Blind Detection of Channel Coding and Interleaving Using Convolutional Neural Networks in Tactical Communications
Seok–Jin Hong, Woong–Jong Yun, Eui–Rim Jeong · 2024
This paper proposes two Convolutional Neural Network (CNN) models for classifying types of channel codings and interleavings blindly in military communications. In tactical communications, eavesdropping of enemy's signal requires blind recognition of the channel coding and interleaving methods. In this study, we introduce a channel coding estimation CNN model that classifies between two widely used convolutional error correction codes and one Reed-Solomon error correction code in military communications. Additionally, our interleaver classifier distinguishes between data processed with block interleavers and data without interleaving. The performance of the proposed CNN models is evaluated through computer simulations and compared with well-known CNN models, VGG and ResNet. The proposed models demonstrate comparable performance with significantly fewer parameters, highlighting their potential for efficient implementation in military communication systems.