Design of likelihood ratio-based energy detectors considering stochastic analysis of non-identically distributed samples in cognitive radio networks

Abstract

The exponential growth of wireless data traffic urged the international wireless industries and academia to lay out a new wireless network standard which is 5G. It is generally believed that the future 5G network may need to be engineered to meet a multitude of stringent requirements in terms of cost, energy and spectral efficiency, number of connected devices and latency. Cognitive Radio (CR) is one of the promising techniques to meet these requirements by exploiting underutilized spectrum bands especially at microwave frequencies where there is a spectrum crunch. CR has applications ranging from the traditional cellular including the recent interesting LTE-U (LTE in Unlicensed band) technology to device-to-device (D2D) and vehicle-to-vehicle (V2V) communication systems via interweave, underlay and overlay dynamic spectrum access.

In this regard, spectrum sensing for primary user (PU) detection is a core concept of the CR networks for ensuring that CRs do not cause harmful interference to PU networks. This thesis basically considers mobile CRs, and for a wireless and mobile system design, it is very important to reflect the distinguishing features of mobile radio propagation. A wireless mobile channel is modeled as a time-varying communication path between two stations, such as from/to one terminal to/from another terminal.

Therefore, this thesis studies signal detection methods in cognitive radios based on both the well-known likelihood ratio test (LRT) and the sequential probability ratio test (SPRT) considering instantaneously non-identically distributed samples. Since, it is rather impractical to perform the ideal test that fully reflects such a change of a sample statistic, we used a practical detection method that optimally approximates the ideal log-likelihood ratio (LLR) with respect to its statistical mean.

In the LRT scenario, we derived closed form expressions for the detection threshold value and the sample size under given false alarm and missed detection constraints and we also found that there exists a nonnegligible performance gap between the ideal and practical tests, and the related experiments and theoretical analyses are also explored. Furthermore, we propose how to design a sequential detector in a robust manner to guarantee predefined false alarm and missed detection constraints for both the ideal and practical SPRTs.