thermal noise (N T ), residual phase noise due to antenna leakage (N φl1 ), lutter refletion (N φ ), human body baksattering (N φh ), mixer leakages through down-onversion mixer (N φl2 ), as well as IQ-demodulator (N φl3 ) [8]. From the analysis result in [8], the main noise soures an be summarized as followings: N φh = 2P T xg T G R G Rx (λ/4) 2 σ h L h φ h (t) 2 (4π) 3 d 4 0 (20) N φ = 2P T xg T G R G Rx (λ/4) 2 σ L φ h (t) 2 (4π) 3 d 4 0 (21) N T = G Rx kt B NF (22) The above noise power an be alulated by using (19) as following: N φh = P ( ) T xg T G R G Rx σ h L h πf 2 S φ (1)(1 Hz) 3 fh ln f L ( 1 2d 2 + ) 0 8d 3 τ m d 4 τm 2 (23) 0

7 Progress In Eletromagnetis Researh C, Vol. 19, N φ = P ( ) T xg T G R G Rx σ L πf 2 S φ (1)(1 Hz) 3 fh ln f L ( 1 2d 2 + ) 0 8d 3 τ m d 4 τm 2 (24) 0 The phase noise power from antenna leakage, whih is one of the most important and dominant parameters, an be easily alulated by using replaement of the delay of the signal bak-sattered by human body, τ d, into the delay from T x antenna to Rx antenna by diret oupling, τ a. The alulated residual phase noise power is as follows: ( ) N φl1 = 2π 2 η a P T x G Rx S φ (1)(1 Hz) 3 fh ln f L ( ( τa ) 2 τ a τ ( m τm ) ) 2 (25) 2 The reeived signal power of the proposed system in (12) is the same as that of the onventional bio-radar system with the arrier frequeny in (11), whih is the same as the fundamental frequeny of the proposed system based on multiplier, beause the relative hest variation to wavelength as well as propagation loss are inreased by the same ratio with frequeny multipliation. On the other side, the residual phase noise power of the proposed system in (25) is larger ompared with the onventional low frequeny diret onversion system in [8] by about 4 times. As a result, the the SNR of the proposed bio-radar system with the multiplied arrier frequeny is degraded by about 4 times. However, it has to be notied that the phase noise of the general VCO is proportional to the osillation frequeny, and the phase noise of the VCO osillating at four times higher frequeny is larger by 4 times or 12 db. It means that the proposed multiplier based bio-radar system using the arrier frequeny generated by frequeny multipliation from lower LO signal is the same in terms of SNR performane ompared with the onventional bio-radar system with the same arrier frequeny generated at higher frequeny. Moreover, the frequeny synthesis in lower frequeny is not only stable but also eonomial. The phase noise power due to the down-onversion mixer as well as IQ-demodulator an be alulated by the same manner. However, it is very interesting to note that the phase noise down-onverted by the down-onversion mixer does not affet the performane of the proposed super-heterodyne bio-radar system, beause the leakage signal is downonverted to DC and the down-onverted signal is filtered by IF filter and modulated again by IQ-demodulator. So, the phase noise due to down-onversion mixer (N φl2 ) does not need to be onsidered for

8 156 Myoung et al. this noise analysis. Moreover, the phase noise effet by leakage of the IQ-demodulator is exatly the same as the leakage effet of the mixer in the onventional diret-onversion based bio-radar system, beause there is no frequeny multipliation for the leakage signal of the IQdemodulator. As a result, the residual phase noise of the leakage signal of the IQ-demodulator (N φl3 ) an be expressed as following: ( ) (τiq N φl3 = 32π 2 η IQ S φ (1)(1 Hz) 3 fh ) 2 ln (26) 2 where, τ IQ is the delay of the leakage signal, and the other parameters are summarized in Table 1. Table 1. Simulation parameters, symbols, and values of the proposed bio-radar system. f L Parameter Symbol Value Fundamental frequeny of LO f 2.5 GHz LO phase noise at 1 Hz S φ(t) (1) +58 db/hz Carrier frequeny of LO - 10 GHz Filter low-utoff frequeny f L 0.5 Hz Filter high-utoff frequeny f H 30 Hz Reeiver gain G Rx 10 db T x Antenna gain G T 8 dbi Rx Antenna gain G R 8 dbi Output power P T x 0 dbm Heart RCS σ h 6.8e 3 Body lutter RCS σ 0.5 Heart refletivity L h 60 db Human body refletivity L 3 db Reeiver noise figure NF 6.0 db Antenna leakage η a 20 db Mixer RF-LO isolation η m 50 db IQ-demodulator isolation η IQ 50 db The idential residual phase noise of the leakage signal of the IQdemodulator with the onventional system with lower arrier frequeny means that the mixer leakage effet is negligible. Finally, the reeived SNR an be alulated as following: SNR = S I N T + N φh + N φ + N φl1 + N φl3 (27)

9 Progress In Eletromagnetis Researh C, Vol. 19, To ompare the system performane of the proposed bio-radar system with that of the onventional diret-onversion based bioradar system using the same arrier frequeny without any frequeny multipliation, the system simulation using MATLAB are performed. The phase noise of 10 GHz LO in the onventional system is assumed as +70 db/hz, whih is 12 db higher than that of the proposed bio-radar system with the 2.5 GHz frequeny synthesizer frequeny, sine the phase noise is proportional to VCO osillation frequeny. Exept for the fundamental frequenies and their phase noise harateristis, it is assumed that all the other parameters of the proposed and onventional systems are idential. The simulation parameters are utilized as summarized in Table 1, and they are based on the onventional and ommerial omponents. Note that the phase noise value of a freerunning VCO may be a positive dbc/hz value with a extremely small offset frequeny [8]. Figure 2 shows the simulated SNR due to the antenna leakage and lutter sattering, whih are the most dominant fators in the system performane, and the overall SNR of the proposed bio-radar system. As shown in Figure 2, the total SNR of the proposed bioradar system is high enough to measure the vital signal up to about 1 m. The alulated SNR of the proposed system is exatly same with that of the onventional diret onversion bio-radar system, and the alulated SNR traes of the onventional system are abbreviated to avoid repetition and onfusion. This simulation result means that the additional IF band signal proessing and benefiial low frequeny PLL implement an be employed without any system performane degradation, and the validity of the proposed bio-radar system based on a super-heterodyne arhiteture with PLL and frequeny multiplier is shown N φ (Clutter) N φ (Antenna leakeage) L1 Total SNR SNR [db] Distane [meter] Figure 2. The simulated SNR with respet of antenna leakage and lutter sattering in the onventional and proposed bio-radar system.

10 158 Myoung et al. 3. EXPERIMENTAL RESULTS Figure 3 shows the photographs of the fabriated bio-radar system. The designed bio-radar is realized in triple layered printed iruit boards (PCB). On the top board, T x as well as Rx array antennas are integrated. The seond board onsists of RF omponents inluding frequeny multiplier, RF bandpass filters, power divider, T x amplifier, and LNA. The bottom board hosts frequeny synthesizer with PLL, IQ demodulator, and IF filter. The size of the fabriated 10 GHz superheterodyne bio-radar system is mm 2. Figure 4 shows the measured return loss and isolation of the fabriated antenna. The measured return loss is about 30 db at 10.2 GHz frequeny band. Moreover, the isolation between T x-rx antennas at 10.2 GHz, best mathing frequeny of the antennas, is about 20 db. (a) (b) () Figure 3. The photographs of the fabriated bio-radar system (40 46 mm 2 ). (a) Antenna board (top board). (b) RF board (seond board). () Baseband board (bottom board). Return Loss [db] Return Loss Isolation Frequeny [GHz] Figure 4. The measured return loss and isolation of the fabriated antenna. Isolation [db] Phase Noise [db/hz] GHz 10 GHz K 10 K 100 K 1 M Offset Frequeny [Hz] Figure 5. Measured phase noise harateristis of 2.5 GHz LO and 10 GHz multiplied signals.

11 Progress In Eletromagnetis Researh C, Vol. 19, Figure 5 shows the measured phase noise harateristis of the 2.5 GHz LO and 10 GHz multiplied signals. The phase noise harateristis depited in Figure 5 show that the designed PLL is well loked, and the measured phase noise value of the 2.5 GHz LO is about 75 db/hz and 112 db/hz at 100 Hz and 1 MHz offset frequenies, respetively. For the 10 GHz multiplied signal, the measured values are 55 db/hz and 100 db/hz at 100 Hz and 1 MHz offset frequenies, respetively. The theoretial inrement of the phase noise due to the quadrupler is 12 db. This phase noise inrement is well mathed for the phase noise for the higher offset frequeny, but the phase noise of the 10 GHz multiplied signal for the lower offset frequeny is rather larger than theoretial predition. It is strongly believed that the main reason of higher phase noise inrement ompared with theoretial inrement is the noise added by the frequeny multiplier. However, the 10 GHz signal shows stable loking operation as well as phase noise harateristis. Figures 6(a) and 6(b) show the measured bio-signals at 50 m and 100 m, respetively. With the 10 GHz radar system the bio-signal measurements are performed in two different ways. First, the biosignal is measured for normally breathing ondition. In this ase, the measured bio-signal in I and Q hannels (I (Res.) and Q (Res.) in Figures 6(a) and 6(b)) inlude heartbeat signals. Next, the biosignal is measured on the same human body without respiration. In this ase, the measured bio-signal in I and Q hannels (I (Non-res.) Q (Non-res.) I (Non-res.) EOG Q (Res.) I (Res.) Time [se] (a) Time [se] Figure 6. Measurement measured bio-signals at 50 m and 100 m. (a) 50 m. (b) 100 m. Q (Non-res.) I (Non-res.) EOG Q (Res.) I (Res.) (b)

12 160 Myoung et al. and Q (Non-res.) in Figures 6(a) and 6(b)) inlude only heartbeat signal. Sine the heartbeat signal is relatively weak ompared to the respiration signal and the respiration signal onsists of many harmoni omponents lose to heartbeat signal frequeny region, the measured heartbeat signal without respiration may provide more objetive bioradar system performane measurement before applying any additional signal proessing. It should be noted that the heartbeat signals are measured by eletroardiogram (ECG) for referene signals to verify the proper operation of the proposed bio-radar system. It is lear that the respiration signals are learly observed in all of the measurement results. It is apparent that the proposed system works very well for lose-in subjet onsidering that there are large amount of T x to Rx oupling. The heartbeat signals measured at the distane of 50 m is also very lear in I hannel (I (Non-res.)) as well as Q hannel (Q (Non-res.)) as shown in Figure 6(a). Moreover, the number and position of the peaks of the heartbeat signal measured at the distane of 100 m also aords with the measured result by ECG, espeially in I hannel (I (Non-res.)) as shown in Figure 6(b). It is worth to note that in this work any additional signal proessing is not employed to proess these experimental results. 4. CONCLUSIONS This paper presents a new 10 GHz super-heterodyne bio-radar system based on frequeny multiplier and phase-loked loop. Sine the phase noise of loal osillator is still dominant fator due to the low isolation between T x and Rx antennas, the PLL employment an greatly improve the performane of Doppler-effet-based bio-radar system. Moreover, the PLL in this paper is realized at 2.5 GHz, and the 10 GHz signal is generated by the frequeny multiplier or quadrupler to ahieve best phase noise performane. The proposed bio-radar system has a super-heterodyne arhiteture, and it has distint advantages over onventional super-heterodyne arhiteture, suh as less stringent requirement of IQ demodulator and IF filtering for required noise figure performane. To verify the validity of the proposed system, the noise harateristis of the proposed system is arefully analyzed, and the result reveals that the same SNR performane are ahieved with the proposed system based on the frequeny multiplier ompared with the onventional system employing idential signal frequeny. This means that the additional IF band signal proessing as well as eonomial realization an be ahieved with the proposed system without any sarifie of the system performane. The experimental results demonstrate the exellent detetion performane of respiration

13 Progress In Eletromagnetis Researh C, Vol. 19, as well as heartbeat signals up to 100 m without any additional signal proessing tehnique. The presented bio-radar an be employed for future medial system, and the proposed super-heterodyne arhiteture based on PLL and frequeny multiplier an also be employed for higher frequeny bands, suh as 24 GHz or 60 GHz industrial, sientifi, and medial (ISM) radio band systems. REFERENCES 1. Budge, Jr., M. C. and M. P. Burt, Range orrelation effets in radars, Reord of the 1993 IEEE National Radar Conferene, , Apr Budge, Jr., M. C. and M. P. Burt, Range orrelation effets on phase and amplitude noise, Proeedings of IEEE Southeaston, 5, Droitour, A. D., O. Bori-Lubeke, V. M. Lubeke, J. Lin, and G. T. A. Kova, Range orrelation and I/Q performane benefits in single-hip silion Doppler radars for nonontat ardiopulmonary monitoring, IEEE Trans. Mirowave Theory and Teh., Vol. 52, No. 3, , Mar Droitour, A. D., V. M. Lubeke, J. Lin, and O. Bori-Lubeke, A mirowave radio for doppler radar sensing of vital signs, IEEE MTT-S International Mirowave Symposium Digest, Vol. 1, , May Xiao, Y., J. Lin, O. Bori-Lubeke, and M. Lubeke, Frequenytuning tehnique for remote detetion of heartbeat and respiration using low-power double-sideband transmission in the ka-band, IEEE Trans. Mirowave Theory and Teh., Vol. 54, No. 5, , May Kim, C. Y., J. G. Kim, and S. C. Hong, A quadrature radar topology with T x leakage aneller for 24-GHz radar appliations, IEEE Trans. Mirowave Theory and Teh., Vol. 55, , Jul Kim, C. Y., J. G. Kim, D. Baek, and S. C. Hong, A irularly polarized balaned radar front-end with a single antenna for 24- GHz radar appliations, IEEE Trans. Mirowave Theory and Teh., Vol. 57, , Feb Jang, B.-J., S.-H. Wi, J.-G. Yook, M.-Q. Lee, and K.- J. Lee, Wireless bio-radar sensor for heartbeat and respiration detetion, Progress In Eletromagneti Researh C, Vol. 5, , Myoung, S.-S., B.-J. Jang, J.-H. Park, and J.-G. Yook, 2.4 GHz

14 162 Myoung et al. bio-radar system with improved performane by using phaseloked loop, Mirowave and Optial Tehnology Letters, Vol. 52, No. 9, , Frequeny Alloations of Amateur Satellite Servie, International Teleommuniation Union, 2009.