Abstract
There has been an increased demand for multicast transmission over wireless networks in recent years. Satellite system, such as Broadband Global Area Network (BGAN) provided by the Inmarsat, is indeed ideal for multicast delivery with its inherent wide geographic reach. However, the heterogeneous channel conditions and capacity constraints in the satellite links motivates employment of an adaptive approach in reliable multicast transmission provisioning in a pure bi-directional satellite network. Within this context, this thesis aims to increase resource efficiency and enhance reliable multicast transmission performance with the emphasis on mechanisms applicable to BGAN. The work carried out to achieve these objectives is divided into four main parts. The first part proposes channel-aware scheduling algorithms utilising channel state information (CSI) from the terminals. The results show that the proposed scheduling algorithms increase forward link resource utilisation whilst providing reasonably good session durations compared to that of the static channel representation benchmark scheme, hi the second part, a Change Detection (CD) mechanism is used in the CSI collection policy with the constraint of limited resources in the return link. The simulation results demonstrate that the traffic volume in the return link is reduced by up to 32% as compared to a system without any mechanism for CSI collection. In the third part, the integration of the forward link channel-aware scheduling algorithms and the CSI collection and suppression policy in the return link are investigated. The integration is studied analytically, resulting in improvement to the mathematical approximations of the performance parameters which are forward link resource efficiency and session duration observed by the terminals. The proposed adaptive techniques differ from other works where the forward and the return link approaches are either separately quantified or inadequately examined. Finally, the adaptive approach is further enhanced by employing an optimum link adaptation mechanism as a fade mitigation technique. The proposed algorithm sets an optimal transmission rate adaptively, in order to increase system throughput whilst offering fair transmission to the terminals. The results show that fewer retransmitted packets and fairer transmissions are achieved compared to other link adaptation techniques. Key words: multicast scheduling, link adaptation, feedback implosion suppression, reliable multicast transmission, BGAN, geostationary satellite networks.