基于神经网络的SD-Bi-LSTM差分跳频智能译码算法
Differential Frequency-Hopping Intelligent Decoding Algorithm Based on SD-Bi-LSTM Neural Network

为解决差分跳频译码算法复杂度高、纠错能力差的问题，在深入分析差分跳频机理和传统译码方式的基础上对差分跳频序列特性进行分析，基于神经网络提出SD-Bi-LSTM差分跳频智能译码算法，提高差分跳频译码准确性与纠错能力。仿真结果表明，提出算法的译码准确率和纠错能力在不同频率序列长度均优于传统算法。
In order to solve the problems of high complexity and poor error correction ability of differential frequency hopping decoding algorithm, based on in-depth analysis of differential frequency hopping mechanism and traditional decoding methods, the characteristics of differential frequency hopping sequence were analyzed. Based on neural network, SD-Bi-LSTM intelligent decoding algorithm was proposed to improve the accuracy and error correction ability of differential frequency hopping decoding algorithm. Simulation results show that the decoding accuracy and error correction ability of the proposed algorithm are superior to the traditional algorithm in different frequency sequence lengths.