Cooperation Of Selfish Nodes In The Ad Hoc Networks

Mobile Ad hoc Networks (MANETs) consist of a number of heterogeneous mobile nodes formed at anytime without using any pre-existing infrastructure. The network has gained rapid progress in business and military area due to characteristics of self-organization, survivability and robustness. End-users can be provided by more flexible information services through cooperation between users. Cooperation plays an important role in the area of Ad hoc networks. While users have limited resources and lower capacity of execution, some selfish nodes aren’t willing to consume their own resources to provide a service for another (e.g. packets forwarding, information share). Previous studies have demonstrate that the performance of network throughput decreased by 16% to 32% when the network accounted for 10% to 40% of selfish nodes.Based on this, the paper proposes the cooperation mechanism of selfish nodes in the MANETs. The main research contents and innovation points are depicted as follows:(1) The resources and applications are regarded as services, thus selfish nodes are stimulated by bargaining services. When each seller dependently provides a service for a buyer with plentiful resources, they use the service-driven negotiation model proposed in our paper to gain their maximal profits. In addition, we optimize the negotiation tactics for negotiators to generate offers in the next rounds and increase the probability of agreement.(2) In the negotiation condition, we further consider the factors such as initial price, reserve price and deadline have influences on agreement. In the service pricing models, the initial price should reflect the resources states in the beginning of negotiation, and reserve price can be recalculated with the advance of negotiation process to avoid invalid agreement. Since the sudden user-applications occupy the computational and energy resources which impact on deadline, we design the adjustment mechanism for deadline in order to provide reference for offering. (3) Ad hoc networks could interconnect with Internet or mobile cellular network. In this condition, the management center needs to detect, control and distribute tasks in the Ad hoc networks. Multiple nodes cooperate to form task coalitions to increase the completion ratio of the tasks. Through the improvement fuzzy c-means algorithm, we cluster the nodes attributes (such as battery, computational capability, mobility, fault rate) to reduce the searching space of coalition formation. In the voting process for coalition leader, we consider the factors of communication delay, resource states, and mobility to vote a stable leader to avoid frequent coalition re-organization. Based on the above-mentioned mechanisms, we design a coalition formation algorithm that ensures task redundancy and the system robustness.(4) Selfish nodes belong to different individuals who generate different emotions respond to each cooperation invitation according to their preferences, current resource states, emotion history and network environment. The cooperation motivation model is constructed for mobile devices to imitate human emotions for pursuing cooperation success. The behaviors of selfish nodes are triggered by cooperation motivation, thus the nodes reduce the energy consumption and network traffic.