Research And Optimization On ⅡC Detection Algorithm In Large MIMO Communication Systems

With the development of modern wireless communication, the requirement of higher transmission rate and better transmission reliability becomes more and more urgent. Considered as one of the key techniques, MIMO has a good performance in dealing with multi-path fading and provides significant increases in spectrum efficiency and channel capacity without extra transmit power and additional bandwidth, which mainly results from spatial multiplexing gain and spatial diversity gain. In order to make full use of benefits from MIMO, a larger number of antennas are introduced into MIMO system, that is, Large MIMO system with tens or hundreds of antennas in both the transmitter and the receiver. Large MIMO system with large number of antennas provides more spatial dimensions and more degrees of freedom, which contributes the higher spectrum efficiency and the higher capacity. However, signal detection algorithm at a receiver is the key bottleneck to the performance for MIMO/Large MIMO communication systems.As is known, fundamental signal detection algorithms are maximum likelihood (ML) detection and linear detection as well as no-linear detection. Briefly, ML detection achieves the optimum performance in terms of bit error rate, while has an exponential order of complexity. Some sub-optimum detection algorithms, such as sphere decoding and successive interference cancellation, have lower order of complexity, but when it comes to Large MIMO systems, they are obviously unavailable. Additionally, linear detectors, including zero forcing and minimum mean square error algorithms, are simple but has a poor bit error rate performance. However, iterative interference cancellation (IIC) algorithm based on maximum ratio combining is efficient in terms of bit error rate performance and effective in terms of complexity, which benefits from the processes of interference cancellation and iteration as well as the feedback structure. Unlike hard decision, partial decision (PD) is a symbol-wise decision method. PD method based on a given threshold deals in-phase components and quadrature components respectively, and the threshold determines those components to be detected in terms of reliability. In this way, significant decreases in the probability of inaccurate decision comes out.Considering the low complexity and good bit error rate performance of iterative interference cancellation algorithm, we apply it to the Large MIMO system. IIC algorithm makes full use of spatial dimension provided by the large number of antennas and achieves better BER performance with increased antennas, that is, large effect. Additionally, the introduction of PD method alleviates the error propagation problem effectively, which increases the order of receive diversity. Therefore, PD-IIC has an improved BER performance with negligible increases in computational complexity. However, IIC/PD-IIC shows poor BER performance in the cases of high order of modulation. Fortunately, IIC is sensitive to the receive diversity, so the multi-stage IIC is proposed. Given multiple stages, M-IIC combined with PD method detects each component of high reliability in layer-wise and stage-wise and removes those detected components from the received signal vector, as a result, the original MIMO system transfers to a system with extra receive diversity, leading to the improvement in terms of BER performance without increases in number of antennas.