MPEG video quality prediction in a wireless system

Wireless systems under design today promise to deliver high-bandwidth applications to mobile users in dynamic channel environments. The protocol used to control the network in a wireless system may not necessarily be able to control bit errors in its payload. Therefore, it is important to assess the impact of bit errors in a delivered data stream. This paper concentrates on the effects of bit errors on an MPEG-2 digital video sequence. A basic decoupling between bit error rate (BER) and peak signal-to-noise ratio (PSNR) was shown on previously published papers, signaling a loss of confidence on BER as a basic metric of digital system performance. This paper introduces a statistical approach to predicting video quality based solely on physical layer parameters. The mean error event length, λ_distance, is a metric derived from the physical layer that can be used to predict the expected video quality across systems with different channel coding algorithms. This metric proved to be uniformly consistent in predicting video quality for sequences corrupted by additive white Gaussian noise and protected by convolutional, Reed-Solomon (RS), and concatenated (convolutional and RS) codes