Analytical performance of soft clustering protocols

The success of a mobile ad hoc network (MANET) is strongly related to the protocol used at the medium access control (MAC) layer. Depending on the requirements and the specific network under concern, the protocol parameters at the MAC layer can be arbitrated to make best use of the channel resources. Typically, extensive simulation studies are used to find the best values for these variables. The problem with this approach is the need for excessive amounts of processing power and time. As the dimensions of the decision space increase, the need for processing power grows exponentially. This paper addresses this problem by developing an analytical model that reflects the relationships between protocol parameters and the overall performance of the protocol under various network conditions. Specifically, we model the MH-TRACE cluster-based protocol, which is capable of supporting real-time data transmission. The model is capable of estimating performance measures such as energy consumption and number of receptions while being simple enough to be run for a large set of parameters. The model can be used to optimize parameters of the protocol (such as the number of frames per superframe) as well as to predict the performance variations as the external conditions (such as data generation rate) vary.