Channel estimation method with improved performance for the UMTS-TDD mode

Abstract

Channel estimation is an essential building block for UTRA-TDD high performance receivers. Once the performance of the channel estimator algorithm proposed by 3GPP is highly dependent on the time spreading between consecutive multi-path components, a Successive Multi-path channel Estimation Technique (SMET) that improves the time resolution is proposed in this paper. A SMET based maximum likelihood approach for vectorial channel estimation, to include the estimation of the direction-of-arrival, is also proposed. This algorithm solves efficiently the complex problem of DOA estimation of multiple users in a multi path propagation environment even when the number of required DOA's exceeds the number of antenna array elements. Another property of the proposed algorithm is its ability to resolve signals from different users arriving from the same direction. This is due to processing in both time and space dimensions. The performance of these algorithms is assessed by resorting to simulations in multi-path environments using the UMTS-TDD specifications, and also by comparing the rms estimation errors against the Crámer-Rao Bound. The effect of imperfect channel estimation on the performance of RAKE and Hard-Decision Parallel Interference Canceller receivers is also analysed. The results show that a good performance can be achieved with SMET, from low to high values of Eb/n0.