60 GHz Ultra-Band Channel Estimation Based on Cluster-Classification Compressed Sensing

In this chapter, we investigated the application of compressed sensing in channel estimation for the emerging 60 GHz millimeter-wave communications. We firstly investigate the regular orthogonal matching pursuit (Regularized OMP) algorithm for 60 GHz systems, and then consider the characteristics of 60 GHz channel, a Cluster-based Classification Compressed Sensing Algorithm is finally proposed on this basis. It may significantly reduce the reconstruction error of the channel estimation. Error ratios of CS-ROMP and algorithm based Cluster-Classification are thoroughly compared and comprehensive analysis is given relying on the experimental simulations. The results show that CS-ROMP algorithms can be properly applied to channel estimation of the 60 GHz system. The developed Cluster-based Classification Compressed Sensing Algorithm shows a superior performance in both the precision of the channel estimation and the complexity of reconstruction.