Dynamic spectrum allocation algorithm based on matching scheme for smart grid communication network

Abstract

Each part of smart grid system is supported by communications network. The amount of smart grid data is developing larger and more complex so that the network faces a lot of challenges to acquire stable and efficient communication in smart grid. The application of cognitive radio can effectively alleviate the shortage of spectrum resources in the smart grid communication network. In this paper, the spectrum allocation scheme of matching algorithm is proposed based on channel idle time and users' priority. Firstly, hidden Markov-model (HMM) is established to predict the idle time of channels in spectrum pool. Baum-Welch, which is the HMM training algorithm of the most commonly used, is formulated to train HMM parameter for each channel to get the most suitable HMM. Secondly, the data of smart grid is divided considering real-time priority, making the higher priority second user (SU) can occupy the channel with longer idle time. Finally, the system total throughput is calculated based on the two factors above. Simulation results show that the total throughput is improved effectively employing the matching algorithm and the stability and the reliability of communication network is obtained. The utilization of spectrum resources is also improved in smart grid.