DITRA: an efficient event-driven multi-objective optimization algorithm for bandwidth allocation in IoT environments

Mehrin Rouhifar,Alireza Hedayati,Vahe Aghazarian

DOI: https://doi.org/10.1007/s10586-023-04214-4

2024-01-19

Cluster Computing

Abstract:Internet of Things (IoT) technology has facilitated different human-related activities and, therefore, has extended to a wide range of applications. However, these networks have the limited resources that must be utilized in an efficient manner. Hence, various studies have introduced several methods and algorithms for better managing these resources. Although these approaches produce acceptable solutions, their performance still needs improvement because of an increasing number of IoT devices and application services. To address such a limitation, the present study proposed a novel IoT environments-specific bandwidth allocation method, which dynamically distributes the wireless bandwidth among the devices according to their heterogeneous nature and the existence of various traffic services. To this end, the Trader metaheuristic algorithm was developed for the discrete problems and formulated as a multi-objective algorithm. To evaluate the performance of the proposed method, it was applied to the six gold standard datasets generated based on the Poisson distribution. The outcomes indicated that the proposed approach surpasses the other introduced state-of-the-art methods in terms of the service success rate by 6.32%, network throughput by 5.79%, and resource efficiency by 3.13%. The results also showed that based on the different statistical criteria, the proposed discrete metaheuristic algorithm yields better outcomes compared to the other efficient algorithms.

computer science, information systems, theory & methods