Simultaneity in discrete-time queues with compound Bernoulli inputs and AF-DF buffer management policies

This paper focuses on discrete-time single server queues with finite capacity and Compound Bernoulli Inputs with geometrically distributed bulk sizes under Arrivals First (AF) and Departures First (DF) buffer management simultaneity policies. The influence of these policies on typical performance measures is investigated and closed-form expressions relating to various state probabilities are based on the principle of maximum entropy (ME) and probabilistic arguments. Extensions to the approximate analysis of discrete-time open queueing networks with or without repetitive-service blocking and random destination (RS-RD) are carried out. Numerical validation results indicate that the ME approximations have a good error level in relation to simulation