Optimal Admission Control For Secondary Users using Blockchain Technology In Cognitive Radio Networks

The optimal admission control for secondary users (SUs) using blockchain technology in a cognitive radio (CR) network is investigated in this paper. When an SU comes to the system, it will decide to either enter the system, or give up its service and then leave the system based on the current number of PU and SU users in the network. SUs are linked as a block in the waiting area to access the spectrum in an opportunistic way without interfering the operations of PUs. When a PU arrives, if the band is occupied by SU(s), the SU service on the band is interrupted and SU(s) is then placed back to the waiting queue. Admitting an SU would bring a reward to the system provider but holding an SU in the system would also incur an expense to the provider. We concentrate on the optimization problem of when to admit or reject an arriving SU in order to achieve the maximum total discounted expected reward for any initial state. By establishing a discounted Continuous Time Markov Decision Process (CTMDP) model, we verify that the optimal policies for admitting SUs are state-related control limit policies. An extensive numerical analysis is implemented to show how to determine the value of optimal threshold and of the corresponding maximum objective function. Especially, a simulation by using the idea of the neural network and machine learning is implemented to show how to determine the optimal threshold value for any new input parameters through learning process for a given input parameters. The idea and the corresponding result of this paper can be applied in designing optimal cognitive radio networks.