ABSTRACT
In order to adapt to the development of modern communication system, in 2004, 3GPP launched a mobile communication Long Term Evolution (LTE). LTE adopt MIMO and OFDM technologies which can provide high spectrum efficiency, high channel capacity and flexible spectrum allocation [1]. It will play an important role of mobile communications in the future. VBLAST (Vertical layered space-time code) encoding which is presented by Bell Labs can provide a high transfer rate [2], so it is very suitable for LTE system. The traditional LTE detection algorithms are mainly Maximum Likelihood (ML) algorithm, Zero-Forcing algorithm (ZF) and Minimum Mean Square Error (MMSE) algorithm. Among the three algorithms, simulation shows that the ML algorithm is optimal, but the shortcomings of the ML algorithm is that the complexity of it is exponential growth which cannot be practical as in [3]. Interference progressive cancellation of ZF and MMSE detection algorithm can obtain a better BER performance in [4,5]. An improved MMSE detection algorithm is put forward in [6] can obtains a low error rate, but also increases complexity. The various detection algorithms based on MMSE is introduced and showed their system capacity in [7,8]. Analysis of existing detection algorithm, this paper presents a bi-directional detection minimum mean square error algorithm of VBLAST in LTE system, simulation results show that this algorithm has better bit error rate performance and system capacity than the interference cancellation MMSE algorithm.
