Intelligent spectrum cooperative sensing method based on reinforcement learning

Abstract

The invention discloses an intelligent spectrum cooperative sensing method based on reinforcement learning, which is applied to the field of spectrum sensing. According to the method of the present invention, for each secondary user SUk, when a calling request is reached, a decision of sensing a channel is made to minimize the scanning overhead according to a channel priority list obtained by learning the channel state closest to the secondary users based on a Q-learning technology; secondly, the node can request other SUs to perform cooperative spectrum sensing to improve the detection probability when detecting the channel and selects the secondary user with strong detection capability to cooperate to improve the detection probability through a bandit gambling mechanism. Once detection of the channel is completed, the state of the detected channel and the detection weight are broadcasted to notify other secondary users. The method of the invention effectively reduces the average number of scans of the channel and reduces the call blocking rate, thereby improving the detection probability.

Description

technical field [0001] The invention belongs to the technical field of spectrum sensing in wireless mobile communication networks, and in particular relates to a cooperative spectrum sensing method in wireless mobile communication networks. Background technique [0002] With the rapid development of the information industry, especially the wireless mobile communication industry, the demand for wireless spectrum resources has increased sharply, and the contradiction between the scarcity of spectrum resources and the inefficiency of existing fixed radio spectrum resource allocation strategies has become increasingly obvious. Spectrum utilization has become the core problem to be solved urgently in wireless communication. Facing the severe challenge of the scarcity of wireless communication spectrum resources, changing the traditional fixed resource allocation to dynamic resource allocation, so as to effectively improve the utilization of spectrum resources has been more and mo...

AI Technical Summary

Problems solved by technology

[0006] In a cognitive radio network, since the channel occupancy of the PU is unknown to the SU, the SU needs to detect the occupancy of each channel by the PU successively, so the above traditional algorithms will cause excessive scanning overhead and delay. Appropriate selection of the sensing and access sequence of the authorized channel plays a crucial role in determining the average delay and effective throughput; secondly, because the detection ability of the node is easily affected by fading and shadow effects, the detection ability of the node is dynamically changed, How to select appropriate sub-users for cooperation in the case of dynamic changes in node detection capabilities is also a problem worthy of attention

Images