Efficient pilot allocation for sparse channel estimation in UWB OFDM systems

摘要

In ultra-wideband (UWB) orthogonal frequency-division multiplexing (OFDM) systems, compressive sensing (CS) is often employed to produce an estimate of the typically sparse channel from the observation of a limited number of allocated pilot subcarriers. Moreover, it is well established that proper pilot subcar-rier allocation (PA) design is the key for CS-based channel estimates to achieve a good performance. In this paper, we propose three novel PA algorithms. These are referred to as deterministic greedy algorithm (DGA), simplified deterministic greedy algorithm (SDGA) and simulated annealing (SA). The proposed algorithms are attractive alternatives for the state-of-the-art method because they are orders of magnitude less complex but yield only a small system performance degradation (SDGA), or they yield a better system performance as well as a lower complexity (DGA and SA).