Showing papers on "Server published in 1969"

Ordered Entry Queueing Networks with Multiple Servers and Multiple Queues

Abstract: The purpose of this research was to investigate the conveyor-fed multiple channel queueing system with ordered entry and storage at each channel. The primary objective was to determine the operating characteristics of such a system, and to develop a general predictive theory regarding these characteristics. A simulation analysis was undertaken to analyze the m-channel homogeneous and heterogeneous service system with recirculation and storage at each channel. Pertinent results are presented and analyzed for a wide range of arrival and service rates, with several different storage combinations at each of the service channels.

Note on the output of a queuing system

Abstract: Consider a queuing system that has c servers and d waiting positions. Assume that the input is Poisson with rate a and the service times are exponential with mean fl-1. Further assume the following: (i) a customer arriving when all servers are busy and all waiting positions are occupied is "cleared" from the system; (ii) a customer arriving when all servers are busy and not all waiting positions are occupied waits with probability I ( and "balks" or "clears" with probability (; (iii) a customer arriving when not all servers are busy commences service immediately (never balks); and, (iv) a customer who is waiting for service may "defect", the distribution of time until a waiting customer defects being given by an exponential distribution with mean y-1. Also, the usual independence assumptions, which make the process that is described by the number in the system at time t Markov, are assumed. An "output" of this queuing system is defined to occur whenever a service completion occurs, or whenever an arrival "clears" or "balks", or whenever a waiting customer "defects". Thus the output is a pooling of service completion epochs, the epochs when arrivals are cleared, the epochs when arrivals balk, and the defection epochs. The result of this note is that under conditions of equilibrium, the output process is, like the input process, a Poisson process with rate a. An application of this result is presented if it is desired to monitor the grade of service that the queuing system gives by quizzing departing customers. The output of the queuing system then becomes the input to the monitoring system, and for engineering and analysis purposes it is useful to know that this input is Poisson. Burke (1956) and Reich (1957) first proved that the output has the same Poisson distribution as the input for the particular case ( = = 0 (no balking and no defections) and d = oo (infinite number of waiting positions). The special case d = y = ( = 0 and c = oo (infinite number of servers) appears in Benes (1957). Finch (1959) generalized, for the single server case, Burke's result to arbitrary service time distribution, and Mirasol (1963) gave a similar generalization of Benes' result. Mirasol's result is given a simple intuitive explanation in Kendall (1964).

A Multiserver Queue with Enforced Idle Times

Abstract: In this paper, exact, recursive, and asymptotic formulas' are derived for a queuing system in which access to the servers takes place at regular intervals. Steady-state results, already known, are extended and utilized to construct delay distributions that are adaptable to computer work.

A Performer-Oriented Approach to Systems Quality

Abstract: The objective of this paper is to present a means of modeling the quality characteristics of a large class of systems. The proposed models depend upon the use of a state variable to describe the quality level at various points in the system and upon, transition matrices to describe the error-producing and error-detecting characteristics of the individual servers and machines that compose the system. Applications of the model to linear, cyclic, and branching systems are developed in detail. Ultimately the model may be used to optimize the number, type, and location of quality control components or to optimize the performance of systems operating under risk.

Investigation of the mean waiting time for queueing system with many servers

Abstract: In this paper, we shall consider two models of queueing system such as shown in the figure below 9 Model I is a GI/M/S(c~) type queueing system with S equal exponential service channels in parallel, each having mean service rate g/S. Model 2 is a GI/M/S-I(c~) type queueing system with ( S l ) equal exponential service channels in parallel, each having mean service rate /~/(S-1). So, the total value of the mean

Pollution-Free Regenerating Codes With Fast Reconstruction Verification for Verifiable Cloud Storage

Abstract: In this article, we present an efficient mechanism to guarantee the integrity of cloud storage systems using network-coding-based regenerating codes. Regenerating codes are a special type of erasure codes with the property that regenerating a failed encoded server, can be done using significantly less information than the size of the original data. Our proposal includes two protocols that allow a file owner to verify an outsourced storage facility is storing its information without downloading the file itself. One of them can be used an unlimited number of times (unbounded use), but involves higher information transmission for a single verification. The other one, must precompute results in advance (bounded use), but has a lower overhead for each verification. For the bounded-use proposal, we make use of the linear properties of the dot product and special key management strategies to reduce the overhead in systems with many servers. In addition to these protocols, we present a way to guarantee the file can be recovered by any combination of at least k servers in polynomial time in the number of servers. As an additional application, we present how our constructions can be used to reduce network usage for users, when uploading personal multimedia libraries to cloud music services. For this application, we devised a group testing method for linear functions based on bisections, whose worst case scenario is not worse than transmitting each file individually. Since our result applies independently of the composition of the input, the method is of independent interest.

Optimum speed of service in the design of customer data communications systems

Abstract: A model is given for selection of optimum bit speed for private line data communications applications based on the tradeoff between the cost of delay and the cost of equipment. Costs of delay are those costs incurred by the communications user which occur outside the communications system as a result of delays introduced by that system. The delays considered result from transmission delay and customer's queuing delay. The model assumes Poisson input, exponentially distributed message lengths, and k servers. A method for including the effect of error control is given. The bit speed selection model is applied to an information retrieval application.

Estimación de tasas de compresión de imágenes mediante métricas de distorsión completamente referenciadas

Abstract: Image remote access is an area of increasing interest because of the need to store, process and display a large amount of images available in web portals and network servers around the world. This need becomes a challenge as the size and quality of images increase. This work presents a method for the estimation of compression rates of images by means of full-referenced distortion metrics and shows its application to medical images that are processed by the JPEG, SPIHT, and EBCOT techniques. This method makes it possible that high-definition images can be transmitted in shorter times. Compression rates for the three mentioned techniques are calculated by considering the perceived- lossless compression limits and the correlation values between measurements. The results obtained in this work show that it is possible to achieve the progressive transmission of images for each of the three techniques, keeping the same perceived quality and shortening the transmission times. .