Distributed Artificial Intelligence Based Cluster Head Power Allocation in Cognitive Radio Sensor Networks
Amrit Mukherjee, Pratik Goswami, Lixia Yang · IEEE Sensors Letters · 2019
This proposed work addresses the problem of dynamic and real-time power allocation to the cluster head (CH) in a cognitive radio sensor network (CRSN). The work is based on spectrum sensing outputs by these secondary users (SUs) for a nonergodic system. In general, a cooperative sensing network consists of multiple nodes communicating with each other about their respective spectrum sensing output, to perform the desired work. Therefore, the SU nodes of each cluster sense the spectrum all the time, which results in a continuous power consumption in CRSN. Now, the total amount of energy, which is used to allocate all the CHs at a particular place for spectrum sensing is being avoided and saved. This is performed, by the use of autocorrelation error for predicting the real-time behavior of primary users extended by the implementation of vector quantization to identify nearby active SU nodes and based on distributed artificial intelligence for power allocation. The simulation results and the mathematical derivation validate the proposed method.