Recognition relay network trust management device and method based on dynamic evolution

Abstract

The invention provides a recognition relay network trust management device and method based on dynamic evolution. The recognition relay network trust management method comprises a trust perception phase, a trust evaluation phase, a trust evolution phase and a stimulation optimizing phase. In the trust perception phase, node behavior information and user custom requirements are collected, and distinguish contribution degrees of the nodes are described as direct trust perspectives on the nodes. The trust evolution phase is interacted with corresponding phases of other nodes belonging to a perception relay network so as to integrate computional node trust. In the trust evolution phase, node behavior multi-strategy dynamic evolution game is performed, and a node behavior strategy is formulated. In the stimulation optimizing phase, node behaviors are optimized and adjusted, and the nodes are encouraged to obtain better and superior services by improving self trust. The recognition relay network trust management device and method can process conditions of united over-layer attacks and the like in the recognition relay network, achieves fast and stable optimization strategy resolving and effectively restrains selfish and vicious behaviors of the nodes.

Description

technical field [0001] The invention relates to a trust management method in a cognitive relay network environment. Specifically, it is a cognitive relay network trust management method that simulates the interactive behavior between network nodes based on dynamic evolutionary games, and promotes the evolution of node groups in the direction of trust and cooperation through cooperative and competitive relationships. Background technique [0002] Cognitive relay networks can be deployed in a centralized, distributed structure or a mixture of the two, which is different from general multi-hop wireless networks. In addition to the basic characteristics of cognitive loops, cognitive relay networks also have cooperative Features such as perception and relay communication. A basic cognitive loop includes six links of perception, judgment, planning, reasoning, decision-making, and execution. The cognitive node judges the usage of the primary user channel according to channel perce...

AI Technical Summary

Problems solved by technology

[0003] The research on trust management in the fields of overlay network and peer-to-peer network is in the ascendant, and has accumulated rich research results, but it cannot be directly applied in the cognitive network environment. The main reasons are: the dynamic network topology determines any centralized trust None of the solutions is suitable, let alone a third-party trust mechanism, such as PKI; limited storage capacity, computing power, and communication capacity make the current trust model that relies on complex calculations lack engineering feasibility; interference with the limited operating mechanism causes the existing Trust detection is prone to misjudgment of selfishness and cooperation of nodes, and will cause major deviations or even subversion of cognitive decision-making; asynchronous communication and timeliness characteristics make traditional trust evaluation and incentive mechanisms effective in dealing with dishonest recommendations, collaborative cheating and some Limited capabilities in complex strategic attacks

Bhattacharjee et al. proposed a trust-based distributed perception fusion scheme, which evaluates the trust degree of neighbor nodes and uses heuristic thresholds to filter malicious nodes. However, this scheme is mainly aimed at Byzantine attacks and requires each node to have multiple neighbor nodes. And cannot prevent collusion (Bhattacharjee S, Debroy S, Chatterjee M. Trust computation through anomaly monitoring in distributed cognitive radio networks [C]. Proceedings of 2011IEEE22nd International Symposium on Personal Indoor and Mobile Radio Communications (PIMRC), 2011: 593-59 )

Yu et al. proposed a scheme based on evidence nodes to prevent collusion attacks. The evidence nodes are clustered by measuring the credibility of neighbor nodes to assist cognitive nodes in selecting trusted nodes. However, this method is only suitable for single-hop authentication nodes. Cognitive network environment (Yu H, Liu S, Kot A C, Miao C, Leung C. Dynamic witness selection for trustworthy distributed cooperative sensing in cognitive radio networks [C]. Proceedings of 2011IEEE13th International Conference on Communication Technology (ICCT), 2011: 1- 6.)

[0004] To sum up, the existing trust management methods focus on the situation where single-hop nodes participate in cooperation, and there is still a lack of corresponding solutions for the cooperation between multi-hop nodes based on the network layer. Comprehensive consideration of dynamics, node mutation, etc., it is difficult to balance contradictory metrics such as cognitive ability, communication load, node behavior, etc., and cannot always effectively deal with threat evolution, joint cross-layer attacks, etc.

Images