Statistical modelling of correlated and independent errors in packet data transmission systems

Abstract

Two types of errors are common in packet data transmission systems, namely the random single bit error and the burst error. We define the burst error as a consequence of the error in bits in a packet. The burst error can be independent or correlated with the single bit error within a packet, depending on the source of errors. Although the conventional models assume either independent or correlated errors, in many practical transmission systems, both types of errors are present. We propose and develop statistical models to compute the average bit error rate by simultaneously considering both the independent and the correlated errors. The results obtained from our models suggest that the conventional models may underestimate the actual bit error rate in certain systems. The proposed models are fairly general and may be applied to any transmission systems, including cellular systems, where the errors are affected by both the independent and correlated phenomena.