Abstract:
Many software/hardware MPEG audio encoding schemes have been proposed to reduce encoding time. At present, software scheme that is used to encode the MPEG audio is implemented using sequential algorithm. Also real time encoding usually requires very high-speed computer. However, approach still has the limitation of encoding time performance due to the limited speed of a computer. It is interesting to improve the encoding algorithm by implementing the MPEG audio encoding in parallel processing architecture to overcome the limitation of encoding time.
In this study, parallel processing based software is proposed to improve the performance of MPEG audio encoding. The implementation algorithm is simulated on RedHat Linux version 6.2 with MPICH release 1.2.0 from Argonne National Laboratory [15]. Various numbers of processors are experimented and evaluated.
Experimental result shows much improvement in the overall encoding time. The MPEG audio encoding and parallel processing can work well together due to the independent data characteristics of MPEG audio and the data management of parallel processing.