Contribuições para o método CPSC em redes de rádios cognitivos

Abstract

Recently,anewtechniqueforspectrumsensinginCognitiveRadio(CR)networkshas been proposed named Cooperative Power Spectral density split Cancellation (CPSC). Thistechniquehastheadvantagesofbeingrobustagainstnoiseuncertaintyandhaving low computational complexity. In this dissertation, a new sensing technique was proposed from a simple modification on the CPSC method. Preliminary results showed that the new technique presents better performance than the original one. Also, two CumulativeDistributionFunctions(CDFs)oftwoimportantrandomvariablesforboth proposed and original methods were derived. For each random variable, two solutions with different approaches were presented. The results showed the level of accuracy of the developed CDFs, specially the one from the second approach, in which it is even moreaccuratethantheonepresentedinthepaperoftheoriginalCPSCmethod. Other contribution is the development of two approximate expressions for the Probability of False Alarm (PFA) for the new method. The results showed that one of the solutions happened to be accurate for high values of PFA, while the other for low levels of PFA. Theconclusionisthatthesecondone,whichisaccurateforlowvaluesofPFA,ismore appropriate to be used in practice as the spectrum sensing should work with low PFA. It is worth mentioning that the derived expressions are very useful in practice because theyallowonetoanalyzestatisticallythesystembehavioroftheoriginalandproposed methods.