Wireless Sensor Network Spectrum Sensing And Allocation Strategy Based On Sequential Decision

With the widely use of wireless devices in recent years, researchers have paid more and more attention to the use of wireless spectrum. The traditional method of using spectrum is that some international organizations allocate spectrum to certain users (as the primary user) and the rest of the users should share the public spectrum. But with the increasing of primary users, spectrum is going to be exhausted. As public spectrum become less and the public users’ increase, the communication quality of public users is decreasing. At the same time, the utility of authorized spectrum is low due to primary user cannot be using it all the time. Therefore, there are a lot of spectrum holes in environment.To improve this situation, researchers have brought the concept of cognitive radio, which means using software to control the access of the spectrum. With cognitive radio, public users (secondary users) could access authorized spectrum opportunistically when primary users are absent. In this way, the lack of spectrum resources could be alleviated and the utility of authorized spectrum could be improved. As cognitive radio involves many areas, there is still much work to do with it.In this paper, we firstly propose a novel spectrum sensing strategy. We consider cognitive radio (CR) system with a single secondary user (SU) and multiple licensed channels. The SU requests a fixed number of licensed channels and must sense the licensed channels one-by-one before transmission. By leveraging prediction based on correlation between the licensed channels, we propose a novel spectrum sensing strategy, namely, to decide which channel is the best choice to sense, to reduce the sensing time overhead and further improve the SU’s achievable throughput. Since the correlation coefficients between the licensed channels cannot be exactly in advance, the spectrum sensing strategy is designed based on the model-free reinforcement learning (RL). The experimental results show that the proposed spectrum sensing strategy based on reinforcement learning converges and outperforms random sensing strategy in terms of long-term statistics.Then I propose a novel spectrum allocating strategy. In traditional CR networks, the spectrum will be allocated to SUs randomly. Our novel spectrum allocating strategy combines the physical layer and network layer. In physical layer, I modeled the change of channels status as a non-stationary hidden Markov chain and predicted the channel quality with this model. Then in network layer, by analyzing the requirements generated by SUs in the current timeslot, the system formatted the current topology of CR network, then analyzed the nodes’importance in the topology graph via decision tree. Then allocate the higher quality channel to a more important channel to achieve the optimal allocation of spectrum resource. Simulation results shows that this optimal spectrum allocating strategy outperforms the random allocating strategy and could optimize the CR network’s throughput.Finally, I put forward some shortcomings of this paper and introduced some future works.