Researches On Key Technologies Of Multi-Point Cooperative Communications

Two distinct features characterize the modern wireless mobile communication systems, which are broadband high-speed transmission and inter-communication in mobility. Those two features pose several challenges for system designs, such as inter-cell interference mitigation, reliable connection during movement, distributed/centralized signal processing and so on. Cooperative communication is regarded as one of the most effective technologies to meet the challenges. In particular, it has been shown that cooperative communication is the core technology to improve the cell-edge throughput as well as the overall throughput. Therefore, researches of multi-point cooperative communication technologies are of high value both in theoretical studies and in practical uses.In this thesis, recent research activities in multi-point cooperative communications are comprehensively addressed and systematically analyzed. This dissertation studies the key technologies of multi-point cooperative communications in interference-cognitive scenarios, the scopes of which cover the enabling technologies of the fourth and future generations of wireless mobile communication systems. The contents include interference coordination technologies, antenna selection techniques in multiple MIMO-relay/base station systems, linear precoding schemes for distributed base station systems, as well as limited-bit feedback methods.In the inter-cell interference coordination technology, we propose a scheme consisting of user classification and resource scheduler from an altruistic perspective, which aims to reduce the interference to adjacent cells. To verify the effectiveness of the proposed scheme, we resort to the approach of system-level cellular network simulations. Simulation results show that compared with the exising schemes, the proposed scheme can increase the efficiency of the tradeoff between the cell edge throughput and the overall throughput by more than 50%. Therefore, the proposed scheme exhibits high effectiveness in interference coordination and can be implemented with small load of inter-base-station signaling, which are very beneficial for the uplink FDMA cellular systems. The novelty of the proposed scheme is that, the base station assigns different levels of clearance for users to access the frequency resources based on the users'resistance of interference and tendency of producing interference. If a user tends to make a large amount of interference to adjacent cells, the base station will restrict the frequency resources for the user to use, thus reducing the network interference level from an altruistic perspective, and hence the cell edge throughput can be leveraged. On the other hand, if a user is not likely to cause much interference to adjacent cells, the base station will grant the user to have access to the edge user resources of the cell if they are available, thus the overall throughput can be improved.In the antenna selection technology for multiple MIMO-relay systems, we propose three greedy antenna selection algorithms, namely, DZF-GCM algorithm, GCM algorithm and GMM algorithms. They are either based on system capacity maximization criterion or based on MSE minimization criterion. Under the assumption of ZF processing at both transmit end and receive end, simulation results show the proposed DZF-GCM algorithm can achieve higher system capacity than the existing schemes, and the gain is more pronounced in the low SNR regime at the relay nodes. Under the assumption of general processing at both transmit end and receive end, simulation results show that the system capacity obtained from the existing schemes increases with the relay number by the rate of O(loglogK), while the proposed GCM algorithm can make the system capacity scale like O(logK) with the relay number, achieving the upper bound of system capacity in the sense of asymptotic capacity. Under the assumption of general processing at both transmit end and receive end, the proposed GMM algorithm can achieve higher diversity gain and lower bit error rate than the existing algorithms. Moreover, simulation results show that the proposed GMM algorithm and the exhaustive search method have similar MSE performance, but the GMM algorithm is much easier to implement. Therefore, the proposed greedy antenna selection algorithms can achieve relatively larger system capacity, relatively smaller MSE and its computational complexity is smaller than that of the exhaustive search method to a great extent, which is of high value for theoretical studies and practical applications. The novelty of the proposed scheme is that, the closed-form expressions of the capacity increase and MSE decrease resulted from adding one more antenna pair are carefully derived, followed by designs of low-complexity greedy antenna selection algorithms.In the linear precoding technology for distributed base stations with perfect backhaul communications, we propose a precoding scheme, which is a combination of antenna selection and single-frequency network precoding. Theoretical analysis and simulation results show that in the case of single-antenna users, the proposed scheme and the weighted local precoding scheme can achieve similar optimal SINR performance compared with the global precoding scheme. Furthermore, when considering the case of multi-antenna users and more practical scenarios, simulation results show that even if restricted search space is exerted on the antenna selection process, and the beneficial operation of antenna power compensation is removed, the proposed scheme still can achieve higher system capacity than those of the local precoding, weighted local precoding and single-frequency network precoding schemes. Thus, the proposed scheme effectively closes the performance gap with the optimal global precoding scheme. In addition, the proposed scheme, compared to the weighted local precoding scheme, requires a lower load of feedback overhead and is easy to implement. The novelty of the proposed scheme is that, transmit antenna selection is first applied to improve the capacity of the virtual single-frequency network channel, and then the single-frequency network precoding process is invoked. In the linear precoding technology for distributed base stations with imperfect backhaul communications, we propose a sequential and incremental precoding scheme based on the criterion of minimizing the largest sub-stream MSE. Simulation results show that when the backhaul links are relatively reliable, the proposed scheme can obtain similar maximum sub-stream MSE performance compared with the existing schemes. However, when the backhaul communications suffer from more serious connectivity problems, the proposed scheme exhibits considerable performance gains compared with the existing schemes. And the robustness of the proposed scheme is more evident in the high SINR regime or when the equal power allocation over the substreams is engaged. In addition, through simulations for limited-bit feedback systems, the proposed scheme retains the aforementioned advantages. Hence, considering the practical backhaul conditions and the limited-bit feedback methods, the proposed scheme is very useful for implementation. The novelty of the proposed scheme is that, the imperfectness of the backhaul communications is modeled as probability events and considered as a relevant factor in the optimization process. We first optimize the precoding matrix for the channel between the serving base station and the user. Then according to the descending order of the activation probabilities of the helper base stations, we sequentially generate the optimal precoding matrix for each helper base station by an iterative algorithm with the previously optimized precoders fixed.Finally, we give a brief summary of the innovative works and suggest some future research areas of multi-point cooperative communications technologies as well.