Resource allocation in SVD-assisted broadband MIMO systems using polynomial matrix factorization

Abstract

Removing channel interference in broadband multiple-input multiple-output (MIMO) systems is a task which can be solved by applying a spatio-temporal vector coding (STVC) channel description and using singular value decomposition (SVD) in combination with signal pre- and post-processing. In this contribution a polynomial matrix factorization channel description in combination with a specific SVD algorithm for polynomial matrices is analyzed and compared to the commonly used STVC SVD. This comparison points out the analogies and differences of both equalization methods. Furthermore, the bit error rate (BER) performance is evaluated for two different channel types and is optimized by applying bit-allocation schemes involving a power loading strategy. Our results, obtained by computer simulation, show that polynomial matrix factorization such as polynomial matrix SVD could be an alternative signal processing approach compared to conventional SVD-based MIMO approaches in frequency-selective MIMO channels.