Cross-Validation-Based SVM for Cooperative Spectrum Sensing Against Probabilistic Byzantine Attack in Cognitive Wireless Sensor Networks

Cognitive radio (CR) enables multiple sensor nodes (SNs) to detect whether the primary node (PN) occupies spectrum bands through cooperative spectrum sensing (CSS), allowing access to unused channels. However, the openness of CSS makes it vulnerable to Byzantine attacks by malicious sensor nodes (MSNs), leading to performance degradation. To mitigate this, we propose a cross-validation-based support vector machine (CVSVM) approach for Byzantine attack resistance in cognitive wireless sensor networks (CWSNs). Cross-validation enhances generalization and reduces performance fluctuations caused by varying sample partitions. The dataset undergoes multiple random stratified splits, with <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{k}$</tex>-fold cross-validation applied each time, and the final evaluation is averaged across iterations. A blacklist management system is generated by comparing classification results with a predefined threshold. Simulations show that CV-SVM outperforms other methods under identical conditions.