| dc.description.abstract | Cooperative communication has recently become a key technology for modern wireless networks such as 3GPP long-term evolution and WiMAX, because in such networks the transmission rate, the communication reliability, and coverage problems could be improved in a cost-effective manner. This, however, faces many design challenges. First, cooperative transmission typically involves a relaying phase which requires extra resources. This may cause a reduction in the spectral efficiency. Second, extra control signaling increases the complexity of operation, which may limit practical implementation. In addition, a wireless channel is time-varying, mainly due to the multipath propagation. As a result, a careful design of efficient cooperative communication systems is required, not only to enhance the spectral efficiency and maintain the quality-of-service (QoS), but also to be practical.
In this dissertation, we aim to address the challenges imposed by cooperative communication and wireless transmission, and design the efficient and distributed systems which can be practically implemented in existing wireless systems. The research work is divided into two main topics: 1) adaptive cooperative wireless systems with variable-rate transmission, and 2) cooperative wireless systems with a power consumption constraint.
The first topic investigates how the spectral efficiency of cooperative wireless communication systems can be improved while maintaining the QoS in terms of bit error rate and outage probability. The spectral efficiency enhancement is achieved by using three techniques: adaptivity over the relay node (i.e., relay node is active or not), adaptivity over the modulation mode, and relay selection. Based on that, we propose several adaptive cooperative schemes for both the decode-and-forward (DF) and amplify-and-forward (AF) protocols. To evaluate these schemes, we provide performance analysis in terms of average spectral efficiency, average bit error rate (ABER), and outage probability over Rayleigh fading channels.
We start with the single-relay cooperative system using DF protocol, in which two adaptive cooperative schemes with variable-rate transmission are proposed. The first scheme, called the minimum error rate scheme (MERS), aims to exploit the transmit diversity to improve the bit error rate. By trading the multiplexing gain against the diversity gain, we propose the second scheme, called the maximum spectral efficiency scheme (MSES), in which cooperative transmission is avoided whenever it is not beneficial. The MERS improves the ABER significantly and achieves equal or better average spectral efficiency compared to the fixed (i.e., non-adaptive) relaying scheme. In contrast, the MSES provides the best average spectral efficiency due to its ability to not only adapt to the channel variation but also to switch between cooperative and non-cooperative transmissions. To further increase the spectral efficiency, we then propose the third scheme, called variable-rate based relay selection (VRRS) scheme, in which a relay node is selected from among the available relay nodes, based on a predefined criterion.
Furthermore, we propose two AF adaptive cooperative schemes, mainly to enhance the spectral efficiency. In the first scheme, we introduce a generalized switching policy (GSP) for a single-relay cooperative wireless system that exploits the variable-rate transmission and useful cooperative regions. The second scheme, called the AF efficient relay selection (AFERS) scheme, extends the GSP to also consider the relay selection technique. Analytical and simulation results verify that the AFERS scheme not only outperforms conventional direct transmission in terms of the average spectral efficiency, but also the AF fixed relaying and the outage-based AF adaptive cooperative scheme.
The second topic investigates the fair power consumption of the relay nodes for AF cooperative wireless communication systems. The fairness is defined as to achieve equal power consumption over the relay nodes. We focus on how the relay selection process can be controlled in a distributed manner so that the power consumption of the relay nodes can be included in relay selection. We first introduce a simple closed-form expression for the weight coefficient used in order to achieve the considered fairness that depends only on the local average channel conditions of the relay path. We then derive closed-form expressions of the weighted outage probability and ABER and show that our proposed strategy not only has less complexity than the conventional centralized one but also provides better accuracy in distributing the total consumed power equally among the relay nodes without affecting the performance. | en |
