Estimation and equalization of fading channels with random coefficients
Signal Processing - Special issue on higher order statistics
Adaptive tracking of linear time-variant systems by extended RLSalgorithms
IEEE Transactions on Signal Processing
Pilot symbol assisted modulation in frequency selective fadingwireless channels
IEEE Transactions on Signal Processing
Tracking time-varying correlated underwater acoustic channels in the signal subspace
Proceedings of the Eighth ACM International Conference on Underwater Networks and Systems
| Hi-index | 0.00 |
Many underwater acoustic channels exhibit correlated (semi-deterministic) multipath arrivals. Such channels are often time-varying with extensive multipath delay and yet have a limited number of degrees of freedom due to inter-path correlation. Traditional channel estimation algorithms do not exploit this correlation structure. To exploit cross-tap correlation, the channel impulse response is projected into a lower-dimensional signal subspace determined by the eigenvectors (bases) of the channel covariance matrix associated with the signal, represented by a set of uncorrelated channel components. Assuming constant bases, a model-based channel tracking algorithm is proposed, in which a priori knowledge of the characteristics of the channel, in the form of an autoregressive (AR) model for the uncorrelated channel components, is incorporated directly into the structure of the algorithm. The AR model determines the state transition matrix for a Kalman filter, which is used to further improve the tracking performance by compensating for the time variation intrinsically within the algorithm recursion. The proposed algorithm tracks only the small signal subspace and produces significant saving in computations. Performance is demonstrated with real data.