Methodology Research For Coordinated Multi-point Communication Under Non-ideal Backhaul Constraint

In wireless networks, coordinated multi-point(CoMP) communication is used to mitigate the cell-edge interference, and improve system capacity, diversity and coverage region. Thus, the data rate of cell-edge user and overall performance of cellular systems are enhanced, which is of great importance to accommodate user experience and the ever growing need for capacity in broadband wireless communications. CoMP has been studied as a core characteristic in the 5th generation of mobile networks.In CoMP systems, the non-ideal backhaul, caused by the low throughput and high delay in physical links among base stations(BSs), is the major problem that hinders the implementation of CoMP in wireless networks. This paper targets the research upon CoMP with non-ideal backhaul constraint, specifically for the issues of CSI quantization, joint source-channel coding, asynchronous diversity strategy, network coding, and beamforming precode. The major work of this paper is five-fold:(1) Propose a CSI quantization method based on Grassmann manifold theory. By reducing the backhaul overhead, the method solves the problem of performance degradation caused by high delay and low bandwidth of non-ideal backhaul. Regarding a non-linear relationship between CSI quantization ratio and system performance of CoMP, quantization ratio regions that are sensitive and non-sensitive to the performance are classified according to the changing speed of quantization gain. The design aim for quantization ratio are constrained to the sensitive region by implementing the method of calculating a quantization gain.(2) For CoMP with band-limited backhaul, according to the channel characteristics, we establish a precoding channel model via zero-forcing beam-forming and Markov chain theory, and propose an unequal joint source-channel coding scheme. The unequal coding scheme, first of all, uses wavelet transform to compress the source with unequal information distribution, and further uses irregular low density parity check(LDPC) codes to fulfill the channel coding with unequal error protection, and therefore improves the system throughput for CoMP with band-limited backhaul. Based on the precoding channel model established by zero-forcing beam-forming and Markov chain, by keeping channel coding gain always above a certain threshold, and decreasing source compression ratio, not only does the scheme improve the system performance, but also it enhances the joint coding gain of CoMP with band-limited backhaul.(3) Propose a delay-tolerant wavelet coding(DWC) method. We prove in the paper that under asynchronous CoMP condition, DWC can almost achieve full diversity as long as the delay is less than the codeword length. In other words, the delay of non-ideal backhaul less than the codeword length will hardly influence the CoMP performance. Compared to existing schemes, the proposed method is easier to implement, more suitable for asynchronous CoMP, and has higher code rate.(4) Propose a wavelet network coding(WNC) scheme to improve the channel capacity for wireless backhauls of CoMP. The excellent frequency-time domain analysis property of wavelet transform can effectively share the system’s frequency-time resource. This offers a new solution to solve the problems of inter-symbol interference(ISI) in traditional sinusoidal modulation, and of the capacity bottleneck caused by Galois field modulation, respectively. WNC scheme can easily group a symbol’s priority in the frequency-time domain based on the symbol’s salience, and further determine whether to use redundant correction channel or not. Thus, a relatively controllable, flexible frequency resource-saving scheme is achieved. Theoretical and simulation results show that, as the number of coordinated points increases, the anti-ISI capability of WNC becomes stronger.(5) Propose a novel precoding scheme named RE-ZFBF for CoMP in 5G networks. It has been analytically shown that the interference can be optimized by the proposed scheme. The numerical results have verified that, with a non-ideal backhaul, RE-ZFBF can mitigate at most half of inter-channel interference and inter-user interference, provide better performance than the traditional ZFBF scheme, and outperform the existing schemes respectively in spectral-efficiency and energy-efficiency.