Combined channel estimation and adaptive prediction for MC-IDMA systems

Combination of Orthogonal Frequency Division Multplexing (OFDM) and Interleave Division Multiple Access (IDMA) to form OFDM-IDMA scheme has been shown as a promising approach to resolving the dual problems of intersymbol interference (ISI) and multiple access interference (MAI). For the optimum performance of this scheme, efficient method of obtaining channel state information (CSI) to be used by the multi-user detector for accurate signal detection is required. This paper proposes channel estimation scheme that employs combination of linear estimation of the frequency domain channel transfer function CTF) and the adaptive prediction of the time domain channel impulse response (CIR). Regularized adaptive algorithm is employed for the CIR prediction in order to exploit the sparsity characteristics of the OFDM channel. The achievable performance of the proposed channel estimation scheme is documented in the context of OFDM-IDMA system operating in both very slow and very fast mobile speed scenarios.