|
This paper considers the power allocation for cooperative sensing when the system's statistical properties cannot be deduced for any given real-time scenario. For such type of systems, the power allocation varies and the standard optimization techniques fail. Therefore, a simple and robust method based on Euclidean Distance (ED ) is implemented which measures the Autocorrelation Error (AE) of the Power spectra of the channel at different instants. Here, ED refers to the shortest path between the two active nodes or users. The proposed method is analyzed and simulated for 10 channels keeping ED constant. Blind Source Separation (BSS) is primarily used to identify dominating frequency components in the corresponding channels followed by the evaluation of AE. The results illustrate the active nodes along with the importance of identifying these nodes for future prediction which will help in 5G Communications using Cognitive Radio (CR) network. Further, it can be inferred that during spectrum sensing there is a need of allocating power to the specific nodes only, rather than all the nodes, which will use the power consumption by the secondary users efficiently. The probability of detection and probability of false alarm is set to QoS degradation by the Primary User (PU).
|
