Two- and three-way relaying for wireless communications

Abstract

This thesis investigates about two- and three-way relaying for wireless communications. The contents of thesis are summarized as follows. First, we study how amount of capacity gain is obtained when the two-hop two-way relaying is applied into the cellular relay networks. Second, the power control method for two-way relaying is studied. Third, the three-way relaying extended from the conventional two-way relaying is proposed and its achievable rate region is specified. More specifically, in Chapter 2, we consider the transmission mode and user selection problem in cellular relay networks. The transmission mode selection denotes the determination between one-hop and two-hop two-way relaying under a BS, relay, and $\It{K}$ mobile-stations (MSs) system configuration. The best transmission mode and user selection is possible when the channel state information (CSI) between BS and RS, between BS and $\It{K}$ MSs, and between RS and $\It{K}$ MSs is fed to BS. However, this increases the feedback quantity as $\It{K}$ increases. Hence, we propose two-step scheduling such that the transmission mode is first determined and the user is selected using the CSI of previously determined transmission mode. This approach can reduce the feedback quantity by half compared to that of full feedback. We propose the metric of transmission mode selection by maximizing sum ergodic rate between one-hop and two-hop two-way relaying. The system level simulation shows that the performance loss of proposed two-step scheduling is marginal compared to that of optimal selection with full feedback. In Chapter 3, the power allocation for orthogonal frequency division multiplexing (OFDM)-based two-way relaying is studied. The conventional method was the optimal power allocation using dual decomposition under the individual power constraint of two terminals and a relay. However, the conventional method inefficiently utilizes the powers and the computational complexity is high. Hence, we propose ...