Optimized Neural Network Using Differential Evolutionary and Swarm Intelligence Optimization Algorithms for RF Power Prediction in Cognitive Radio Network: A Comparative study

摘要

Cognitive radio (CR) technology has emerged asuda promising solution to many wireless communication problemsudincluding spectrum scarcity and underutilization. The a prioryudknowledge of Radio Frequency (RF) power (primary signals and/udor interfering signals plus noise) in the channels to be exploited byudCR is of paramount importance. This will enable the selection ofudchannel with less noise among idle (free) channels. ComputationaludIntelligence (CI) techniques can be applied to these scenarios toudpredict the required RF power in the available channels to achieveudoptimum Quality of Service (QoS). In this paper, we developed audtime domain based optimized Artificial Neural Network (ANN)udmodel for the prediction of real world RF power within the GSMud900, Very High Frequency (VHF) and Ultra High Frequencyud(UHF) TV bands. The application of the models produced wasudfound to increase the robustness of CR applications, specificallyudwhere the CR had no prior knowledge of the RF power relatedudparameters such as signal to noise ratio, bandwidth and bituderror rate. The models used, implemented a novel and innovativeudinitial weight optimization of the ANN’s through the use ofuddifferential evolutionary and swarm intelligence algorithms. Thisudwas found to enhance the accuracy and generalization of theudANN model. For this problem, DE/best/1/bin was found to yielduda better performance as compared with the other algorithmsudimplemented.