Performance analysis of a finite-buffered DQDB station

The paper analyzes the performance of a DQDB station with a finite buffer capacity. The authors take a per-node approach to the analysis and make a Bernoulli assumption for the bus outcomes. The model, which is based on a discrete-time Markov chain, is solved by means of EPA (equilibrium point analysis). By utilizing this method, one can calculate the steady-state probabilities of the Markov chain; then, one can evaluate throughput, average bus access delay and average MAC delay. The accuracy of the analytical results are examined through simulation. For light to medium loads, the analytical results show little discrepancy when compared to simulation results. A small discrepancy is found for some stations, and it is explained in part by the correlation between outcomes of consecutive slots, which is not captured in the the model