The Research On Content-based Packet Loss Recovery For Music Streaming

Bandwidth efficiency and error robustness are two essential and conflicting requirements for streaming media content over error-prone channels, such as wireless.published works in audio packet loss recovery have focused on speech. Relatively few published works have dealt with high quality music, their quality requirements in streaming applications are quite different. Technologies which are applied to the audio packet loss do not suit to the music streaming. Technologies do not use the features of the music. a result proposed a novel scheme proposed "content-based unequal error protection". The main contents of this dissertation are as follows:1.For the transients, which exist in the music streaming, we processed sub-frame and extract the data frames with the onset point. Through this process, we determine the position of the onset by calculating the temporal energy/harmonic structure/zero crossing rate/phase change.2.For the different characteristics of the onset point, we use the SVM and the dispenser wrapper mode, according to the features-the temporal energy/harmonic structure/bandwidth; we form a feature vector to classify the feature space into two.3.According to the characteristics of different types of onset point, we do some recovery respectively. We will send the secondary bit stream to the receiver side. We can recover the packet loss by the secondary bit stream.The experiments show that our scheme improves audio quality significantly over simple muting and packet repetition is achieved with a negligible amount of redundancy, which is transmitted to the receiver ahead of playback.