The Design And Implementation Of Speech Subsystem Of A Supervising System Of Radar And Speech

It is the most important thing for civil aviation work to keep safety. As the authorities of safe work of civil aviation, every Air Traffic Control Center consistently takes more care of safety work. It has acquired good performance on safety work. In the meantime, with the fast increasing of air traffic, difficulty and intensity of air traffic control work are increased. It results in serious effect on safe work. To enhance its supervising on safety work, every Air Traffic Control Center wish it could watch and listen to the real-time information of real air traffic control work, and it could extract unsafe information from history speech and radar data that have been recorded to qualitatively and quantitatively analyze. As a result, it can more easily get objective conclusion according to these analyses. The supervising system of radar and speech related to this thesis has been developed to meet the requirements of Air Traffic Control Center. It has the ability to collect real-time data of radar and speech, synchronized play real-time and history data of radar and speech, analyze and process history data etc. it is satisfying to user. The author of the thesis has participated in design of system architecture, and taken charge of development of speech subsystem independently. The speech subsystem has the ability to collect speech data, code and decode speech data, transport real-time speech data, synchronized play real-time and history speech data, and store speech data, etc. the thesis is based on the work of author. It mainly includes development of speech subsystem. It has been divided to the following 6 points. Firstly, the thesis discusses the design of architecture of the supervising system and design of module of collection of speech and radar. Secondly, the thesis discusses how to collect speech data by hardware and code data in ADPCM format. It discusses in detail the theory and algorithm of ADPCM. It also discusses the characteristic and process of decoding by software. Thirdly, the thesis discusses how to transport real-time speech data to play subsystem. It manly discusses multicast and RTP/RTCP technology. It also discusses an algorithm of adaptively adjusting length of buffer and a strategy about feedback control. Fourthly, the thesis discusses how to synchronized play real-time and history speech data in real-time and history situations. It discusses the synchronization strategy in the two situations, it also discusses the characteristic and process of play speech data by service of windows OS. Fifthly, the thesis discusses how to store the speech data to oracle database. It mainly discusses supports provided by oracle database to store multimedia data and the OCCI interface provided by oracle database to access and manipulate oracle database. Sixthly, the thesis discusses how to implement a class which is safe in the environment of multithread. It discusses the issue from three arrangements of language, win32 API, and MFC.