The Research Of Cellular Interference Alignment With Limited Feedback

The inter-cell interference (ICI) is one of the main factors that limits the system capacity in cellular systems. Interference alignment (IA) is a precoding technique which can manage the ICI and improve the system capacity effectively. It constrains the ICI into a signal subspace of small dimensions, allowing the interference free transmission of the desired signals in the remaining dimensions. As a result, the degrees of freedom (DoF) of the system is increased and the system capacity is improved. The channel state information at the transmitters (CSIT) is required for IA. In practice, the CSIT is fed back to transmitters after the quantization at receivers. The feedback information after quantization is called limited feedback. The limited feedback results in residual ICI, which limits the system capacity. The main contributions of the essay are as follows:Firstly, a quantization strategy is proposed. In the existing limited feedback IA schemes, the CSI is quantized and fed back by each receiver independently. Different from these schemes, the proposed strategy is a joint quantization strategy in uplink cellular systems. The precoding vectors are worked out at the base stations (BSs). Then, all the precoding vectors in the system are quantized together and then fed back by only one BS. The proposed strategy can achieve higher system throughput compared with the existing schemes with the same feedback cost.Next, a cellular limited feedback IA scheme is proposed based on an existing scheme. In this essay, the residual ICI power has been analyzed and a receiving filter has been proposed for minimizing the residual ICI power. The proposed filter is computed using the CSI and the quantized precoding vectors and can realize the minimization of the residual ICI power for each transmission. As a result, the system throughput has been improved.Lastly, an cellular IA algorithm special for limited feedback has been proposed in the interfering multi-access channels (IMAC). Though analysis, the concepts of the effective ICI alignment space is proposed. It is pointed out that to realize the minimization of the residual ICI power, the receiving filter should eliminate the energy of the ICI in the effective ICI alignment space and the precoding vectors should minimize the power leakage of the ICI outside the effective ICI alignment space. By transceiver design, the proposed scheme realize the minimization of the residual ICI power for each transmission under the condition of given codebooks. Therefore, the system throughput has been increased significantly. In addition, the proposed scheme extends the feasibility of the limited feedback cellular IA effectively.