Qualitative and Quantitative Analysis of the HEC Mechanism in ATM

Fault tolerance measures like the CRC-based Header Error Control (HEC) algorithm in ATM can be used to reduce failure probabilities and by this enhance the performance/Quality of Service (QoS) of a transmission service. However, a detailed analysis of the applied algorithms is crucial for a system design that is well balanced with respect to the trade-off of redundancy cost and benefit. The analysis cannot be restricted to the quantitative aspect of a global failure probability but, instead, must include an examination of failure types and their respective ‘severeness’ and mapping through the layers. Taking the HEC mechanism of ATM as an example, it is shown how such a qualitative and quantitative analysis can be carried out. As an additional result, it is derived that the design decision to cover the entire cell header by the HEC algorithm is not optimal with respect to the needs of some commonly used AALs.