Resource Management and Interference Control in Distributed Multi-Tier and D2D Systems

In order to improve the capacity and spectrum efficiency of next generation wireless networks, multi-tier wireless networking and device-to-device (D2D) communication are widely considered as strong candidates for 5G. In this thesis, I have developed new theories and design guidelines to improve the performance of large-scale multi-tier and D2D networks by studying their resource optimization and interference management. In the first part of this thesis, we study optimal power allocation for distributed relays in a multi-channel system with multiple source-destination pairs and an individual power budget for each relay. We focus on designing the optimal relay beamformers, aiming at minimizing per-relay power usage while meeting minimum signal-to-noise guarantees. Showing that strong Lagrange duality holds even for this non-convex problem, we solve it in the dual domain. Further, we investigate the effect of imperfect channel information by quantifying the performance loss due to either quantization error with limited feedback or channel estimation error. In the second part of this thesis, we study optimal inter-cell interference control for distributed relays in a multi-channel system. We design optimal relay beamforming to minimize the maximum interference caused at the neighboring cells, while satisfying minimum signal-to-noise requirements and per-relay power constraints. Even though the problem is non-convex, we propose an iterative algorithm that provides a semi-closed-form solution. We extend this algorithm to the problem of maximizing the minimum signal-to-noise subject to some pre-determined maximum interference constraints at neighboring cells. In order to gain insight into designing this system in practice, we further study the received worst-case signal-to-interference-and-noise ratio versus the maximum interference target. In the third part of this thesis, we consider D2D communication underlaid in a cellular system for uplink resource sharing. Under optimal cellular user (CU) receive beamforming, we jointly optimize the powers of CUs and D2D pairs for their sum rate maximization, while satisfying minimum quality-of-service (QoS) requirements and worst-case inter-cell interference limit in multiple neighboring cells. The formulated joint optimization problem is non-convex. We propose an approximate power control algorithm to maximize the sum rate and provide an upper bound on the performance loss by the proposed algorithm and conditions for its optimality. We further extended the results of the third part in the fourth part of this thesis, where we jointly optimize the beam vector and the transmit powers of the CU and D2D transmitter under practical system settings. We consider a multi-cell scenario, where perfect channel information is available only for the direct channels from the CU and D2D to the base station. For other channels, only partial channel information is available. The uncertain channel information, the non-convex expected sum rate, and the various power, interference, and QoS constraints, lead to a challenging optimization problem. We propose an efficient robust power control algorithm based on a ratio-of-expectation approximation to maximize the expected sum rate, which is shown to give near-optimal performance by comparing it with an upper bound of the sum rate.Ph.D