A comparative analysis of local and global adaptive threshold estimation techniques for energy detection in cognitive radio

In this paper, we compare local and global adaptive threshold estimation techniques for energy detection in Cognitive Radio (CR). By this comparison, a sum-up synopsis is provided regarding the effective performance range and the operating conditions under which both classes best apply in CR. Representative methods from both classes were implemented and trained using synthesized signals to fine tune each algorithm's parameter values. Further tests were conducted using real-life signals acquired via a spectrum survey exercise and results were analyzed using the probability of detection and the probability of false alarm computed for each algorithm. It is observed that while local based methods may be adept at maintaining a low constant probability of false alarm, they however suffer a grossly low probability of detection over a wide variety of CR spectra. Consequently, we concluded that global adaptive threshold estimation techniques are more suitable for signal detection in CR than their local adaptive thresholding counterparts.