On adaptive reduced complexity soft output equalization for EDGE

Abstract

An adaptive soft output equalizer structure of moderate complexity, which involves the BCJR maximum a posteriori (MAP) is proposed to mitigate the adverse effects of a fast time-varying frequency selective fading channel such as the case of a high speed train for EDGE modulation. For practical low complexity implementation, the delayed decision feedback sequence estimation (DDFSE) is applied and using per-survivor processing (PSP), the channel derived parameters, which the soft output equalizer depend on for delivering soft values, are jointly updated. The sub-optimum soft-output algorithm (SSA) is also considered as another reduced state soft output option to MAP but at the expense of sub-optimality. For further complexity reduction, the reduced state sequence estimation (RSSE) technique is applied and the overall complexity of the joint scheme is effectively reduced as parameter update is carried out with a smaller trellis. Combining the PSP with reduced state soft output algorithms, adaptive soft output extensions of various complexity are proposed for EDGE. The results suggest an adaptive 2 state MAP is sufficient and capable of supporting mobiles of high vehicular speeds that exceeds EDGE requirements.