Worst Deterministic Pattern Allocation: a viable approach to attain statistical gain in ATM

This paper deals with a novel strategy for Connection Admission Control in ATM networks. This strategy, called Worst Deterministic Pattern Allocation (WDPA), aims at reaping statistical gain of ATM, through a simple and controllable bandwidth allocation. WDPA is based on the concept of constraining information sources to emit their cells according to a superimposed deterministic pattern. This forced regularity is exploited in order to easily control the cell transfer within the network. Practically, WDPA establishes the worst emission pattern of the source. Resource allocation is performed taking into account only the parameters of the declared deterministic pattern. The same allocation rule is utilized in every network section. Application of WDPA to a large class of switching structures with output buffer is also discussed and its compatibility with this architecture is demonstrated. The presented performance study aims at evaluating the impact of the shaping on the original source cell stream and at determining general criteria for the choice of the deterministic pattern parameters. Finally, we compare the efficiency resulting from application of WDPA with that relevant to peak rate, mean rate and an ideal statistical allocation strategies.