An improved model for service time on playthrough rings

Models are presented for message service time on playthrough rings, a class of multiserver computer communication subnetworks that uses circuit-switching on a ring topology. The playthrough protocol uses a perpetually circulating control token, source-removed update control messages, and destination removed data messages to provide concurrent transfer of multiple messages of arbitrary length on a single ring. Data messages share ring bandwidth with and are frequently interrupted at source transmitters by update control messages associated with other concurrent message transfers and the circulating token. The rate of interruptions to message service varies according to message outbound distance from source to destination. The authors use the concept of service completion time in a queuing system that experiences preemptive-resume interruption to predict mean and variance of message service time on playthrough rings for messages with geometrically distributed lengths. The models' predictions are validated using simulation data. These models are suited for use in analytical models for approximating queuing and total transfer times for messages.<>