Research And Validation Of MIMO Detection In LTE-Advanced Systems

In recent years, with the development of wireless technology, mobile communication has become an indispensable tool for people. Due to the dramatic increase in the number of mobile communication users and the rapid change of broadband multimedia services, the demand for transmission rate and performance of mobile communication system becomes increasingly high. On the promise that is without increasing the system bandwidth and power, Multiple Input Multiple Output (MIMO) technology can increase system capacity exponentially, which has become the key component of the fourth generation mobile communication system. However, with the increase of the number of antennas and modulation order, the MIMO detection algorithm complexity will be rapid growth in the system receiver. So it is an important issue to look for an excellent performance and low complexity detection algorithm that can be implemented in hardware product.This paper first introduces the development history and technology character of mobile communication system. It briefly describes the basic knowledge of OFDM system, LTE-A physical layer, downlink reference signal design and downlink MIMO technology, especially the signal generation for spatial multiplexing and transmit diversity. Then the paper researches the MIMO detection algorithms, including ML algorithm, linear detection algorithms, interference elimination algorithms and simplified ML algorithms, and analyses the performance of these MIMO detection algorithms through simulation. Finally, according to the signal processing procedure in hardware products, we build a LTE-A physical layer downlink simulation platform based on fixed-point operation, and add auxiliary signal processing modules and MIMO detection module to the platform. Auxiliary signal processing modules are used to compensate receiving signal’s frequency offset, time offset or amplitude loss. MIMO detection module selects MMSE algorithm for fixed-point operation, and mainly introduces the realization of noise measurement process and simplification of signal detection process. The module’s output is soft estimated symbols. The results show that the implemented MMSE detection has good performance while low processing delay, which can reach the real-time requirement. All modules meet the demands of system design and can be used in hardware products.