Computationally Efficient Linear Predictor for Communication Over Doubly Selective Fast-Fading Channel

This paper proposes multidimensional diversity for the high-velocity channel while enabling pertinent tradeoffs among error performance and complexity. It also analyzes and compares the impact of transmit diversity of information transmission over wireless communication system. The paper considers a high-velocity outdoor system like a vehicular user for which the channel varies too rapidly to be tracked. A blind channel-estimation approach that incorporates a linear prediction and MMSE is proposed, which reduces the effect of any perturbation introduced by linear prediction, noise, and multipath. The comparison results reveal that the proposed method using recursive modified Cholesky (RChol) algorithm is computationally efficient when compared with the existing linear prediction based methods. This algorithm scales down the correlation matrix time averaging and hence unveils a significantly computationally efficient algorithm for the time domain equalizer.