| In the past decades, on one hand, various novel technologies such as cognitive radio, cooperative relay and multiple input multiple output have appeared in the wireless communication. On the other hand, in order to decrease the energy consumption and environmental contamination, and to satisfy the requirements of high spectrum or energy efficiency, the wireless communication network has developed from the simple cellular network to a number of advanced wireless networks, such as device-to-device networks, wireless full-duplex networks, and virtual multiple-input multiple-output networks, etc. In these advanced wireless networks, it is crucial to effectively allocate the limited resources, select the optimum communication mode, and manage the interferences for the improvement of the network performance. This dissertation focuses on the hybrid networks of cellular and D2D users, and studies how to use the power allocation, mode selection and interference management to maximize the network capacity while taking QoS provisioning into account. The main contributions of this dissertation lie in:1. The QoS-guaranteed power allocation and mode selection schemes are investigated. We build the channel analysis model based on the imperfect CSI, and develop the joint power allocation and mode selection (JPAMS) algorithm, which provides a new prospect on the resource allocation and mode selection schemes. The JPAMS algorithm aims at minimizing the total transmit powers and selecting the optimum communication mode subject to the given QoS constraint. Firstly, the impact of imperfect CSI on the power allocation is analyzed. By converting the power allocation optimization problem into a convex optimization problem, we derive the closed-form solution to the power allocation scheme. Secondly, the optimum mode selecton is proposed based on the principle of minimizing the total powers, which guarantees the QoS for both the cellular and D2D users. The analysis result demonstrates that considering the effect of the imperfect CSI on the proposed JPAMS algorithm can avoid consuming more power to some extent. Thus, it is very important to consider the impact of the imperfect CSI on the system performance for D2D communications.2. The interference alignment scheme for the single-cell uplink hybrid D2D communications system based on TDD (Time Division Duplexing) mode is investigated. The proposed interference cancellation scheme can obtain the optimum degrees of freedom of the hybrid networks. The degrees of freedom of the k-user D2D hybrid networks increases by (k-l)/2 compared with the traditional cellular network. Then, we develop the optimum power allocation scheme based on interference alignment to obviously improve the sum-rate performance. However, the interference alignment algorithm in this part needs the global state information of the hybrid networks at the transmitters, which results in great network load and poor practicability. Thus, the interference alignment algorithm of this part is only a theoretical explore of the proposed algorithm. In the future work, we need to design a distributed interference alignment algorithm based on TDD mode without the global channel state information.3. The interference alignment scheme for the multi-cell downlink hybrid D2D communications system based on TDD (Time Division Duplexing) mode is investigated. The receivers calculate the transmit precoding matrix by use of the self-interference channel information to further complete interference alignment. Due to the less performance consumption, the interference alignment algorithm of this part is feasible and practical. In particular, we compare the sum-rate performance of the interference alignment strategies under different transmit precoding matrices design. Our analysis indicates that the proposed interference alignment-based precoding design can obtain larger capacity than the selfish precoding design algorithm.4. The interference management problem of the D2D full-duplex network underlaying the cellular network is investigated. For the full-duplex bidirectional communication link of the hybrid networks, we build a k-user interference channel model. Unlike the existing self-interference cancellation techniques for the full-duplex communication, we propose the joint mutual interference and self-interference interference alignment strategy based on the principle of interference alignment. The scheme cancels the aligned interferences by developing the precoding matrices at the transmitter and the receiver respectively. For the full-duplex relay D2D unidirectional transmission, in order to solve the problem of the limited number of users for the 2-phase iterative interference alignment algorithm, we propose the multi-phase full-duplex iterative interference alignment scheme applying to any k-pair D2D users. Under this premise, the proposed scheme can also guarantee the degrees of freedom of the hybrid networks.
|
PDF Full Text Request