Implementation of the Levinson algorithm for MMSE equalizer
Minsu Kim, Jin‐Yong Lee, Younglok Kim · 2008
In wireless communication system applying the minimum mean square error (MMSE) as its equalization algorithm, the equalizer is one of the most critical part in terms of the computational complexity of the baseband signal processing. Hence more efficient algorithm implementing MMSE equalizer is needed to save power consumption and reduce the hardware complexity. In this paper, we implement the recursive Levinson algorithm which is to solve the linear system solution in MMSE equalizer. The Levinson algorithm can be implemented with reduced hardware complexity and better operation speed compared to Cholesky decomposition which is one of linear system solution algorithm. In addition, the recursive algorithm has flexibility on the service quality due to the trade-off between performance and complexity which can be determined by the threshold value.