Adaptive bandwidth allocation scheme for wireless networks

Backoff manipulation is a MAC layer misbehavior that can degrade the performance of wireless networks. Existing misbehavior handling strategies operate on the basis that all nodes in the network generate similar type of CBR traffic. This does not reflect the true nature of wireless networks that may consist of multiple traffic types. In this paper, a scheme that enhances the existing misbehavior handling strategy is proposed. It correlates misbehavior with the traffic characteristics generated by the node deviating from the stipulated protocol. It eliminates the stereotyping of backoff deviation by deriving a multi-scale deviation definition ranging from misbehavior to underestimated allocation or resource. The scale adaptively enables varying types of misbehavior according to resource requirement. From the performance analysis, it is determined that the scheme is successful in differentiating between misbehavior due to selfishness and under-allocation of resource. In addition, the scheme also enables nodes that are allocated with inadequate resource to achieve higher share of throughput. The performance analysis of the proposed scheme was conducted with the use of a discrete-event simulator. From the analysis, it is concluded that with the employment of the proposed scheme, if a node deviates from the protocol and it is allocated with inadequate resources, the scheme will adaptively allow the node to gain the resource as per the degree of its resource under-allocation and hence, enabling for the increase of overall network throughput.