Periodic signal recognition algorithm and quantization based on BP neural network

Jinlei He; Hanwen Ou; Yongquan Zhang

doi:10.1117/12.2656593

20 October 2022 Periodic signal recognition algorithm and quantization based on BP neural network

Jinlei He, Hanwen Ou, Yongquan Zhang

Proceedings Volume 12451, 5th International Conference on Computer Information Science and Application Technology (CISAT 2022); 124514E (2022) https://doi.org/10.1117/12.2656593
Event: 5th International Conference on Computer Information Science and Application Technology (CISAT 2022), 2022, Chongqing, China

Abstract

Aiming at the low accuracy of the traditional periodic signal recognition algorithm and the complexity of recognition circuit, this paper utilizes a lightweight back-propagation neural network (BPNN) model. The model achieves self-learning of the digital characteristics of four periodic signals, avoids the feature extraction under the traditional recognition algorithm, improves the accuracy of recognition , simplifies the structure of the circuit, reduces the cost and difficulty of hardware design. The trained neural network model has a recognition accuracy of 99% on the test dataset. By using 8-bit quantization algorithm inspired by the IEEE-754, the model capacity is reduced to 1/8 of the original, and the model recognition rate is reduced by only 1.3%.

Citation Download Citation

Jinlei He, Hanwen Ou, and Yongquan Zhang "Periodic signal recognition algorithm and quantization based on BP neural network", Proc. SPIE 12451, 5th International Conference on Computer Information Science and Application Technology (CISAT 2022), 124514E (20 October 2022); https://doi.org/10.1117/12.2656593

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available