Authors : Topside E. Mathonsi, Okuthe P. Kogeda and Thomas O. Olwal
Abstract: Most Small and Medium Enterprises (SMEs) setup wireless networks instead of wired networks in order to make the process of communicating and sharing resources among their employees easier, more flexible, faster, cheaper and more ubiquitous. Furthermore, as the SMEs tend to grow in terms of its number of employees, the additional employees create the need for increased number of wireless communication terminal devices such as PCs, PDAs, iPads, etc. which are mostly used by employees to efficiently and effectively accomplish their workplace tasks. However, increasing the size of wireless communication devices constrains the scarcely available spectrum bandwidth. As a result, there exist bandwidth allocation inefficiency, high network latency, link congestion and high number of lost packets. Similar studies have attempted allocating bandwidth between the communicating users in uncooperative fashion, leading to inefficient bandwidth allocations and also not may not be suitable for a cooperative architecture of the SMEs considered in the current study. In order to enhance spectrum bandwidth allocation an integrated transferrable payoff coalitional game theory and standard-Dijkstra algorithm has been proposed in this study. This algorithm involves modelling of the spectrum bandwidth allocation problem in terms of an integrated transferrable payoff coalitional game and the least load shortest path model. The Dijkstra-Transferrable Payoff (DTP) algorithm is then proposed to solve the formulated model. Several computer simulations showed that the proposed algorithm provides reduced link congestion an end-to-end delay and a minimized number of lost packets at various packet sizes of the network injected traffic when compared to the standard-Dijkstra algorithm and Game-Theoretic Bandwidth Allocation (GTBA) algorithm.
Topside E. Mathonsi, Okuthe P. Kogeda and Thomas O. Olwal, 2017. Transferrable Payoff Based Bandwidth Allocation for Small and Medium Enterprises. Asian Journal of Information Technology, 16: 440-450.