3rd International ICST workshop on mobile media delivery

Research Article

Unequal loss protection and multiple description coding: a performance comparison

  • @INPROCEEDINGS{10.4108/ICST.MOBIMEDIA2009.7418,
        author={Enrico Baccaglini and Tammam Tillo and Gabriella Olmo},
        title={Unequal loss protection and multiple description coding: a performance comparison},
        proceedings={3rd International ICST workshop on mobile media delivery},
        publisher={ACM},
        proceedings_a={EUMOB},
        year={2010},
        month={5},
        keywords={},
        doi={10.4108/ICST.MOBIMEDIA2009.7418}
    }
    
  • Enrico Baccaglini
    Tammam Tillo
    Gabriella Olmo
    Year: 2010
    Unequal loss protection and multiple description coding: a performance comparison
    EUMOB
    ICST
    DOI: 10.4108/ICST.MOBIMEDIA2009.7418
Enrico Baccaglini1,*, Tammam Tillo2,*, Gabriella Olmo1,*
  • 1: Dipartimento di Elettronica, Politecnico di Torino, Corso Duca degli Abruzzi 24 - 10129 Torino - Italy.
  • 2: Department of Electrical and Electronic Engineering, Xi’an Jiaotong-Liverpool University, 111 Ren Ai Road - Suzhou 215123 - P.R. China.
*Contact email: enrico.baccaglini@polito.it, tammam.tillo@xjtlu.edu.cn, gabriella.olmo@polito.it

Abstract

In this paper we present a performance comparison between multiple description coding and unequal loss protection as tools to encode a layered source. We address a rate-distortion- based multiple description coding scheme and a stateof- the-art unequal loss protection algorithm based on Reed Solomon FEC codes. The comparison is performed using as a case study JPEG 2000 coded images transmitted over lossy packet networks. Complexity aspects are also considered. The simulation results show that both schemes allocate the same amount of redundancy for any given encoding output rate to protect the transmitted information. Whereas MDC, besides being computationally less intensive, achieves a smoother performance degradation, the ULP scheme yields superior performance in terms of the expected PSNR.