Open Access

ARTICLE

Task Allocation Approach for Minimizing Make-Span in Wireless Sensor Actor Networks

Mohammad Reza Okhovvat¹, Mohammad Taghi Kheirabadi^1,*, Ali Nodehi¹, Morteza Okhovvat²

1 Department of Computer Engineering, Gorgan Branch, Islamic Azad University, Gorgan, Iran
2 Health Management and Social Development Research Center, Golestan University of Medical Sciences, Gorgan, Iran

* Corresponding Author: Mohammad Taghi Kheirabadi. Email:

Computer Systems Science and Engineering 2021, 39(2), 165-178. https://doi.org/10.32604/csse.2021.05468

Received 23 March 2020; Accepted 13 July 2020; Issue published 20 July 2021

Abstract

Wireless Sensor Actor Networks (WSANs) have contributed to the development of pervasive computing wherein time consideration to perform the tasks of pervasive applications is necessary. Hence, time constraint is one of the major challenges of WSANs. In this paper, we propose an analytical approach based on queuing theory to minimize the total time taken for completion of tasks, i.e., make-span, in WSANs with hybrid architecture. The best allocation rates of tasks to actor nodes are figured out through solving inequities and qualities resulting from a steady state analysis of the proposed model. Applying the calculated tasks arrival rates at each of the actors, the make-span could be minimized. To assess the accuracy of the tasks assignment rates to each of the actors attained from the suggested analytical approach and to provide a graphical representation of the WSAN a formal model in terms of the generalized stochastic Petri net (GSPN) is presented. The proposed GSPN model is analyzed, tasks distribution weights to the actors are determined, and then tasks allocation rates can be computed. Comparing the results achieved from the analytical approach and the GSPN model demonstrates that allocation rates and hence, the make-span figured out from proposed approach and the formal model are the same. Experimental results in typical scenarios show shorter make-span and longer network lifetime compared to when one of the two popular traditional task allocation algorithms, namely, opportunistic load balancing (OLB), and stochastic allocation (SA) algorithms, is used.

---

Keywords

---