Hybrid Optimized Secure Cooperative Spectrum Sensing for Cognitive Radio Networks

Abstract

Spectrum utilization is an important factor in Cognitive radio networks, which is accomplished by analyzing the unused spectrum bands of primary users (PU). The secondary users are allowed to access the resources, when the spectral bands are vacant by sensing the spectrum status and thus it reduces the spectrum scarcity among the users. Researchers have paid more attention towards spectrum sensing along with its security factors in cognitive radio networks. In this process, cooperative spectrum sensing is widely adopted in cognitive radio networks due to its robustness. However, the security concerns in cooperative spectrum sensing against attacks must be addressed. The performance of cooperative spectrum sensing will get affected if the fusion center gets wrong information from malicious user. This leads to wrong decision in the fusion center and results into false observations and affects the decision process. In order to overcome these challenges, this research work proposes a hybrid nature inspired and optimized cooperative spectrum sensing against attacks in cognitive radio networks. The proposed model allows the fusion center to remove the uncharacteristic data in the fusion process, which results from the malicious users. The performance analysis of spectrum sensing process under different attacks are analyzed through simulation and later it is compared against conventional methods such as genetic algorithm, particle swarm optimization and differential evolution schemes to validate the improved performance.

Regional autonomous security cooperative spectrum sensing method based on trust value

“…Although the advantages of CSS in spatial diversity mitigate the impact of hidden terminals, the presence of MU still greatly reduces the sensing performance [7,8]. This paper uses the trust value update algorithm of regional autonomy, and the sensing flow chart is shown in Figure 2.…”

“…The CSS can be classified into centralized, distributed, and relay based on the interaction modes between the users. However, CSS also brings some new security issues, such as PU emulation attacks [6], spectrum sensing data falsification (SSDF) attacks [7,8], channel blocking attacks [9] etc., which greatly reduce the performance of CSS.…”

“…By authenticating users, malicious users (MU) are eliminated and legitimate SUs are perceived. In order to solve the problem of multi-band CSS, Rangaraj et al [14] proposed spectrum sensing based on the ant colony optimization algorithm, which removes non-characteristic data generated by MUs during the fusion process, which has better performance in attack scenarios. Recently, with the boom of machine learning, more and more CSS research is being done using machine learning.…”

Less sample‐cooperative spectrum sensing against large‐scale Byzantine attack in cognitive wireless sensor networks

“…A hierarchical cat and mouse learning machine was proposed to minimize Byzantine attack in reference 19. In reference 20, an ant colony optimization and differential evolution model was used to secure CSS against attack in CRNs. A support vector machine was used to learn the malicious behavior and identify malicious users in references 21,22.…”