Recognition of Channel Codes based on BiLSTM-CNN
Xingrong Huang, Shujun Sun, Xi Dang Yang, Shengliang Peng · 2022
Channel code recognition, which aims to recognize the channel code adopted by the received signal, plays an important role in the fields of non-cooperative communications. Deep learning based channel code recognition methods have been attracting great attention due to their superiority in learning from massive signals and extracting signal features automatically. However, these methods mainly use a single type of neural network and suffer from low recognition accuracy. In this paper, we propose a channel code recognition algorithm based on two types of neural networks including bi-directional long short-term memory (BiLSTM) and convolutional neural network (CNN). According to the proposed algorithm, the received signal is firstly fed into BiLSTM and then handled by CNN, which inherits the advantages of both BiLSTM and CNN. Experimental results show that the proposed algorithm outperforms the existing TextCNN based algorithm, and the improvement of average recognition accuracy is about 4% at the low signal to noise ratio region.