A Channel Selection Method Based on Time Series Prediction in Cognitive Radio

Abstract

Disclosed is a channel selection method which is in cognitive radio and based on time series predictions. A cognitive user can equably divide an authorized frequency band into a plurality of sub-frequency-bands. Before service transmission is started, current spare sub-frequency-bands are perceived and acquired through frequency spectrum, and the history occupation probability of an authorized user of all sub-frequency-bands is inquired. The spare sub-frequency-bands are divided into a plurality of spare sub-frequency-band groups according to the number of the spare sub-frequency-bands and the positions of the spare sub-frequency-bands. As for each spare sub-frequency-band group, the occurrence probability of the authorized user inside corresponding sub-frequency-band groups and within a service continue duration of the cognitive user is predicted by using a time series prediction technology based on distance factors, and a sub-frequency-band group which has minimum occurrence probability of the authorized user is selected as a communication channel. The method takes history occupation information of the authorized user into account, achieves that a passive mode of channel selection of the cognitive user is prone to an active mode, effectively reduces the switching probability of the cognitive user in the service transmission process, and improves the handling capacity of the cognitive user.

Description

technical field [0001] The invention relates to a channel selection method suitable for cognitive radio, belonging to the technical field of communication. Background technique [0002] With the rapid development of wireless communication technology and wireless applications, the shortage of spectrum resources has become a bottleneck restricting the sustainable development of wireless communication and services. The traditional fixed radio spectrum allocation and access methods can no longer meet the development needs of wireless communication technology and services. On the one hand, a large number of frequency bands are allocated to authorized users with low active probability, resulting in a large amount of idle and waste of spectrum resources; On the one hand, the increasingly crowded spectrum environment makes a large number of wireless services and applications unable to be effectively applied and developed due to the shortage and limitation of frequency resources. Th...

AI Technical Summary

Problems solved by technology

In the existing cognitive radio channel selection methods, some methods adopt a "greedy" strategy, that is, when an available spectrum is detected, cognitive users occupy as much idle spectrum as possible for service transmission, which improves to a certain extent However, as far as the entire cognitive wireless network is concerned, such a behavior will make other cognitive terminals with service transmission needs unable to complete the normal service requests of users because they cannot detect idle spectrum.

Some methods use random access for channel selection. Although this method effectively avoids the problem of service congestion in the entire cognitive wireless network due to the existence of "greedy" users, it does not make full use of the licensed resources in the vacant licensed frequency band. The historical active information of the signal, its channel selection has a certain blindness

Moreover, the existing cognitive radio channel selection technology adopts a passive response method for the appearance of authorized users, that is, interrupts the current communication after detecting the authorized user signal, and selects a new idle frequency spectrum to continue business transmission. Fully consider the active law of the licensed signal in the authorized frequency band, so that it misses the best access frequency band, and it is very likely that the cognitive user will be interrupted during the communication process due to poor channel selection, reducing the throughput of the cognitive user

Images