The impact of G.826 on the performance of transport systems

Study Group XIII of the ITU-TS has recently approved a new performance recommendation G.826 which the ITU-R has recommended as a standard for satellite systems in recommendation S.1062. G.826 establishes guidelines on the performance levels of digital systems operating at or above primary rate. These performance levels are stated in terms of block error rates. We discuss the impact of the block-based error performance objectives on the performance of transport systems. However, the impact of G.826 is difficult to evaluate directly since the performance recommendations are not given in terms of bit-error rate (BER) values. Hence, we present a procedure to convect block error rates to BER. This conversion is based on the statistical modeling of the error process and includes an analysis of bursty errors as well as random errors. Rather than using the equivalent BER values as simple thresholds, we produce BER against time distributions which satisfy G.826 and can be used as masks for a comparison with measured data and also for route design. It is shown that these masks are a crucial tool in any study of G.826 and contain useful information for system design and planning. Indeed it is shown that false conclusions can be drawn if the masks are not used. Finally, the impact of G.826 is discussed with particular reference to apportionment strategies, system margins, signal processing, and error correction schemes.