Queueing Model Approximation for Multi-Threaded Architecture With Feedback Jobs
Publication Date
1-14-2026
Document Type
Article
Publication Title
Methodology and Computing in Applied Probability
Volume
28
Issue
1
DOI
10.1007/s11009-025-10227-w
Abstract
Queueing theory plays a major role in the modern era applications like streaming, scheduling etc. to improve the service and processing time of the applications. The proposed system designs a dataflow model in multi-threaded architecture using a closed queueing model with feedback jobs. The model uses two types of processors namely synchronization processor (SP) and execution processors (EP) to take care of load store operations and logic, control operations respectively. A single synchronization processor and n execution processors are used in this model. The model is developed, considering the following two cases to maximize the utilization of functional units: (i) when there are only feedback jobs from the EP’s (ii) when there are jobs from both the SP and the feedback channel to the EP’s. The system is numerically validated by plotting the average system throughput as a function of total number of active jobs, degree of multiprogramming for both the cases. The response measures such as average time of execution, queue length, response time, utilization of servers and system throughput are calculated. In addition, the execution process of the model has been discussed using SimEvents toolbox of MATLAB.
Keywords
Execution processors, Response time, Synchronization processors, Throughput, Utilization
Department
Computer Engineering; Mechanical Engineering
Recommended Citation
K. Vimhala, G. Lavanya, Vidhyacharan Bhaskar, V. Subburayan, and Annie Uthra. "Queueing Model Approximation for Multi-Threaded Architecture With Feedback Jobs" Methodology and Computing in Applied Probability (2026). https://doi.org/10.1007/s11009-025-10227-w