论文部分内容阅读
本文提出并实现了一种基于人工神经元网络理论的通信网信号识别系统的研制方案。该系统可以在复杂的线路及干扰情况下,实时完成传真、计算机Modem通信、BP机自动寻呼、拨号音、双音频、电子合成音、热线音乐、噪声和话音的自动识别工作。系统最大监测容量可达16384线,普通巡检单元的检出率超过90%,重点捕捉单元的检出率超过99.99%,系统虚警率低于10~(-5)。
In this paper, we propose and implement a communication network signal identification system based on artificial neural network theory. The system can automatically identify fax, computer Modem, BP automatic paging, dial tone, dual audio, electronic synthesized sound, hotline music, noise and voice in real time under the complicated circuit and interference conditions. The maximum monitoring capacity of the system can reach 16,384 lines, the detection rate of ordinary inspection units exceeds 90%, the detection rate of key capture units exceeds 99.99%, and the false alarm rate of the system is less than 10 ~ (-5).