الفهرس | Only 14 pages are availabe for public view |
Abstract Abstract Increasing the throughput is an important objective for wireless ad-hoc networks. Many methods have been innovated for this purpose and on top of them is the network coding. The existing network coding schemes, such as COPE and its updated versions, have succeeded to provide a remarkable throughput gain in case of unicast flows, while they failed to provide the same performance in case of the multicast scenario. With the notable flourish of conference-based and multimedia streaming applications that are mainly depending on multicast flows, it becomes crucial to find a method that is able to deal efficiently with both unicast and multicast flows. In this thesis, we provide a novel enhanced network coding scheme, which we call Graph-Based Network Coding ”GBNC” that is able to handle both unicast and multicast flows simultaneously with the same performance. The proposed scheme incorporates the graphic theory and the elimination technique to efficiently discover all possible coding opportunity and avoid the draw backs of the previous coding methods. It even provides the option to favor one type of flow over the other when needed. The extensive simulation results reports the ability of the proposed scheme to achieve similar throughput gain to that of COPE in unicast flows and nearly double the gain in case of multicasting. |