基于apes的音频数据丢失恢复技术计算机科学与技术专业论文.docxVIP

哈尔滨工业大学工学硕士学位论文 哈尔滨工业大学工学硕士学位论文 Abstract Audio steaming application has become very popular in recent years, owing to their low cost and convenience. But now the Internet can only provide a best-effort service, network congestions will introduce the occurrence of packet loss and delay, which will create an annoying gap in the streamed media. In order to ensure the quality of service, some strategies need to be taken to reduce the adverse effects of audio quality caused by packet loss and delay. Usually these strategies are divided into receiver-based and sending-based. This paper presents a type of receiver-based method based on the APES(the amplitude and phase estimation of a sinusoid)spectrum estimation method for audio packet loss recovering. First we introduce the basic principle of APES spectrum estimation method , including its maximum likelihood estimation explanation and filterbank derivation. Then we propose several kinds of simplified method based on its property in order to make it available in the long-time data sequence analysis. Next we introduce the data recovering method based on the simplified APES into the audio data loss recovering in time domain. The experimental results show that its performance is much better than the simple method such as the waveform substitute no m atter in the objective evaluation or the subjective comparison. But to recover the data based on the APES in time domain has its limitation, with its high time complexity it could be difficult to real-time realization. So we propose a method to use the recovery algorithm in the transform domain for MP3 audio frame. But the encoded domain of MP3-MDCT(modified discrete cosine transform)coefficients has a fast-changing sign makes them difficult to interpolate. Working in a complex domain, which has a less fluctuating representation of the signal, should provide better interpolation results. For this reason, we combining the MDCT coefficients as real part and the corresponding MDST(mo