Performance analysis of K-best detection with adaptive modulation

Abstract

In this paper, the K-best detection algorithm with an adaptive modulation scheme in multiple input multiple output (MIMO) systems is proposed. A simplified error probability approximation method based on the union bound (UB) of the Maximum-Likelihood detector (MLD) is proposed to predict the bit error rate (BER) of the K-best algorithm. In specific, the simplified approach only uses the minimum Euclidean distance (MED) events which is suitable for the adaptive modulation scheme with much reduced computational complexity. In order to improve the accuracy of prediction, the signal-to-noise ratio (SNR) gaps between the UB with MED events and the full UB are estimated. Finally, simulation results have clearly shown the adaptive K-best algorithm applying the simplified approximation method has much reduced computational complexities while maintaining a promising BER performance.