Performance Comparison between 6-Port and 5-Port homodyne circuits for DOA Estimation

Similar documents
Theory and Proposed Method for Determining Large Signal Return Loss or Hot S22 for Power Amplifiers Using Phase Information

90 and 180 Phase Shifter Using an Arbitrary Phase-Difference Coupled-line Structure

Lab 12. Speed Control of a D.C. motor. Controller Design

3G Evolution. OFDM Transmission. Outline. Chapter: Subcarriers in Time Domain. Outline

CH 7. Synchronization Techniques for OFDM Systems

Time of Arrival Estimation for WLAN Indoor Positioning Systems using Matrix Pencil Super Resolution Algorithm

Comparison of Conventional Subspace-Based DOA Estimation Algorithms With Those Employing Property-Restoral Techniques: Simulation and Measurements

DETERMINATION OF ELECTRONIC DISTANCE MEASUREMENT ZERO ERROR USING KALMAN FILTER

RECOMMENDATION ITU-R M.1828

Introduction to Digital Signal Processing

Introduction to Medical Imaging. Signal Processing Basics. Strange Effects. Ever tried to reduce the size of an image and you got this?

Package: H: TO-252 P: TO-220 S: TO-263. Output Voltage : Blank = Adj 12 = 1.2V 15 = 1.5V 18 = 1.8V 25 = 2.5V 33 = 3.3V 50 = 5.0V 3.3V/3A.

Linearization of Two-way Doherty Amplifier by Using Second and Fourth Order Nonlinear Signals

Low Cross-Polarization Slab Waveguide Filter for Narrow-Wall Slotted Waveguide Array Antenna with High Gain Horn

PERFORMANCE ANALYSIS OF DIGITAL BEAMFORMING ALGORITHMS

Available online at ScienceDirect. International Conference On DESIGN AND MANUFACTURING, IConDM 2013

Analysis the Performance of Coded WSK-DWDM Transmission System

Online Publication Date: 15 th Jun, 2012 Publisher: Asian Economic and Social Society. Computer Simulation to Generate Gaussian Pulses for UWB Systems

Investigation of Power Factor Behavior in AC Railway System Based on Special Traction Transformers

The Trouton Rankine Experiment and the End of the FitzGerald Contraction

Analysis and Practical Comparison of Wireless LAN and Ultra-Wideband Technologies for Advanced Localization

Coexistence between WiMAX and Existing FWA Systems in the Band 3500 MHz

Enhancing the Performance of Ultra-Tight Integration of GPS/PL/INS: A Federated Filter Approach

IEEE Broadband Wireless Access Working Group <

On parameters determination of multi-port equivalent scheme for multi-winding traction transformers

A simple automatic classifier of PSK and FSK signals using characteristic cyclic spectrum

ANALYSIS ON THE COVERAGE CHARACTERISTICS OF GLONASS CONSTELLATION

TALLINN UNIVERSITY OF TECHNOLOGY. IRO0140 Advanced Space Time-Frequency Signal Processing. Individual Work

Performance Analysis of BLDC Motor for Sinusoidal and Trapezoidal Back-Emf using MATLAB/SIMULINK Environment

PAPR REDUCTION TECHNIQUES IN OFDM SYSTEMS USING DCT AND IDCT

Logic Design 2013/9/26. Outline. Implementation Technology. Transistor as a Switch. Transistor as a Switch. Transistor as a Switch

A Synthesis-free Directional Modulation Transmitter using Retrodirective Array

WIDEBAND SPECTRUM SENSING FOR COGNITIVE RADIO

Signals and Systems Fourier Series Representation of Periodic Signals

Direction Finding and Positioning Algorithm with COLD- ULA Based on Quaternion Theory

Frequency Estimation of Unbalanced Three-Phase Power Systems Using the Modified Adaptive Filtering

1.1 Transmission line basic concepts: Introduction to narrow-band matching networks

Chalmers Publication Library

Matched filter based algorithm for blind recognition of OFDM systems

DPCCH Gating Gain for Voice over IP on HSUPA

Impact Analysis of Damping Resistors in Damped Type Double Tuned Filter on Network Harmonic Impedance

Common Collector & Common Base Amplifier Circuits

Test Results of a Digital Beamforming GPS Receiver in a Jamming Environment Alison Brown and Neil Gerein, NAVSYS Corporation

EECE 301 Signals & Systems Prof. Mark Fowler

η = ; (3) QUANTITATIVE INTERPRETATION OF PRECIPITATION RADAR 7R.3 MEASUREMENTS AT VHF BAND Edwin F. Campos 1*, Frédéric Fabry 1, and Wayne Hocking 2

Conducted EMI of Switching Frequency Modulated Boost Converter

Performance Analysis and Architecture Design of Vector-Based Ultra- Tightly Coupled GPS/INS Integration on satellite Faults

IONOSPHERIC ABSORPTION OF HF RADIO WAVE IN VERTICAL PROPAGATION *

Semi Blind Channel Estimation with Training-Based Pilot in AF Two- Way Relaying Networks

Determination of Antenna Q from the Reflection-Coefficient Data

Inverter fault Analysis in Permanent Magnet Synchronous Motor using Matlab & Simulink

Chapter 2 Fundamentals of OFDM

Efficient loop-back testing of on-chip ADCs and DACs

cos The points in an Argand diagram which represent the numbers (iii) Write down a polynomial equation of degree 5 which is satisfied by w.

Integrated INS/GPS Navigation System

LNA IN GND GND GND GND IF OUT+ IF OUT- 7. Product Description. Ordering Information. GaAs HBT GaAs MESFET InGaP HBT

Performance of Extended Super-Orthogonal Space -Time Trellis Coded OFDM system

Using SigLab for Production Line Audio Test

RADAR SIGNAL FEATURE EXTRACTION SYSTEM BASED ON A SAW DISPERSIVE DELAY LINES

ESX10-10x-DC24V-16A-E electronic circuit protector

Narrow-wall slotted waveguide array antenna with low cross-polarization filter

PAPR REDUCTION OF OFDM SIGNALS USING SELECTIVE MAPPING AND CLIPPING HYBRID SCHEME. Victor Cuteanu, Alexandru Isar

Fuzzy Anti-Windup Schemes for PID Controllers

HSMS-2823 RF mixer/detector diode

16 th Coherent Laser Radar Conference (June 20, 2011, Long Beach CA, USA)

A Pilot Aided Averaging Channel Estimator for DVB-T2

QUAD PRECISION MICROPOWER CMOS VOLTAGE COMPARATOR WITH DRIVER

A Self-tuning Fuzzy PID Control Method of Grate Cooler Pressure Based on Kalman Filter

A DSP-based Discrete Space Vector Modulation Direct Torque Control of Sensorless Induction Machines

Bi-Directional N-Channel 20-V (D-S) MOSFET

Grid Impedance Estimation for Islanding Detection and Adaptive Control of Converters

Real Time Speed Control of a DC Motor Based on its Integer and Non-Integer Models Using PWM Signal

Study of Distribution of Transient Voltages in the Winding of a Transformer Subjected to VFTO and Lightning Surges

J. Electrical Systems 9-3 (2013): Regular paper

Analog Integrations Corporation 4F, 9 Industry E. 9th Rd, Science-Based Industrial Park, Hsinchu, Taiwan DS-385B

Lecture 19: Common Emitter Amplifier with Emitter Degeneration.

AOZ8904 Ultra-Low Capacitance TVS Diode Array

A SIMULATION MODEL FOR LIGHT RAIL TRANSPORTATION SYSTEM

ETSI TS V1.2.1 ( )

Dynamic Walking of Biped Robots with Obstacles Using Predictive Controller

SGM8521/2/4 150kHz, 4.7µA, Rail-to-Rail I/O CMOS Operational Amplifiers

Adaptive Design of OFDM Radar Signal With Improved Wideband Ambiguity Function

SPX mA Low Drop Out Voltage Regulator with Shutdown FEATURES Output 3.3V, 5.0V, at 400mA Output Very Low Quiescent Current Low Dropout Voltage

Uplink CoMP under a Constrained Backhaul and Imperfect Channel Knowledge

Multipath Rejection Through Spatial Processing

Advanced I/Q Signal Processing for Communication Systems

Pacing Guide for Kindergarten Version GLE Checks for Understanding Vocabulary Envision Textbook Materials

Department of Humanities & Religious Studies Assessment Plan (REV 6/16)

A Fast and Safe Industrial WLAN Communication

symbols s 1 and s 0 are transmitted simultanously from the transmitters at time instant t. At time instant t + T,the

QUAD PRECISION CMOS VOLTAGE COMPARATOR WITH PUSH-PULL DRIVER

SGM8621/2/3/4 250µA, 3MHz, Rail-to-Rail I/O CMOS Operational Amplifiers

US6H23 / IMH23 V CEO 20V V EBO 12V. 600mA R k. Datasheet. Outline Parameter Tr1 and Tr2 TUMT6 SMT6

SGM8631/2/3/4 470µA, 6MHz, Rail-to-Rail I/O CMOS Operational Amplifiers

SGM721/2/3/4 970µA, 10MHz, Rail-to-Rail I/O CMOS Operational Amplifiers

4.5 COLLEGE ALGEBRA 11/5/2015. Property of Logarithms. Solution. If x > 0, y > 0, a > 0, and a 1, then. x = y if and only if log a x = log a y.

3A High Current, Low Dropout Voltage Regulator Adjustable, Fast Response Time

EMA5 / UMA5N / FMA5A. V CC -50V -100mA 2.2kW 47kW I C(MAX.) R 1 R 2. Datasheet

Polyphase Modulation Using a FPGA for High-Speed Applications

Channel Tracking in Wireless OFDM Systems

Transcription:

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 85 Prformanc Comparison btwn 6-Port and 5-Port homodyn circuits for DOA Estimation Ayman Abdallah, Soubhi Abou Chahin, Kamal Sabbah, ouhammd Rammal, Guillaumm Nvux, Patrick Vaudon Abstract In this papr, w propos two approachs to provid dirction of arrival using homodyn dtction systms tchnologis lik six-port and fiv port rflctomtrs. Th choic of th rflctomtrs' typ bcoms critical in paralll of th circuit capability, i.. in trms of th numbr of signals to b dtctd simultanously. W prsnt our study by implmnting USIC algorithm at th output of two diffrnt homodyn rcivrs. Th rsults ar dducd from th simulation of th charactristics of ach systm, using th Advancd Dsign Systm (ADS) and ATAB. 1. Introduction Th problm of localization of sourcs radiating nrgy by obsrving thir signal rcivd at spatially sparatd snsors is of considrabl importanc, occurring in many filds, including radar, sonar, mobil communications, radio astronomy, and sismology.[1] Thus a varity of mthods, using htrodyn or homodyn tchniqus, for th Dtction Of Arrivals (DOA) stimation ar usd including spctral stimation, minimum-varianc distortionlss rspons stimator, linar prdiction, maximum ntropy, and maximum liklihood [2]. In addition to prvious mthods th most famous mthods usd in DOA ar ignstructur mthods, including many vrsions of USIC algorithms [3], minimum norm mthod [4], ESPRIT mthod [5], and th wightd subspac fitting mthod. Whil th most rcntly publishd paprs dscrib th DOA for thr signals at most [6], [11], th main goal of this papr is to implmnt USIC algorithm with two typs of homodyn dtction circuits. W us ths typs of circuits bcaus of thir simplicity, low cost and high fficincy in th dtction of arrivals. Th purpos is to compar th quality of th dtctd angls, in trm of prcision, for both circuits, spcially whn th numbr of th incoming RF signals incrass till svn. This papr consists of six main sctions. Sction 2 prsnts a gnral introduction of smart antnnas. Sction 3 dscribs 5-port and 6-port homodyn rcivrs and thir calibration tchniqus. Sction 4 xplains th thory of th USIC algorithm. Th simulation rsults as function of local oscillator powr, for 6- port and 5-port circuits, ar prsntd in sction 5. Finally th main conclusions ar summarizd in sction 6. 2. Smart antnnas dsign systms Smart antnnas ar ssntial parts of th nw gnration mobil systms. Thy can b usd to xploit spatial and spctral charactristics of th incoming signals to provid highly accurat location information [1]. Th dirction of arrival (DOA) algorithm is on of th most important parts of such tchniqus. A smart antnna systm can contain an antnna array that rcivs th incoming signals to dtrmin thir arrival angls and to control th transmittd signals. RF 2 RF 1 Array A2 Array A1 Phas Shiftr Control Bus R 1 R n Figur 1: Smart antnnas systm. Figur 1 rprsnts a smart antnna systm which can dtct th DOA for two RF signals simultanously. To do so, an antnna array A 1, that has thr patchs, rcivs th incoming wavs from th signals RF 1 and RF 2. Th output of ach patch is dirctly connctd to a rflctomtr (R i ). From th linarizd and calibratd outputs of th rflctomtrs, thr or four dpnding on his typ; fiv or six ports rspctivly, w can obtain thr low frquncy complx nvlops (ratio of incoming signal to a rfrnc signal). Ths ar th inputs for th USIC algorithm. This algorithm provid as an n D S P Transmission Rcption

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 86 Eignvalus output a matrix that is th rsult of projcting th b2 a1 b1 a5 a2 b5 b3 a3 b4 a4 V3 V4 V5 Figur3. Fiv-port rflctomtr In cas of six-port, a couplr is rquird to provid a rfrnc powr from th ocal Oscillator (O) input for normalization as shown in figur4. P 3 P 4 P 5 signal plan ovr th nois plan as shown in figur2. Figur 2: DOA of 2 signals using ADS Figur 2 rprsnts th rsult of simulation for th DOA of two incoming RF signals from a -18.1 0 and 7.3 0. Th simulation is mad using ADS (in a noislss nvironmnt) to dsign th thr rflctomtrs in addition to circuits for simulating th incoming signals [7]. As a conclusion, to dtct N signals N+1 patchs and rflctomtrs in addition to biggr capability of th DSP programs. 3. Homodyn rcivrs Th architcturs of th homodyn rcivrs ar simpl, with low nrgy consumption and small numbr of componnts. Actually, thr ar two typs of homodyn rcivrs that ar usd as ntwork analyzrs which ar fiv-port rflctomtrs and sixport rflctomtrs. Ths rflctomtrs prcisly dtrmin th complx ratio btwn two input signals. 3.1 Six-port and fiv-port rflctomtrs 50Ω a 1 Dirctiv couplr D 3 Figur 4: rflctomtr six-ports Ths dvics prmit th masurmnt of th rflction cofficint Γ of a dvic undr tst or th ratio of two RF input signals (complx nvlop). Th output dtctd powrs ar submittd to linarization and calibration functions to gt Γ or th ral part I(t) and imaginary part Q(t) of th complx nvlop. 3.3 Dsigning powr dtctors for th homodyn rcivrs Th powr dtctor HSS2852 is mad as shown in figur 5. HF C.a 1 T.a 1 D 4 Ring 5 accss D 5 D.U.T a 2 D 6 P 6 50Ω C R b 2 DC Th six-port and fiv-port rflctomtrs ar microwav masurmnt dvics that wr first introducd by G. F. Engn [8]. Th fiv-port rflctomtr consists of a fiv-port microstrip ring with thr Schottky powr dtctors as shown in figur 3. Figur 5: Powr dtctor It is locatd at ach of th output of th homodyn rcivrs. It convrts RF signals to ow frquncy signals or to DC signals. Th rlation btwn input powr and th output voltag of this dtctor is prsntd in figur 6.

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 87 inarizd Outputs Ral Outputs Figur6: inarization of powr dtctors For low input powr, th law of opration of diod dtctor is quadratic, th output voltag is linarly dpndant of input powr. Howvr, for highr input powr (-15dBm), th output voltag is proportionnal to th amplitud of input signal (nvlopp dtctor), so a linarization function is rquird [10], [11]. powrs to th local oscillator and th tst masurmnt port. Th thortical valu of th rflction cofficint is 0.25 (-12dB) sinc th coupling cofficint of th couplr is -8dB. A good agrmnt is obtaind with th simulatd rsults and as consqunc it can b usd with high fficincy for our systms. Two mthods ar usd to calibrat th fiv port rflctomtr [9], [14]. Th first on uss th fiv port as a I(t), Q(t) dmodulator. This mthod dtcts th complx nvlop of two RF signals at th input of th fiv-port systm. This nvlop will b srvd as input for USIC algorithm to gt th DOA as mntiond bfor. Whil th scond mthod prmits to obtain th rflction cofficint. As in cas of sixport, th thortical magnitud is 1 which agrs with th rsult of th application of a mobil short circuit, as shown in figur 8. Im(Γ) 3.4 Calibration of th homodyn rcivrs any calibrating mthods wr proposd for th six port rflctomtr. Som of thm dtrmin th constants of th rflctomtr by using crtain known loads with matrix calculations. An accurat tchniqu is th six port to four port rduction mthods, providd by Engn. Th main goal of th rduction is to provid th complx ratio btwn th incidnt wavs of two of th four dtctors as function of th four masurd powrs. In this papr w us this mthod providd by Widmann [10]. Figur7 prsnts th calibration mthod. R(Γ) Figur8: Calibration of a fiv port rflctomtr 4. USIC algorithm q 5 q 4 Th multipl signal classification (USIC) [3] mthod is a rlativly simpl and fficint ignstructur that uss subspac dcomposition mthod for stimation of th frquncis of complx sinusoids obsrvd in additiv whit nois [12]. For m rciving antnnas, a signal y(m) can modld as: y P ( ) ( m) A j 2π F K m + φk = n( m)(1) K = 1 K + Figur7: Calibration of a six port rflctomtr Figur7 prsnts th rsult of th simulation of a mobil short circuit. This mobil short circuit is obtaind by applying 0dBm and -20dBm as input q 6 Whr A k, F k, Φ k, ar rspctivly th amplitud, th phas and th frquncy of ach of th P incoming sourc. An N-sampl vctor y=[y(m),..., y(m+n 1)] of th noisy signal can b writtn as: y = x + n = Sa + n (2)

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 88 Whr th signal vctor x=sa is dfind as X X X ( m) ( m + 1) = A1 A2 AP ( ) m + N 1 πf1 N 1) j 2πF2 N 1) πfp N 1) j 2π φp πf1 m πf1 1) πf2m j 2πF2 1) πfp m πfp 1) (3) Th matrix S and th vctor a ar dfind on th right-hand sid of Equation (3). Th autocorrlation matrix of th noisy signal y can b writtn as th sum of th autocorrlation matrics of th signal x and th nois as follows: π φ1 π φ2 In this sction w will rprsnt th rsults of smart antnnas systms outputs whn using six port rflctomtrs and fiv port rflctomtrs with USIC algorithm and a local oscillator powr of 0dBm. All simulations ar mad in a quasi-idal nvironmnt whr w th problm of intrfrncs [13] is not considrd. Th following figur rprsnts th DOA of two incoming RF signals. R yy = R xx = SPS + R H nn 2 + σ I n With R xx =SPS H and R nn = σ n 2 I ar th autocorrlation matrics of th signal and nois procsss, th xponnt H dnots th Hrmitian transpos, and th diagonal matrix P dfins th powr of th sinusoids as: Whr [ P P P ] H P = aa = diag, 1 2,..., 2 i A i P = is th powr of th complx nvlop. Th st of ignvctors associatd with th P largst ignvalus span th signal subspac and ar calld th principal ignvctors. Th signal vctors si can b xprssd as linar combinations of th principal ignvctors. Th scond subst of ignvctors spans th nois subspac and has σ 2 n as thir ignvalus. Sinc th signal and nois ignvctors ar orthogonal, it follows that th signal subspac and th nois subspac ar orthogonal. Hnc th sinusoidal signal vctors si which ar in th signal subspac, ar orthogonal to th nois subspac, and w hav: P Figur10: DOA of two RF signals Th abov figur shows th USIC rspons of two incoming signals to two smart antnna systms that on of thm uss six port rflctomtrs and th scond on uss fiv port rflctomtrs. Th thortical angls was 7.3 0 and -19.5 0, as shown in th xprimntal rsults both smart antnna systms gav qual and prfct rsults. Th sam xprimnt was mad to dtct till 7 signals, th rsults wr: G Six-Port B Fiv-Port Figur 9: Dcomposition of th ignvalus of a noisy signal into th principal ignvalus and th nois ignvalus. Figur9 concluds that th application of th USIC algorithm can lad to sparat th incoming signals from ach othr as wll as from th whit nois, so indd a prcis dtction of arrival angls. 5. Simulation rsults 5.1 Stabl ocal Oscillator Powr cas Arrival Angl Dgr Figur11: DOA of svn RF signals Again th rsults wr qual and prcis for both systms. So, w can us any systm sinc th input O powr is stabl. 5.2 Unstabl ocal Oscillator Powr cas In this sction, w will show th rsults of simulations for DOA of 7 signals using six-port and fiv-port rflctomtr rcivrs in an unstabl powr nvironmnt for local oscillator.

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 89 5.2.2 Fiv-port rcivrs cas 5.2.1 Six-port cas: To do so, w simulatd th six-port rflctomtrs by applying 0dBm, -3dBm, -6dBm as local oscillator powr. Whil th calibration was don at 0dBm O powr, th rsults ar rprsntd in th figur blow. In this sction w followd th sam stps as in 5.2.1 but instad of utilizing six-port rcivrs w usd fiv-port rcivrs. 0dBm -6dBm -3dBm G P_O 0dBm B P_O -3dBm R P_O -6dBm Eignvalus Arrival Angl Dgr Figur14: DOA of 2 RF signals using fiv-port rcivrs Figur12: DOA of svn RF signals for 3 six-port rcivrs Th abov figur rprsnts th rsult for DOA of svn signals for th 3 cass. To dtct th ffct of th local oscillator powr (P_O), w can mak a zoom on aras in th figur13 as shown blow. G P_O 0dBm B P_O -3dBm R P_O -6dBm Th abov pictur rprsnts th DOA for two incoming RF signals which coms from -22.9 o and 33.8 o. As dscribd abov th fiv-port rflctomtrs ar calibratd on a 0dBm as a local oscillator powr. W can asily conclud that th systm prform wll if th O powr is 0dBm as in calibration. Th angl of arrivals shows larg rrors if th O powr falls down to -3dBm and -6dBm. Th rspctiv arrival angls ar -1.5 o and 15.7 o for th first cas, whil thy ar -0.4 o and 14.7 o for -6dBm. It is to b notd that th simulation is limitd for two signals du this dficincy which shows clarly th diffrnc btwn both systms. 6. Conclusions Figur13: comparison of th DOA of a 7.3 0 RF signals for six-port rcivrs This figur shows th rspons of an incoming RF signal form a 7.3 0. W may say that th fficincy DOA for th six-port rcivrs did not affctd badly by local oscillator powr variations (Powr instability). W may conclud from th abov rsults that ths kinds of rcivrs rsists to th P_O instability. This papr has shown th comparison prformanc of 5-port and 6-port homodyn rcivrs as applid to DOA in smart antnna rcivrs. A brif summary of structur and calibration of both systms ar dscribd. Th USIC algorithm was usd to stimat th dirction of arrival. W hav dmonstratd that svn signal dirctions can b obtaind using both systms. Th simulations wr mad as function of th local oscillator powr stability. W concludd that in a stabl nvironmnt, both rcivrs workd proprly, but in an unstabl nvironmnt only th six-port rcivrs consrvd good prformanc. 7. Rfrncs [1] Godara,.C., Application to antnna arrays to mobil communications. Part II: Bamforming and dirction of arrival considrations, IEEE Proc., 85, 1195 1247, 1997. [2] al Chand Godora, "Smart Antnnas", CRC PRESS C 2004.

Procdings of th 4th WSEAS Int. Confrnc on Elctromagntics, Wirlss and Optical Communications, Vnic, Italy, Novmbr 20-22, 2006 90 [3] R. O. Schmidt, ultipl mittr location and signal paramtr stimation, IEEE Trans. Antnna and Propag., vol. 34,pp. 276-280, 1986. [4] Buckly, K.. and Xu, X.,., Spatial spctrum stimation in a location sctor, IEEE Trans. Acoust. Spch Signal Procss., 38, 1842 1852, 1990. [5] Yun, N. and Fridlandr, B., Asymptotic prformanc analysis of ESPRIT, highr ordr ESPRIT, and virtual ESPRIT algorithms, IEEE Trans. Signal Procss., 44, 2537 2550, 1996. [6] Van Ym Vu, A.Judson Braga, Xavir Bgaud and Brnard Huyart, "ulti-path dlay masurmnt basd on fiv-port discriminator", Europan Confrnc on Wirlss Tchnology, Amstrdam 2004 [7] Amant Garcia, Concption d un radar d aid à la conduit automobil utilisant un systèm discriminatur d fréqunc typ six-port, Thès ENST, 2003 [8] G.F. ENGEN, "Th six-port rflctomtr: An altrnativ ntwork analyzr" IEEE Trans. icrowav Thory Tzch., TT-25, p 1075-1080, dcmbr 1977. [9] G.Nvux. dmodulatur dirct d signaux RF multimod t multi-band utilisant la tchniqu cinq-port, ths ENST Paris 2003. [10] Frank Widmann, Dévloppmnts pour ds applications grands publiqus du réflctomtr six-ports: algorithm d calibrag robust, réflctomtr à très lrg band t réflctomtr intgr IC. Thsis ENST Paris 1997. [11] T. ack1, T. Eirinr1, T. üllr1, J.-F. uy1, A Digital mm-wav Smart Antnna Rcivr basd on Six- Port Tchnology for Nar Rang Radar Applications, 34th Europan icrowav Confrnc - Amstrdam, 2004 [12] Ildar Urazghildiiv, USIC with Outlir Rjction: A Nw Tool for Improving th Estimation Prformanc at ththrshold Rgion of th SNR,Europan Radar Confrnc, Amstrdam 2004 [13] Amin,.G., Concurrnt nulling and locations of multipl intrfrncs in adaptiv antnna arrays, IEEE strans. Signal Procss., 40, 2658 2668, 1992. [14] F. R. D SOUSA, B. A. Garcia, G. NEVEUX, B. HUYART, «Fiv-Port Junction : In th way of Gnral Public Application», Europan icrowav Confrnc EUC 2002, ilan, Sptmbr 2002