Spectrum prediction switching method based on transfer learning strategy

Abstract

The invention provides a spectrum prediction switching method based on the transfer learning strategy. The method comprises the following steps: dividing a frequency spectrum switching range of a secondary user SU into N + 1 PU channels which are not overlapped with each other, and sequencing the N PU channels which are not occupied by the secondary user SU from small to large according to a central frequency, phi i being an ith PU channel, and i being a positive integer; dividing each time slot of the HCRNs system into the following sub-steps: dividing each time slot of the HCRNs system intosub-time slots and sub-time slots; the method comprises the following steps of: selecting M idle channels capable of enabling a secondary user SU to be subjected to frequency spectrum switching accessas the ith idle channel from N PU channels by a base station SBS corresponding to the secondary user SU through frequency spectrum prediction sensing in a frequency spectrum sensing TS time period according to a frequency spectrum sensing TS, a switching acknowledgement TACK and a transmission stage TD; and detecting whether the secondary user SU receives an acknowledgement signal indicating thatthe spectrum switching is completed within the switching acknowledgement TACK time interval, and ending the spectrum switching process when the acknowledgement signal is detected. According to the invention, in the HCRNs time slot, the switching confirmation TACK link is added, so that the robustness of frequency spectrum switching is ensured.

Description

technical field [0001] The present invention relates to the field of cognitive radio technology, in particular to a spectrum prediction switching method based on a transfer learning strategy. Background technique [0002] Cognitive Radio Networks (CRNs) can effectively solve the contradiction between the shortage of wireless spectrum resources and the low utilization rate of wireless spectrum through dynamic spectrum access DSA (Dynamic spectrum access) and spectrum resource management technology, and improve the communication of the system. capacity. Hybrid cognitive radio network HCRNs (Hybrid Cognitive Radio Networks) is a mode that unifies Interweave and Underlay spectrum access. Because it has higher spectrum utilization than single-mode spectrum access technology, research on HCRNs technology more theoretical and practical. [0003] However, the HCRNs system faces many technical challenges, one of which is the spectrum switching technology. According to whether the ...

AI Technical Summary

Problems solved by technology

Although this method can minimize the number of spectrum switching times, since the spectrum switching time of this method is fixed at K time slots, it cannot reduce the spectrum switching time, and does not consider the more important system throughput in spectrum switching

[0005] The above spectrum switching method based on prediction and judgment still has the following problems: 1. Only consider the situation of data transmission between a pair of sending and receiving SUs, without considering the influence of surrounding SUs; 2. Only consider the single spectrum access mode CRNs spectrum handover scenario, without considering the hybrid spectrum access HCRNs application scenario with higher spectrum utilization and closer to the actual situation, and the multi-SU spectrum handover problem in this scenario; 3. The success rate or failure of SU handover is not considered rate for analysis

Images