Advanced MIMO Detection Algorithms In LTE System

MIMO technology is a key technology in LTE system, for its improving of capacity and stability of the system without additional cost of spectrum resource. The former gets gain from spatial multiplexing, while the latter form spatial diversity. This paper mainly studies the MIMO detection algorithms in spatial multiplexing cases and provides simulation results of these algorithms in two typical LTE scenarios, the simple single-cell scenario and the dual-cell scenario with co-channel interference.MIMO detection algorithms can generally be divided into two kinds, linear detection algorithms and nonlinear detection algorithms. Zero Forcing (ZF) algorithm and Minimum Mean Square Error (MMSE) algorithm are common linear algorithms, while Maximum Likelihood (ML) algorithm is a typical nonlinear one. However, the complexity of basic linear algorithms is low, and the performance is bad. ML algorithm is considered to be the optimum MIMO detection algorithm, but as the number of transmitter antennas increases and the modulation order is increased, its complexity grows exponentially, which is unacceptable. Some MIMO detection algorithms with iterative structure also have excellent performance, such as Successive Interference Cancellation (SIC) and Parallel Interference Cancellation (PIC). Therefore, the study of advanced MIMO detection algorithms in LTE system with excellent performance and acceptable complexity has two aspects. On the one hand is to simplify the ML algorithm to find an advanced algorithm having the same performance with ML algorithm but a much lower complexity. On the other hand is focusing on the iterative structure, looking for a suitable iterative detection algorithm with excellent performance.For the single-cell scenario, the paper gives out the system model and deduces MMSE algorithm, MMSE-Parallel Interference Cancellation (MMSE-PIC) algorithm and List Sphere Decoding (LSD) algorithm based on the model. On this basis, we designed LSD-Parallel Interference Cancellation (LSD-PIC) algorithm. The paper also gives the system model for dual-cell co-channel interference scenarios, and the adaptive MMSE algorithm, MMSE-Interference Rejection Combining (MMSE-IRC) algorithm. On this basis as well, using the idea of SIC, two algorithms called MMSE-Interference Cancellation (MMSE-IC) and LSD-Interference Cancellation (LSD-IC) algorithm are designed.The paper gives simulation results of the above algorithms in real LTE scenarios and analyzes the performance and complexity of these algorithms. Considering all the factors, we get a conclusion that MMSE-PIC algorithm is the best compromised algorithm in single-cell scenario considering both performance and complexity. Although LSD-PIC algorithm has a better performance than MMSE-PIC algorithm, LSD-PIC algorithm’s complexity is too high while MMSE-PIC algorithm’s performance is moderate.In the case of dual-cell co-channel interference scenarios, MMSE-IC algorithm and LSD-IC algorithm almost have the same performance, which is significantly better than other linear algorithms’. Taking into account the high complexity of LSD-IC algorithm, it is considered that MMSE-IC algorithm is the best compromised algorithm considering both performance and complexity.