Maximum Entropy Formalism of Multi-traffic Censored Priority Queue

The principle of maximum entropy (ME) proposes a stable method of inference for approximating the form of an unknown discrete state probability distribution, built on information expressed in terms of true expected values. The most fundamental in the analysis of complex queueing system is the provision of a convincing interpretation for a probability assignment free from arbitrary assumptions. In this paper, ME formalization is used to investigate performance of a censored priority system where we have considered three types of traffic i.e. video, voice and data entering the system to get service and the constraints involve only first two moments of inter arrival and service time distribution. The traffic arrival and departure process follow Generalized exponential distribution i.e. calls/packets arrive for service in burst/s and after getting service/or without getting service departs in burst/s. The video calls/packets that arrive when all the channels/servers are busy processing calls/packets wait in queue, data and voice calls/packets that arrive at the system and find shared channels busy are blocked. These block packets/calls are lost and never return to the system.