An QPSL Queuing Model for Load Balancing in Cloud Computing

Load balancing is the process of distributing a workload among various servers. Queuing is the most common scenario for day-to-day applications. Queuing theory is used to study the problem of waiting lines. Queuing theory bridges the gap between service demands and the delay in replies given to users. The proposed QPSL Queuing Model makes use of M/M/k queue with FIFO queue discipline for load balancing in cloud computing. The model makes use of exponential distribution for calculating service rates and Poisson distribution for calculating waiting lines. The proposed QPSL queuing model is also compared with other existing queuing models for load balancing on various parameters. The experimental analysis depicts that QPSL model performed better in terms of service rate and response time.

The inclusion structure of partially lossy queue monoids and their trace submonoids

We model the behavior of a lossy fifo-queue as a monoid of transformations that are induced by sequences of writing and reading. To have a common model for reliable and lossy queues, we split the alphabet of the queue into two parts: the forgettable letters and the letters that are transmitted reliably. We describe this monoid by means of a confluent and terminating semi-Thue system and then study some of the monoid’s algebraic properties. In particular, we characterize completely when one such monoid can be embedded into another as well as which trace monoids occur as submonoids. Surprisingly, these are precisely those trace monoids that embed into the direct product of two free monoids – which gives a partial answer to a question raised by Diekert et al. at STACS 1995.