Normalization method for online learning on radio access technology identification in cognitive radio

Asakura, Kenta; Ichikawa, Haruhisa; Kawakita, Yuusuke

doi:10.23919/apcc.2017.8304025

Cited by 1 publication

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we present the method proposed in [10] for normalizing radio features for online learning to identify RATs that exist in the same frequency band. Through sorting, dividing, and digit aligning as shown in Fig.…”

Section: Radio Feature Normalization Methodsmentioning

confidence: 99%

“…In the present paper, we evaluate our previously proposed method [10] for normalizing signal features for online learning to improve RAT identification accuracy. The proposed scheme involves implementing a RAT identification system that uses online learning methods to adapt easily to RAT diversification.…”

Section: Introductionmentioning

confidence: 99%

“…However, online learning is less accurate than batch learning. We studied RAT identification in the sub-gigahertz bands and used online learning to improve the accuracy of identifying RATs with similar feature [10]. Additionally, we resolved the stability-plasticity dilemma [11] caused by using online learning and the difficulty in identifying RATs that exist in the same frequency bands [10].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Normalization of Radio Feature for Online Learning for Identification in Cognitive Radio

Asakura

Ichikawa

Kawakita

2019

SICE Journal of Control, Measurement, and System Integration

Self Cite

View full text Add to dashboard Cite

The identification of the radio access technology (RAT) of the primary user by secondary users is important to avoid interference for spectrum-sharing techniques using cognitive radio (CR). RATs have become more diversified with the introduction of various services that use wireless communication. Therefore, it is desirable that a RAT identification system, which can easily cope with the diversification of RATs, is applied to CR. The purpose of this study is to use online learning to identify multiple RATs in the same frequency band. To improve the accuracy of identifying RATs with similar features, we evaluate a normalization of radio feature method proposed in our previous research for the features extracted from the signal's spectrogram. We evaluate the RAT classifier created using the proposed method by calculating the curve of receiver operating characteristics (ROC) and the identification accuracy. The results for the ROC curve show that the proposed method is effective for several supervised learning methods. Moreover, the results for the identification accuracy show that the proposed method improves the identification performance compared to the identification accuracy of conventional methods.

show abstract

Section: Radio Feature Normalization Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%