Md Shamsul Alam
Modeling and Performance Analysis of Relay-based Cooperative OFDMA Networks
Shen, Sherman X. and Mark, Jon W. (Adjunct)
Next generation wireless communication networks are expected to provide ubiquitous high data rate coverage and support heterogeneous wireless services with diverse quality-of-service (QoS) requirements. This translates into a heavy demand for the spectral resources. In order to meet these requirements, Orthogonal Frequency Division Multiple Access (OFDMA) has been regarded as a promising air-interface for the emerging fourth generation (4G) networks due to its capability to combat the channel impairments and support high data rate. In addition, OFDMA offers flexibility in radio resource allocation and provides multiuser diversity by allowing subcarriers to be shared among multiple users.
One of the main challenges for the 4G networks is to achieve high throughput throughout the entire cell. Cooperative relaying is a very promising solution to tackle this problem as it provides throughput gains as well as coverage extension. The combination of OFDMA and cooperative relaying assures high throughput requirements, particularly for users at the cell edge. However, to fully exploit the benefits of relaying, efficient relay selection as well as resource allocation are critical in such kind of network when multiple users and multiple relays are considered. Moreover, the consideration of heterogeneous QoS requirements further complicate the optimal allocation of resources in a relay enhanced OFDMA network. Furthermore, the computational complexity and signalling overhead are also needed to be considered in the design of practical resource allocation schemes. In this dissertation, we conduct a comprehensive research study on the topic of radio resource management for relay-based cooperative OFDMA networks supporting heterogeneous QoS requirements. Specifically, this dissertation investigates how to effectively and efficiently allocate resources to satisfy QoS requirements of 4G users, improve spectrum utilization and reduce computational complexity at the base station. The problems and our research achievements are briefly outlined as follows.
Firstly, a QoS aware optimal joint relay selection, power allocation and subcarrier assignment scheme for uplink OFDMA system considering heterogeneous services under a total power constraint is proposed. The relay selection, power allocation and subcarrier assignment problem is formulated as a joint optimization problem with the objective of maximizing the system throughput, which is solved by means of a two level dual decomposition and subgradient method. The computational complexity is finally reduced via the introduction of two suboptimal schemes. The performance of the proposed schemes is demonstrated through computer simulations based on OFDMA network. Numerical results show that our schemes support heterogeneous services while guaranteeing each user's QoS requirements with slight total system throughput degradation.
Secondly, we investigate the resource allocation problem subject to the satisfaction of user QoS requirements and individual total power constraints of the users and relays. The throughput of each end-to-end link is modeled considering both the direct and relay links. Due to non-convex nature of the original resource allocation problem, the optimal solution is obtained by solving a relaxed problem via two level dual decomposition. Numerical results reveal that the proposed scheme is effective in provisioning QoS of each user's over the conventional resource allocation counterpart under individual total power constraints of the users and relays .
Lastly, decentralized resource allocation schemes are proposed to reduce the computational complexity and CSI feedback overhead at the BS. A user centric distributed (UCD) scheme and a relay centric distributed (RCD) scheme are proposed, where the computation of the centralized scheme is distributed among the users and relays, respectively. We also proposed suboptimal schemes based on simplified relay selection. The suboptimal schemes can be combined with the distributed schemes to further reduction of signalling overhead and computational complexity. Numerical results show that our schemes guarantee user's satisfaction with low computational complexity and signalling overhead, leading to preferred candidates for practical implementation.
The research results obtained in this dissertation can improve the resource utilization and QoS assurance of the emerging OFDMA networks.