Method and device for predicting spectrum occupancy state based on neural network

Abstract

The invention discloses a method and a device for predicting a spectrum occupancy state based on a neural network. The method for predicting the spectrum occupancy state based on the neural network comprises the steps of S1: constructing the neural network and using a pruning algorithm to determine initial parameters of all layers and all neurons of the neural network; S2: inputting all elements in a set (bi, bi-1,..., bi-m) into the neural network, comparing an output result of the neural network with a representation variable hi+1 which represents the occupancy state of the preset frequency band of an (i+1) timing sequence, and correcting initial parameters to obtain predictive parameters according to the comparison result, wherein bi is the signal intensity of the preset frequency band of an i timing sequence, i is a positive integer, and m is a positive integer smaller than i; and S3: inputting all elements in a set (bi+1, bi,..., bi-m+1) into the neural network simultaneously, outputting a predictive representation variable Hi+2 which represents the occupancy state of the preset frequency band of an (i+2) timing sequence by the neural network, and judging the occupancy state of the preset frequency band of the (i+2) timing sequence according to the predictive representation variable Hi+2. The method has the advantages of being low in computation complexity and easy to implement.

Description

technical field [0001] The present invention relates to the field of communication technologies, in particular to a method and device for predicting spectrum occupancy status. Background technique [0002] In the process of dynamic spectrum allocation, the secondary user needs to temporarily use the idle spectrum of the primary user to communicate in order to achieve the purpose of improving spectrum utilization. The premise of communication based on the foregoing dynamic spectrum allocation is that the spectrum occupancy state can be accurately predicted, and an idle spectrum is found for communication of secondary users. Existing spectrum occupancy state prediction methods such as Kalman filter model, hidden Markov model and other prediction methods are all based on current and / or historical time series spectrum occupancy state and spectrum distribution information. However, in the specific implementation process, it is difficult for the secondary user end to obtain spect...

AI Technical Summary

Problems solved by technology

However, in the specific implementation process, it is difficult for the secondary user to obtain the spectrum occupancy distribution information, and because of the spectrum occupancy distribution, there are problems such as large amount of calculation, high computational complexity, and inaccurate prediction results.

Images