Learning for Robust Routing Based on Stochastic Game in Cognitive Radio Networks

TL;DR

This paper introduces a stochastic game-based routing scheme for cognitive radio networks that enhances robustness against malicious nodes by learning optimal policies and evaluating node trustworthiness.

Contribution

It proposes a novel stochastic game model with smooth fictitious play and multi-arm bandit trust evaluation to improve secure routing in CRNs with malicious threats.

Findings

The routing scheme effectively enforces cooperation among malicious SUs.

It reduces the impact of insider attacks on routing decisions.

Simulation results confirm improved robustness and security.

Abstract

This paper studies the problem of robust spectrum-aware routing in a multi-hop, multi-channel Cognitive Radio Network (CRN) with the presence of malicious nodes in the secondary network. The proposed routing scheme models the interaction among the Secondary Users (SUs) as a stochastic game. By allowing the backward propagation of the path utility information from the next-hop nodes, the stochastic routing game is decomposed into a series of stage games. The best-response policies are learned through the process of smooth fictitious play, which is guaranteed to converge without flooding of the information about the local utilities and behaviors. To address the problem of mixed insider attacks with both routing-toward-primary and sink-hole attacks, the trustworthiness of the neighbor nodes is evaluated through a multi-arm bandit process for each SU. The simulation results show that the proposed routing algorithm is able to enforce the cooperation of the malicious SUs and reduce the negative impact of the attacks on the routing selection process.