Joint Angle Delay Estimation in MIMO Communication Systems

In this chapter, we investigate a joint angle-delay estimation in the THz communication system with a large-scale array. Due to a relatively high dimension of THz array and wideband of signal, the traditional narrow-band-array assumption that signals received in different antennas are coherent could not be satisfied. Meanwhile, high accuracy in angle-delay estimation could lead to a huge computation complexity. To deal with this contradiction, we first give out the system model of a large-scale array system and show the fact that signal bandwidth could be spread. After that, we present the CRLB of joint angle-delay estimation. Based on this system model, we prove that signals with different angle-delay are approximately orthogonal. Furthermore, two estimators are proposed: MLE with grid, and progressively reached CRLB estimator without grid. To overcome the huge complexity, we provide a new CS structure with a sparsely uniform line array, random analog matrix, and WMC. The GAMP-based algorithm is listed to solve this CS problem. Simulation shows that the estimators proposed in this chapter outperform those based on the multiple signal classification (MUSIC) method. In particular, MLE could reach CRLB but limited by the grid, while the other one could reach CRLB at a high SNR. The GAMP-based algorithm jointly estimates angle-delay at the cost of a lower complexity, whose accuracy reaches the Rayleigh limit.