Hardware and software implementations of an MMSE equalizer for MIMO-OFDM based WLAN

Communication over a MIMO (multiple-input-multiple-output) channel promises several advantages: increase in channel capacity, reduced transmit power, greater coverage, and improved link robustness. The minimum mean squared error equalizer (MMSE) is a potential algorithm in addressing the MIMO detection challenge. This paper presents four implementations for the computation of the MMSE equalizer coefficients. The studied options are: (1) a general-purpose microprocessor (ARM926EJ-S); (2) a traditional general purpose DSP (RD16024); (3) an embedded vector processor (EVP₁₆); and (4) a dedicated hardware solution. We show that, the equalizer requires acceleration for real-time processing. In order to obtain a low cost, flexible multi-standard WLAN baseband implementation, EVP₁₆ is the only feasible solution among the four studied.