In order to attract more users, cloud vendors always provide free service in many cloud applications. If a Non-VIP (Non-Very Important Person) user is satisfied with the free cloud service, the user may well register as a VIP (Very Important Person) user. In this paper, we focus on the registration process. In order to provide a timely registration service, we propose a novel cloud architecture where free service and registration service are deployed on the same VM (Virtual Machine). We present a multiple-server queueing model with a second optional service for analyzing the stochastic behavior of cloud users. By establishing a two-dimensional Markov chain, we derive the steady-state distribution of the queueing model. We construct an individual benefit function to investigate the Nash equilibrium arrival rate of anonymous users, and propose a new intelligent searching algorithm to obtain the socially optimal arrival rate of anonymous users. For coinciding these two arrival rates of anonymous users, we impose an appropriate registration fee with a pricing policy.

Pricing policy for a cloud registration service with a novel cloud architecture

In this paper, we investigate a single server batch arrival non-Markovian retrial queueing model with random break down and Bernoulli vacation. Customers arrive in batches accordingly Poisson process with arrival rate but are served one by one with rst come rst served basis. All customers demand

/pdf/m-x-g-1-retrial-queueing-system-with-second-optional-service-51wie1kqvr.pdf

M[X]/G/1 Retrial Queueing System with Second Optional Service , Random Break down ,Set Up Time and Bernoulli Vacation

In this paper we consider a single server batch arrival non-Markovian retrial queueing model with non persistent customers. In accordance with Poisson process, customers arrive in batches with arrival rate  and are served one by one with first come first served basis. The server is being considered as unreliable that it may encounter break down at any time. In order to resume its service the server has to be sent for repair, but the repair does not start immediately so that there is a waiting time before the repair process. The customer, who finds the server busy upon arrival, can either join the orbit with probability p or he/she can leave the system with probability 1-p.  More details can be found in the full paper. Key words : Batch size, break down, delay time, transient solution, steady solution,  reliability indices.

/pdf/transient-behavior-of-m-x-g-1-retrial-queueing-model-with-4jdck1uke8.pdf

Transient behavior of M[x]/G/1 Retrial Queueing Model with Non Persistent Customers, Random break down, Delaying Repair and Bernoulli Vacation

This paper deals with unreliable M/G/1 queueing system with different vacation policies, nonpersistent customers and repeated attempts. The jobs arrive in Poisson fashion. On finding the server busy, under setup, under repair or on vacation, the jobs either join to the orbit or balk from the system. The jobs in the orbit repeat their request for service after some random time. The inter retrial time of each job is general distributed. The jobs are served according to FCFS discipline. If the primary call, on arrival finds the server busy, it becomes impatient and leaves the system with probability (1 ) and with probability , it enters into an orbit. The server provides preliminary first essential service (FES) and followed by second essential service (SES) to primary arriving calls or calls from the retrial group. On completion of SES the server may go for i th (i = 1,2,3··· ,M) type of vacation with probability i(i = 1,2,3,...,M) or may remain in the system to serve the next call, if any, with probability 0 where P M=0 i = 0.

Analysis of Poisson repeated queue general service with set up time and different vacation policies

T his paper investigates a single server retrial queuing system with two stages heterogeneous serviceCustomers arrive in batches in accordance with compound Poisson processes After the completion of first stage service, the second stage service starts with probability 1 In addition to this, the server takes Bernoulli vacation and setup times We assume that the retrial time, the service time, the repair time, the vacation time and the setup time of the server are all arbitrarily distributed We obtain the time dependent probability generating functions in terms of their Laplace transforms and the corresponding steady state results explicitly Also we derive the average number of customers in the queue and the average waiting time in closed form with numerical illustration

• transient solution of m[x]/g/1 retrial queue with two stage heterogeneous service, extended vacation, non-persistent customers and setup time

In this paper, we study an M/G/1 queue with two phases of heterogeneous service. A first essential service is provided to all arriving customers. Upon completion of this service, a customer can either opt for a second phase of service or can leave the system. A customer, who finds the server busy, either leaves the system with probability (1-α )o r joins an orbit with probability α. From the orbit the customer makes repeated attempts to obtain service. We assume the inter retrial times are exponential random variables. We also assume that upon completion of a service, the server either remains in the system with probability β0 or leaves the system for an ith type of vacation with probability

/pdf/an-m-g-1-retrial-queue-with-non-persistent-customers-a-2gr2u008fe.pdf

Kasturi Ramanath

Papers

An M/G/1 Retrial Queue with Non-Persistent Customers, a Second Optional Service and Different Vacation Policies