Joint Channel and Data Estimation for MIMO Communications using Sparse Pilots

In this paper, we present a joint channel and data estimation algorithm for 2 × 1 Space-Time Code (STC) and 2 × 2 Spatial Multiplexing (SMX) schemes. Very sparse pilots are exploited for the channel estimation, in order to reduce the pilot-related overhead. In the proposed algorithm, a trellis skeleton is established to simultaneously predict and track channel coefficients of both STC transmission branches. The a posteriori Probability (APP) detection is taken into account to compute the soft output bits. The proposed algorithm can be straightforwardly integrated into an iterative receiver. Namely, significant performance enhancement can be achieved even with little available a priori knowledge. Link layer simulations show that the ergodic capacity can be gradually approached with incremental a priori knowledge. Furthermore, we present the robustness of the proposed algorithm by tuning variation of the channel and sparsity of pilots, with which the conventional channel estimators can hardly deliver sufficient performance.