Interference management for two-way relay networks

Abstract

Relaying is a promising technique to provide high data service in wireless networks. Recently, many researchers have studied the throughput of two-hop or multi-hop relay networks. In a relay network with bidirectional transmission, the main drawback is the loss of spectrum efficiency caused by the half-duplex signaling. Inspired by the idea of network coding, two-way relaying is introduced to improve the spectrum efficiency by reducing the transmission time. With two-way relaying, the system allows all the nodes to send signals to the relay simultaneously. However, the interference caused by the simultaneous transmission undermines the performance of the system. Thus, interference management is necessary to improve the overall performance of two-way relay networks. In this thesis, we first investigate the power control of a three-node two-way relay network. In order to mitigate the interference, adaptive power control is applied to each node. We derive the optimal transmit power of each node to maximize the throughput and minimize the system outage probability, respectively. With practical individual power constraints, we show that the solutions based on the two metrics are different, which is in contrast to the case with a sum power constraint. Simulation results demonstrate that both the sum throughput and the system outage are improved by the proposed power control. We then extend the model to the multi-antenna and multi-user case and study the beamforming design for a two-way relay cellular network. Aiming to improve the throughput of the network, transmit beamforming at the base station (BS) and the relay is optimized. We first formulate the design problem, and show the difficulty of finding the optimal solution due to the non-convexity of the objective function. We therefore propose a suboptimal solution with a fixed beamforming structure, which decomposes the system into multiple parallel single-antenna two-way relay channels. Based on the proposed beamforming structure, the original problem is transformed into a joint power allocation problem. Finally, with high signal-to-noise ratio (SNR) approximations, we divide the joint power allocation problem into two convex subproblems. An efficient algorithm is then proposed to find the suboptimal beamforming. Simulation results demonstrate that the proposed beamforming achieves better throughput than the existing beamforming schemes. In particular, the gap grows when the relay position is away from the intermediate point between the BS and the users. In a two-way relay cellular network with a very large number of users, the before-mentioned beamforming is not able to completely remove the inter-user interference caused by the simultaneous transmissions of all the users. To avoid the inter-user interference and improve the overall performance of the system, we resort to user selection by selecting a certain number of active users from all the candidate users. With the objective of the sum throughput maximization, we formulate the problem of user selection based on the developed beamforming. After making a series of approximations, we derive a user selection algorithm with low computational complexity based on the channel vector projection. Finally, numerical results demonstrate that the proposed user selection achieves a similar sum rate compared with the optimal selection via the exhaustive search.

Date of Award	2013
Original language	English
Awarding Institution	The Hong Kong University of Science and Technology