The queuing simulation presented was of high quality. In order to get good results, the simulation looks at two different scenarios with each of the three arrangements of the processors. The three different arrangements are centralized, distributed, and hierarchical arrangements. In each of these three arrangements, the simulation looks at both a low load and high load scenario. Using these six different results, it is possible to look at the best results for the arrangements. The simulation is very useful as it gives a range of results that can be analyzed and compared. Using these results, it is possible to decide the best arrangement for each scenario. The centralized queuing is seen to work best in small systems with few tasks. Distributed systems are seen to work better in larger systems with more tasks. However, it has a major disadvantage of having idle capacity. The hierarchical arrangement is seen to be the best as it carries the advantages of both the distributed and centralized arrangements.
A key change I would have suggested to the authors would be to increase the size of the task. This would have given more data and using this new data, it would have been possible to calculate the factor by which the hierarchical arrangement is more advantageous. This new statistic would be the best way to compare the difference in performance between the three advantages. As it stands, the results just show the advantage of the hierarchical arrangement but there are no figures to show by how much it is advantageous. This would show the efficiency of the hierarchical system over the distributed or centralized system.
Queuing simulations are usually simulations used to show how systems with limited resources distribute these resources. The most common places where queuing simulation is used is in telecommunications, call centers, traffic, computer networks and predicting computer performance. In such places, the tasks that need to be done are more than the resources that carry them out. Therefore, there is a system for arrival of the tasks, a system for arranging the tasks in a queue and finally, the performance of the task in the best order. The type of queue can be either arbitrary or deterministic. The most common queuing system is first come first served, also known as first in first out. This is usually the method used in centralized systems where the tasks are carried out in the order in which they arrived. Other more specialized and complex queuing methods are also available. These ensure that tasks with the highest priority are processed first. These queuing methods all have to take some factors into consideration; the arrival process, the service process and the number of servers. The arrival process can be either arbitrary or deterministic while the servers can be arranged in parallel or serially.
Reading about a simulation is more difficult as opposed to actually experiencing it. When reading, I come across figures and symbols and I have to actually visualize the setup in my mind in order to understand the simulation. When reading about a simulation, I cannot adjust anything and so it is impossible for me to know what would have happened in other circumstances. When actually performing a simulation, it is easier to understand what is going on in the whole process. I am free to perform adjustments in the input to study the different results.
References
Deo, N. (2009). System simulation with digital computer. New Delhi, India: Prentice-Hall of India.
Stewart, W. J. (2009). Probability, Markov chains, queues, and simulation: The mathematical basis of performance modeling. Princeton, N.J: Princeton University Press.
