Blind Sensing In Cognitive Radio

We are now faced by an era marked by an exponential growth in technological advancement that use radio frequency spectrum in communication, transport, medicine, and security. This has resulted to wireless communication shifting to more active internet data interaction.In the near future we will have more vehicles endowed with the capability to’talk’to each other in order to improve on road safety. RF transponders will in the near future too be massively deployed in a tiny chip, implanted in human beings, to improve health care efficacy. Affordable smart phones are now widespread, even in areas that were initially considered as remote. All these advancements have and will in one way or the other exert immense constrain on the naturally occurring RF spectrum resource. However advancement in cognitive radio technology is aware of the challenges it faces. Many detectors have been developed in the last decade to improve on the simple energy detector in the blind scenario case. More and more efficient algorithms are being developed to match up with technological trends.This work focused on the energy detector (ED) for its simplicity. Two types of EDs were simulated in MATLAB(?) R2012a software in order to study the effects of manipulating the test statistics coefficients (TS) on their performance notably ED1and ED2, simple energy detector and improved energy detector respectively. Their performance was investigated under the following parameters:probability of detection, probability of false alarm, variance, frequency, SNR and threshold. For simplicity the noise variance was set to one and a preferred bandwidth of1kHz for the primary signal was simulated.In order to further improve detection probability (≥0.9) at low signal to noise ratio while maintaining the probability of false alarm as low as possible (≤0.1) we derived Improved Test Statistic based on the coefficients of the test statics of the two EDs. Similarities between the test statistics of ED1and ED2was the salient think tank of our novel approach. Provided that the test statistic manipulation doesn’t affect the probabilities of detection and false alarm as well as the threshold equations of the two detectors, we derived the new test statistic namely Improved Test Statistic (ITS).The new (ITS) was then applied in both EDI and ED2in order to generate the Improved Test Statistic energy detector EDI (ITS) and ED2(ITS) respectively. The performance of ED1(ITS) and ED2(ITS) was similarly investigated under the same parameters as in ED1and ED2.Simulations showed that our improved test statistic energy detectors far outperformed their predecessors under the parameters set in the scope of our study. The performance of the (ITS) coefficients in both ED1and ED2was also investigated with a total of eight (ITS) coefficients investigated. The simulations were done using real synthetic PU signals with additive white real Gaussian noise in a MATLAB environment and1000Monte Carlo simulations. The Simulations were analyzed using ROC curves as well as probability versus SNR curves.