An Active and Passive Reputation Method for Secure Wideband Spectrum Sensing Based on Blockchain

Xinyu Xie,Yaochi Zhao,Min Chen,Zhuhua Hu,Yong Bai

doi:10.3390/electronics10111346

Abstract

Spectrum is a kind of non-reproducible scarce strategic resource. A secure wideband spectrum sensing technology provides the possibility for the next generation of ultra-dense, ultra-large-capacity communications to realize the shared utilization of spectrum resources. However, for the open collaborative sensing in cognitive radio networks, the collusion attacks of malicious users greatly affect the accuracy of the sensing results and the security of the entire network. To address this problem, this paper proposes a weighted fusion decision algorithm by using the blockchain technology. The proposed algorithm divides the single-node reputation into active reputation and passive reputation. Through the proposed token threshold concept, the active reputation is set to increase the malicious cost of the node; the passive reputation of the node is determined according to the historical data and recent performance of the blockchain. The final node weight is obtained by considering both kinds of reputation. The proposed scheme can build a trust-free platform for the cognitive radio collaborative networks. Compared with the traditional equal-gain combination algorithm and the centralized sensing algorithm based on the beta reputation system, the simulation results show that the proposed algorithm can obtain reliable sensing results with a lower number of assistants and sampling rate, and can effectively resist malicious users’ collusion attacks. Therefore, the security and the accuracy of cooperative spectrum sensing can be significantly improved in cognitive radio networks.

Highlights

In cognitive radio (CR) networks, spectrum sharing strategy based on dynamic access provides a feasible scheme for efficient spectrum utilization [1,2,3,4]
For the correction value of recent credibility, the nearest accuracy judgment value of L = 4 times is taken into account, and the memory weight is respectively taken as θn−1 = 0.4, θn−2 = 0.3, θn−3 = 0.2, θn−4 = 0.1, and the decision parameter is β = 0.5
In the Collaborative spectrum sensing (CSS) network, to solve the problem that the sensing performance is significantly reduced when sensing data falsification (SSDF) attacks were launched by malicious nodes, a blockchain-based APR_SWSS algorithm was proposed

Summary

Introduction

In cognitive radio (CR) networks, spectrum sharing strategy based on dynamic access provides a feasible scheme for efficient spectrum utilization [1,2,3,4]. This technology allows the secondary user (SU) to access the idle spectrum when the signal of the primary user (PU) is not detected. When users provide false sensing results maliciously, the performance and security of spectrum sensing will be significantly affected [13] Zhang Linyuan et al discuss various situations of Byzantine attack and defense in reputation-based cooperative spectrum sensing [14]. How to reduce the impact of malicious users (MU) uploading false sensing data and how to ensure the accuracy and security of spectrum sensing are hot issues worthy of research [15,16]

Results

Discussion

Conclusion