Network Traffic Prediction Based on Firefly Algorithm Optimized Support Vector Machine

To meet the demands of intelligent network management in increasingly congested network environments, high-precision network traffic prediction is crucial. In order to enhance prediction accuracy, a network traffic prediction model incorporating Firefly Algorithm optimized Support Vector Machine (SVM) has been proposed. Firstly, simulation is performed based on the statistical properties of self-similar network traffic, and the Firefly Algorithm is employed to rapidly search for the optimal parameters of SVM. With these optimal parameters, the simulated network traffic is learned to derive an optimal network traffic prediction model. Finally, a comparative analysis is conducted between the optimized SVM-based network traffic prediction model and the non-optimized model. Experimental results demonstrate that this approach achieves higher prediction accuracy in network traffic forecasting while meeting the requirements of network management.