Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Electrical and Computer Engineering

First Advisor

Sang Wu Kim


The concept of cooperative relay is an essential technique for future cellular networks such as wireless mesh networking or wireless ad-hoc networking. In a practical relay network, channel coding, network coding, and antenna arrays, will coexist and yet the joint optimization of these conventional coding schemes and cooperative relay is not well understood. To build a design guideline for relay network, this dissertation develop a joint optimization methodology for multiple coding schemes in multiple access relay network.

There are four major contributions in this thesis: First, we jointly optimize conventional coding schemes and radio resources of multiple access relay network with multiple antennas. The combined design of MIMO transmission modes, channel coding at the source, network coding at the relay have been investigated. We develop optimal design rule that minimize the end-to-end error probability. Second, we derive the fundamental tradeoff between achievable rate and reliability of multiple access relay network with multiple antennas. We consider three MIMO transmission modes, spatial multiplexing (SM), Alamouti coding as transmit diversity (TD), and Golden Coding, and random linear network coding at the relay. We compare the average decoding error probability of each transmission mode. Third, we present an interference cancellation scheme for multi-user MIMO. The proposed Log-likelihood-ratio (LLR) ordered successive interference cancellation (SIC) scheme provides 1 ~ 3 dB gain over the conventional SNR-ordered SIC and the gain increases with increasing number of users. Finally, we present a new architecture for MIMO receivers that cancel the co-channel interference (CCI) using a single radio frequency (RF) and baseband (BB) chain, while still achieving nearly the same bit error rate that can be provided by the conventional receiver requiring multiple RF/BB chains.

Copyright Owner

Young Jin Chun



Date Available


File Format


File Size

141 pages