Passive Identification of Multipaths Channels
Identification Passive de Canaux à Trajets Multiples
OPEN ACCESS
RF transmissions are often done along multipath channel, due to specular reflections. A physical model of propagating along such a channel is available, and takes into account few parameters as angles bearing of rays on the array, group delay for each path, Doppler shift, polarization. In order to compensate Rayleigh fading, a spatio-temporal separation of multipaths is proposed . Usually, this is done by transmitting a training sequence (known), which reduces the data rate. We show in this paper that passive identification can be performed, using only received signals. Proposed algorithm proceeds in two steps : the first step is a blind deconvolution, and then a parametric estimation of the channel is performed .Many simulations exhibit performances of proposed algorithms.
Résumé
Les transmissions radioélectriques sont souvent effectuées sur des canaux à trajets multiples, dus essentiellement à des réflexions sur desobstacles physiques présents dans l'environnement. Une modélisation physique des trajetsde propagation a conduit à une paramétrisation de la réponse impulsionnelle du canal, utilisant les angles d'incidence des rayons sur la station réceptrice, ainsi que les temps de retard degroupe dechaque trajet, le décalage Doppler et la polarisation. Afin de pallier les évanouissements fréquentiels, une séparation spatio-temporelle des trajets est proposée. Habituellement, une telle séparation est réaliséeparl'envoi périodique d'une séquencetestd'égalisation (connue), qui a poureffetderéduire le débitde la transmission. Nous montrons dans cetarticleque l'identification peut être menéedemanière passive,uniquement à partir des signaux reçus. L'algorithme proposé procèdeendeux étapes: la premièreétape effectue une déconvolution autodidacte, puis laseconde procède à uneestimation paramétriquedu canal. De nombreusessimulationsmontrent les performancesdes algorithmes proposés.
Space time processing, multi path separation, blind deconvolution, parametric estimation, maximum likelihood.
Mots clés
Identification spatio-temporelle, séparation de multi-trajets, déconvolution autodidacte, estimation paramétrique, maximum devraisemblance.
[1] K. Abed-Meraim, E.Moulineset P. Loubaton «Prediction Error Method for , Second-order Blind Identification>>, IEEE Transactions on Signal Processing, vol45, No. 3, pp. 694-704, Mars 1997.
[2] S. Anderson, M. Millnert, M. Viberg and B. Wahlberg, «An Adaptive Array for Mobile Communication System» IEEE Transactions on Vehicular Technology,Vol. 40, No. 1, pp. 230-236, February 1991.
[3] J.C.Bic, D. Isnier, M. Jay et P. Merton,«Les Spécificités de la Propagation en ZoneUrbaine», REE,No. 7, pp. 11-24,Juillet 1996.
[4] G. Bienvenu et L. Kopp, «Principe de la gionométrie passive adaptative», Actesdu GRETSI, pp. 106/1-10, mai 1979.
[5] G.E.P. Box and G.M. Jenkins, «Time Series Analysis : Forecasting and Control», San Fransisco, CA: Holden-Day, 1970.
[6] H. Clergeot, S. Tressens, A. Ouamri, «A new maximum likehood method for estimation of correlated sources. Comparison with existing methods», Proceedings of Eusipco, Florence, pp 71-75, Sept. 1987.
[7] H. Clergeot, S. Tressens, A. Ouamri, «Performances of High Resolution Spectral Methods to theCramerRao Bounds», IEEE Transactions on Acoustics Speech and Signal Processing,Vol.37, N'5, pp. 1701-1720, November 1989.
[8] J.F. Denisse et J.L. Delcroix,Théorie des Ondes dans les Plasmas, Éditions Dunod, 1961.
[9] U. Dersch and E. Zollinger, «Physical Characteristics of Urban Micro-Cellular Propagation», IEEE Transactions on Antennas and Propagation, Vol.42, No. 11, pp. 1528-1539, November 1994.
[10] G. H. Gollub and C. F. Van Loan, Matrix Computation, Baltimore, John Hopkins University Press, 1990.
[11] P. Gounon, «Analyse spatio-temporelle haute résolution à l'aide d'une antenneactive»,RevueTraitementdu signal, vol11,n°5, pp.351-360,1994.
[12] P. Gounon and S. Bozinoski, «High Resolution Spatio-temporal Analysis by an Active Array»,Proceedings of ICASSP, Vol. 5, pp. 3575-3578, 1995 .
[13] J. Grouffaud, «Identification spatio-temporelle decanaux de propagation à trjets multiples», Thèse de doctorat de l'ENS de Cachan, 16 Juin 1997.
[14] Y. Hua, «Fast Maximum LikehoodforBlind Identification ofMultipleFIR Channels»,IEEE Transactions on Signal Processing, vol.44, N°44, pp. 661672, March 1996.
[15] T. Lo and J. Litva, «Use of a Highly Deterministic Multipath Signal Model in Low-Angle Tracking» , IEE Proceedings, Vol. 138, No. 2, pp. 163-171, April 1991.
[16] O. Macchi, C.A. FariaDa Rocha, J.M. Travassos-Romano, « Egalisation Adaptative Autodidacte par Rétroprédiction et Prédiction», Proceedings of GRETSI, pp. 491-494, Sept. 1993.
[17] T. Manabe and Y. Furuhama, «Recent Propagation Studies in Japan», IEEE Antennas and Propagation Magazine,Vol. 36, No. 5, pp. 7-13, October 1994.
[18] E. Moulines, P. Duhamel, J.F. Cardoso and S. Mayrargue, «Subspace Methods for the Blind Identification of Multichannel FIR Filters», IEEE Transactions on Signal Processing, Vol. 43, No. 2, February 1995.
[19] M.A. Panas and G. Jourdain, «Active High Resolution Time Delay Estimation for Large BT Signals », IEEE Transactions on Signal Processing, Vol. 39, No. 4, pp. 781-787, April 1991.
[20] A. J. Paulraj and C. B. Papadias, «Space-Time Processing for wireless communications>>, IEEE Signal Processing Magazine, pp.49-83, November 1997.
[21] C. Rush, « Ionospheric Radio Propagation Models and Prediction . A minireview», IEEE Transactions on Antennas and Propagation, Vol. 34, No. 4, pp. 1163-1170, April 1986.
[22] R.O. Schmidt, «A Signal Subspace Approach to Multiple Emitter Location and Spectral Estimation», Ph D Thesis, Stanford University, CA, 1981.
[23] P. Stoica, K.C. Charman, «Maximum likelihood Methods for Direction of Arrival Estimation», IEEE Transactions on Acoustics Speech and Signal Processing,Vol. 38, N°7, pp. 1132-1142, July 1990 .
[24] L. Tong, G. Xu and T. Kailath, «A New Approach to Blind Identification and Equalization of Multipath Channels», Proceedings of the ASILOMAR Conference, Pacific Grove, CA, pp. 856-860, 1991.
[25] L. Tong, G.Xuand T. Kailath, «Blind Identification and Equalization ofMultipath Channels», Proceedings oflnternational Conference Communications, pp. 1513-1517, June 1992.
[26] M. Wax, T. Kailath, «Detection of Signals by Information Theoretic Criteria», IEEE Transactions on Acoustics Speech and Signal Processing, Vol. 33, N°2, pp. 387-392, April 1985.
[27] G. Xu and T. Kailath, «Fast Subspace Decomposition», IEEETransactions on Signal Processing, Vo1.42, No. 3, pp 539-551, March 1994.
[28] G. Xu, H. Liu, L. Tong and T. Kailath, «A Least Square Approach to Blind ChannelIdentification>>, IEEETransactions on Signal Processing, Vol. 43, No. 12, pp. 2982-2993, December 1995.
[29] G. Xu, Y. Cho and T. Kailath, «Application of Fast Subspace Decomposition to Signal Processing and Communication Problems»,IEEE Transactions on Signal Processing, Vol.42, No. 6, pp. 1453-1461, June 1994.