Study On Interference Alignment Techniques For Multi-cell Multi-user MIMO Systems

The performance of the multi-cell multi-user multiple-input multiple-output(MIMO) systems is fundamentally limited by inter-cell interference(ICI) and inter-user interference(IUI). Since both ICI and IUI are bottlenecks for multi-cell multi-user MIMO systems to achieve high spectral efficiency, developing an intelligent interference suppression scheme to mitigate those interferences is an important concern. The recently emerging techniques of interference alignment(IA) present a major breakthrough in understanding and managing interference to enhance the system throughput for MIMO systems.The dissertation describes and explores the benefits of IA for multi-cell multi-user MIMO systems. The main objective of this dissertation is to provide a series of designs for multi-cell multi-user MIMO systems, including optimal design for iterative linear precoding scheme, non-cooperative IA schemes, IA schemes with base station cooperation, and IA schemes with user cooperation. The main/major contributions and innovations of the dissertation are listed as follows:1. Design of IA requires the precoding techniques. This dissertation proposes two new iterative linear precoding algorithms. Based on generalized zero-forcing channel inverse precoding method, the first one can greatly reduce the complexity as against the conventional iterative precoding algorithm without compromising any channel capacity. The second one is based on generalized minimum mean square error channel inverse precoding method and also can reduce the complexity as against the conventional iterative precoding algorithm. Taking into consideration the effect of noise on channel capacity, this second algorithm has larger capacity than the first one.2. The dissertation proposes two new IA schemes for multi-cell multi-user MIMO systems without channel state information(CSI) exchange. Based on the idea of conventional zero-forcing interference alignment(ZF-IA) method, the first proposed IA scheme extends the conventional ZF-IA method from two-cell MIMO systems to generalized multi-cell MIMO systems with multiple stream transmission. And then, to enhance the sum rate performance with minimum additional complexity, a second non-iterate IA scheme is proposed.to modify the zero-forcing precoders of the extended ZF-IA scheme. Simulation results show that the second proposed IA scheme outperforms the first one. Besides, after inference, an antenna-constrained condition is presented for the two schemes in terms of the number of users, the number of cells and the number of data streams dispatched to each user.3. A new IA algorithm is proposed to exploit another channel resource available for base station cooperation. Based on the criterion of minimum mean square error with CSI exchange across two neighboring BSs, the proposed IA algorithm jointly designs the precoding vector at the transceiver and beamforming vector at the receiver to suppress the IUI and ICI from dominant cell. The ICI from aggregate cell is suppressed through the whitening filter at the receiver which uses the information about the aggregate cell interference covariance matrix for each user. The proposed scheme reduces the number of receive antennas compared to the extended ZF-IA scheme. Simulation results have confirmed the effectiveness of the proposed IA scheme.4. A new IA scheme exploiting user cooperation techniques for downlink multi-cell multi-user MIMO systems is proposed. The proposed scheme requires that the minimum number of antennas at each user should be smaller than the minimum number of antennas at each BS. The new scheme not only can reduce the complexity and make up for the loss of antenna gain compared to the conventional grouping method IA scheme, but also can greatly increase the channel capacity gain when the number of transmit antennas is larger than the required minimum number. And compared to the conventional grouping method IA scheme, the proposed IA scheme has larger capacity at low signal to noise ratio(SNR) when the number of transmit antennas equals the required minimum number.5. A new IA scheme for uplink multi-cell multi-user MIMO systems is proposed. The new scheme first constructs the transmit beamforming vectors so as to align the effective ICI channel. Based on this ICI channel alignment, each BS is able to treat these ICI channels as one ICI channel. Therefore, each BS can suppress ICI and IUI by designing receive beamforming vectors. Simulation results show that the proposed IA scheme outperforms its conventional counterpart. The proposed scheme requires that the minimum number of antennas at each user should be smaller than that at each BS.