Doubly-iterative equalization of space-time bit-interleaved coded modulation

We address the doubly-iterative space-time equalization of broadband wireless signals for frequency selective multiple-input multiple-output (MIMO) systems employing M-ary modulation through bit to symbol mapping. The proposed receiver consists of a soft-information aided MMSE space-time equalizer processing antenna array observations at the front-end, followed by the serial concatenation of two soft-input soft-output (SISO) modules: a symbol-to-bit demapper and a channel decoder. In the proposed scheme, a new equalization stage is started only after several inner demapping/decoding iterations are applied to improve the soft decisions. The doubly iterative structure offers not only a reduced overall computational complexity but also a significant equalization gain as illustrated by the results.