Adaptive Spectral Separation Two Layer Coding with Error Concealment for Cell Loss Resilience

This paper addresses the issue of cell loss and its consequent effect on video quality in a packet video system, and examines possible compensative measures. In the system's enconder, adaptive spectral separation is used to develop a two-layer coding scheme comprising a high priority layer to carry essential video data and a low priority layer with data to enhance the video image. A two-step error detection scheme using cell numbering and variable length decoder is incorporated in the system to detect cell loss in the bit-stream without long mis-interpretation. In the system's decoder, cell loss is compensated by a two-step recovery procedure based in temporal error concealment to maintain high picture quality. The simulated performance of the proposed two-layer system is studied and compared with that of the single layer system. It is shown that good resilience and PSNR improvement can be achieved even at quite high error rates. Statistics such as the cell-rate distribution and autocorrelation function of both high and low priority streams generated by the adaptive system are analyzed. It is observed that two different classes of gamma function can represent the cell distribution in each layer. A multiplicative – ARIMA (autoregressive integrated moving average) model is proposed to represent and forecast the number of cells per frame for the encoded video traffic and its goodness-of-fit is compared with AR (1) (autoregressive) model. The cells per frame measured from a sample video sequence compare favorably with those obtained by the ARIMA model. Simulations were performed using three five-second video sequences to demonstrate the efficiency of the two-layer enconding algorithm for sequences with different characteristics. Statistics gathered from these prioritization schemes are claimed to be useful for network design and resource management.