An Optimal Auction-Based Routing Scheme for Detecting and Managing Selfish Nodes in Delay-Tolerant Networks

Abstract

Nodes in Delay-Tolerant Networks (DTNs) and other opportunistic networks work collaboratively to ensure efficient message forwarding. The decision of selfish nodes to neither forward nor discard incoming messages had a detrimental effect on the performance of such networks. The focus of this paper was to propose a novel routing scheme for DTNs. This scheme incorporated an optimal auction mechanism which played a crucial role in detecting and managing selfish nodes, ultimately leading to improved efficiency and reliability of the network. Existing detection methods, which encompass neighbor monitoring, acknowledgment, autonomy, reputation, and credit-based approaches, have been found to suffer from drawbacks such as network overhead, trust concerns, and the issue of unreliable data. Within the framework of the proposed auction model, the source node initiated the process by broadcasting a message to its neighboring nodes. This message contained a predefined value, prompting the neighbors to formulate bid strategies that took into account their competence and the costs they would incur for transmitting the bids. The source node made a selection of relay nodes in order to enhance the performance of the network, ensure honesty, and deliver fair rewards. The implementation of this strategy-proof mechanism not only encouraged participants to behave honestly but also effectively deterred any malicious actions and significantly enhanced the overall reliability of the system. Through simulations, it was shown that this method significantly enhanced network performance, detection accuracy, and incentive fairness when compared to the baseline method, ProPhet. In terms of the proposed algorithm, the median delivery ratios varied between approximately 0.6 and 0.7. The median delivery ratios of the ProPhet algorithm ranged from around 0.35 to 0.75.