基于贝叶斯压缩感知理论的超宽带通信信道估计

摘要

Ultra Wide-band (UWB) is a newly developing high-speed wireless communication technology.It is difficult to sample it directly as its wider bandwidth.However,compressed sensing(CS) provides a feasible way with lower sampling speed.Considering the current UWB channel estimation reconstruction algorithm which is based on CS must be used on the assumption that channel sparsity is known,an UWB channel estimation method based on bayesian compressed sensing(BCS)is proposed in this paper.UWB channel estimation is translated into CS reconstruction problem.Then BCS method is used in reconstruction to get the channel estimation value and error bound,finally decode UWB signal.BCS method introduces the sparse bayesian learning theory into CS.It sets posterior probability density function that is controlled by hyperparameters to each value in the vector that demand reconstruction.In the process of updating,many of the hyperparameters tend to infinity for those values that have insignificant amplitudes,and these corresponding posterior probability tends to zero.This method can reject trivial multipaths,find the critical paths in channel vector automatically and reconstruct them with regression algorithm.The experiment results show that this method can reconstruct original channel effectively under the circumstance that channel sparsity is unknown.%超宽带是一种新颖的高速无线通信技术.其过高的带宽给采样带来了困难,压缩感知理论提供了一种可行的低速采样方法.针对目前的压缩感知超宽带信道估计方法必须假设信道稀疏度已知,论文提出了基于贝叶斯压缩感知理论的超宽带信道估计方法.将超宽带信道估计转化为压缩感知理论中的重构问题,并使用贝叶斯压缩感知方法进行重构,得到信道估计值及其误差范围,最终实现信息解调.贝叶斯压缩感知理论将稀疏贝叶斯学习理论引入到压缩感知中,给需要重构向量中的每个值设置受超参数控制的后验概率密度函数,在超参数的更新过程中,零值所对应的超参数将趋向于无穷大,与之对应的后验概率将趋向于零,通过这种方法剔除非重要多径,自适应地找出信道向量中的重要多径,并使用回归算法进行重构.实验结果表明在信道稀疏度未知的情况下,该方法能够对原信道进行有效的重构.