Spectrum Sensing using NMLMF algorithm in Cognitive Radio Networks for Health Care Monitoring Applications

Abstract

To use unlicensed version of cognitive radio spectrum without any interference spectrum sensing method is used. Energy detection is mostly used technique for sensing spectrum because it does not need any prior information of primary input. Spectrum sensing performance is good by using multiple antennas. Performance is low for noise uncertainty values at selecting threshold point and for low SNR values. Detection probability is improved by using dynamic threshold point, but there is noise present at the receiver output. To avoid this noisy output proposed an adaptive filter algorithm i.e., Maximum Normalized Least Mean Fourth (MNLMF) algorithm. By the proposed algorithm, noisy outputs are decreased, also variable step size considered in weight update equation of NMLMF for better estimation of error estimates. Then we get better probability detection performance for decreased false alarm probability and for various values of signal to noise ratio values. Performance is measured in terms of convergence, get the better results for sign regressor based algorithm when compared to other two variants of MN LMF algorithm.