Using high level information to make services and applications sparse MANETs aware

Context awareness in wireless networks with mobile nodes will enable services or applications to adapt their behavior to the network situation by reconfiguring their underlying protocols. We studied whether high-level information, captured in the semantics of the events during emergency and rescue scenario might be exploited to improve network performance by mapping alternative solutions in dependence on the space-time graph specifics. ldquoBinary behavioral templaterdquo - a new functional feature for estimating the network context is suggested. Fuzzy C-mean clustering and fuzzy reasoning for predicting the future state of space-time paths that services and applications will use for reconfiguring their underlying protocols, are proposed. RoboCup rescue scenario is used to evaluate the feasibility of the proposed fuzzy model.