Embedded Voice Acquisition System Design And Attack Detection In Data Transmission

With the rapid development of electronic technology,voice broadcasting technology is widely used in a variety of specific scenarios,such as railway stations,airports and so on.In order to improve the security of the broadcasting system and reduce the work intensity of people,it is urgent to propose a broadcast audio collection and recognition system suitable for a specific scene.Audio signal acquisition and transmission system is one of the most important links.In order to monitor the broadcast of each waiting room and platform of Taiyuan Railway Station and send the broadcast to PC for identification,it is necessary to ensure the security of voice transmission.If the network is attacked during the voice transmission,the result of voice recognition will be affected.However,the current research on cyber attack detection is still in the field of networked control,and there are few involved in the field of voice communication.Therefore,it is of great significance to study whether the voice is affected by cyber attack in transmission.This paper will design a voice acquisition transmission device and study the security of voice in transmission.The main work of this paper is as follows:(1)This article integrates STM32 MCU,VS1053 audio acquisition module and ESP8266 Wi Fi module on one circuit board,and applies two Wi Fi modules.One Wi Fi sends the voice signal,that is,the train station broadcast,and the other Wi Fi sends the broadcast position.Multi-node voice acquisition transmission system,which can receive broadcasts from different waiting rooms of the railway station and transmit the voice to PC for identification,and the real-time and reliability of the transmission is further improved.(2)This paper will use the autoregressive(AR)modeling method to build a model of the audio signal,and then preprocess the audio signal.At the same time,it analyzes the possible cyber attacks in audio transmission,and models three cyber attacks: denial of service attack(DOS),false data injection attack(FDI)and replay attack.Different methods are used to detect three kinds of cyber attacks.The Kalman filter and the chi-square detector are combined to detect the denial of service attack.The cosine similarity matching method is used to detect the false data injection attack,and the residual method is used to detect the replay.attack.Finally,by constructing the experimental platform,the collected transmission voice signal is transmitted to PC,and the attack detection algorithm is applied to identify whether the signal is subjected to a malicious cyber attack.The simulation experiment verifies the effectiveness of the proposed method.