We address the problem of evaluating the impact of forward error correction (FEC) coding on the peak signal-to-noise ratio (PSNR) quality of video communication over burst-loss channels. The probability of a given packet loss pattern and the associated mean square error (MSE) video distortion are formulated such that we can compute, at the encoder, the expected MSE distortion of the reconstructed video signal at the receiver, both in the presence or absence of FEC. The proposed model is applicable to any communication channel as long as the loss pattern probability is computable for that channel. Using our analytical model, the expected PSNR quality is computed in the case of transmission over various Markov channels. These results match with high accuracy actual simulation-obtained data thereby validating the proposed analysis. Furthermore, based on the proposed framework, we study the impact of FEC on PSNR video quality for different average packet loss ratios, average burst lengths, and FEC code rates.

Figure 1. Video streaming with FEC error control: white rectangles denote video packets; gray rectangles denote redundant packets.