TB GOMP channel estimation algorithm for shallow underwater acoustic communication

DOI:	10.11809/bqzbgcxb2023.05.032
Keywords:	TB GOMP; OFDM; channel estimation; shallow underwater acoustic channel
Abstract:	Aiming at the problems of generalized orthogonal matching pursuit (GOMP) algorithm, including high complexity, large estimation error, an excessive number of required pilots, over dependence of estimation performance on atomic selection numbers and insufficient consideration of noise, this paper proposes a GOMP algorithm based on atomic threshold and backtracking (TB GOMP). The proposed algorithm is also applied to the compressed sensing channel estimation framework of the underwater acoustic orthogonal frequency division multiplexing (OFDM) system. It improves the GOMP algorithm from multiple perspectives as well. Firstly, a reasonable atomic threshold is proposed for precisely screening atoms. Using the atomic threshold not only improves the support set reliability but also shortens the running time by reducing the complexity of operations. Secondly, the idea of backtracking is introduced to eliminate the misselected atoms contained in the algorithm to improve channel estimation accuracy. Finally, through an adequate consideration of noise, the iteration stop condition is set to the L2 norm of the noise. The multiple experiment results demonstrate that the proposed algorithm can effectively estimate shallow underwater acoustic channels in a noisy environment, and has obvious advantages over the comparison algorithm in estimation accuracy, running time, spectrum saving, robustness and other aspects.
Issue:	Vol. 44 No. 5 (2023)
Published:	2023-05-28
PDF HTML