Fig. 4
From: A novel cooperative spectrum signal detection algorithm for underwater communication system
The curve of the relationship between observed values and detection probability. This result in Fig. 4 indicates that the spectrum sensing performance of the MP algorithm and the OMP algorithm is gradually improved when the number of observation points of the compressed sensing is increased. This is because that the more sampled values, the more information is obtained. Besides, the MP algorithm and the OMP algorithm are algorithms for a single task, regardless of the spatial diversity information, and the only information available is from the compressed sensing observations. Compared with the MP and OMP algorithms, the proposed algorithm has no change in the ROC curve when the number of observation is greater than 20 and the detection probability approaches 1. Because it makes full use of the compressed sensing data information collected by each SU and the number of iterations of the algorithm is not related to the number of samples. Even at lower sample values, it can still obtain batter detection performance and effectively reduce the system overhead of spectrum sensing. From the ROC of the three algorithms and considering the lower number of observation, it is obvious that the correct detection probability of the paper algorithm is better than the MP and OMP algorithms, and the spectrum sensing performance is better