The topic of the paper is the distribution of tasks in mobile ad hoc networks (MANETs), where tasks are performed by the cooperating nodes. Node computing resources are limited by assumption, and selfish nodes are not interested in cooperation. To enforce node cooperation, the authors implemented an estimation system using the reputations of the nodes and the price for package transfer. Reputation ratings classify nodes as trustworthy and untrustworthy. Virtual money is used to measure the cost of packet-forwarding services. The game theory approach (see [1] for various applications of this methodology in computer science) is used to investigate the effectiveness of the system of incentives, compared to solutions without such motivation methods. The results are based on the simulation of the network equipped with the proposed reputation evaluation methods and information about the cost charged by the nodes for the transfer of packets. From this, the authors argue that a clever, rich, or selfish node can manipulate the operation of the network. Theoretical and simulation studies of the interaction of nodes show the superiority of the proposed integrated system in finding a solution using individual assessments of reputation and valuation of node services.
The game theory approach to the analysis of packet forwarding in MANETs is the subject of many research projects [2,3]. The main aim of existing models is to construct the strategy of individual nodes based on payoffs for service--their own and that of others. This paper tries to solve the challenge of encouraging the nodes to cooperate in order to maximize the overall system benefit.