Abstract: Cloud computing paradigm, like the paradigm of network computing, is based on the clustering of resources and on the usage of network and Internet technologies. In general, the cloud computing refers to a new way of delivering computing resources in a form of service, such as data, software and hardware components (processing elements, memory and storage).
Cloud computing is a current and important multidisciplinary field, followed by a large number of published papers in the state-of-the-art international journals, as well as in the proceedings of the world-renowned international conferences. Based on the scientific results which have been gathered from the published papers in this domain, it can be concluded that there are many challenges and problems in the cloud environment, which can be more efficiently tackled by applying improved methods, techniques and algorithms. One of the most important challenges in cloud computing is scheduling of end users' requests on a limited set of available resources (virtual machines). A scheduling problem in cloud environment can be defined as an execution schedule of tasks on a limited set of available resources, taking into account the potential constraints and objective function.
Task scheduling is performed by scheduling algorithms, which can be divided into static and dynamic. In the case of a static scheduling, where it is not possible to dynamically switch tasks from over utilized to underutilized virtual machines, tasks are being allocated for execution on available virtual machines before the scheduling algorithm execution. In the case of the dynamic scheduling methods, which are known in the literature as load balancing approaches, a workload allocation between active virtual machines is being performed during the scheduling algorithm run-time. Requests' redistribution is executed by dynamically switching from over utilized to less utilized virtual machines. Heuristics and metaheuristics optimization methods are mostly used for dynamic scheduling, where they have achieved great results.
Task scheduling and load balancing problems in cloud computing belong to the group of NP hard combinatorial and/or global problems with or without constraints. Based on the published results in the relevant literature, it can be concluded that the swarm intelligence metaheuristics have been successfully tested on benchmark and practical NP hard optimization problems, and that they have achieved better results in terms of convergence speed and the solutions' quality, than other methods, techniques and algorithms. In our experiments, it is examined whether it is possible to further improve the task scheduling and load balancing in cloud computing environment by applying swarm intelligence metaheuristics.
During the experimental research, several swarm intelligence metaheuristics were improved and adapted for solving task scheduling and load balancing problems in cloud environment. Some of the swarm algorithms that proved state-of-the-art performance in terms of results’ quality, as well as of convergence speed are monarch butterfly optimization (MBO), whale optimization algorithms (WAO), elephant herding optimization (EHO), tree growth algorithm (TGA) and grey wolf optimization (GWO). The algorithms were implemented in both, original and modified/hybridized versions. The robust environment of CloudSim platform was utilized as the simulation platform.
In this speech, I will highlight the most significant results that the swarm algorithms obtained in the domain of cloud computing task scheduling and load balancing.

Biography: Nebojsa Bacanin received his first Ph.D. degree in 2014 from the domain of applied computer science, and second Ph.D. degree from Faculty of Mathematics, University of Belgrade in 2015 (study program Computer Science, average grade 10,00). He started University career in Serbia 15 years ago at Graduate School of Computer Science in Belgrade. He currently works as an associate professor and as a Vice-Rector for Scientific Research at Singidunum University, Belgrade, Serbia. He teaches 16 courses on bachelor, master and Ph.D. studies from the domain of computer science.

He is involved in scientific research in the field of computer science and his specialty includes stochastic optimization algorithms, swarm intelligence, soft-computing and optimization and modeling, as well as artificial intelligence algorithms, swarm intelligence, machine learning, image processing and cloud and distributed computing. He has published more than 100 scientific papers in high quality journals and international conferences indexed in Clarivate Analytics JCR, Scopus, WoS, IEEExplore, and other scientific databases, as well as in Springer Lecture Notes in Computer Science and Procedia Computer Science book chapters. He has also published 2 books in domains of Cloud Computing and Advanced Java Spring Programming.

He is a member of numerous editorial boards, scientific and advisory committees of international conferences and journals. He is a regular reviewer for international journals with high Clarivate Analytics and WoS impact factor such as Journal of Ambient Intelligence & Humanized Computing, Soft Computing, Applied Soft Computing, Information Sciences, Journal of Cloud Computing, IEEE Transactions on Computers, IEEE Review, Swarm and Evolutionary Computation, Knowledge-based Systems, Future Generation Computer Systems, Computer and Information Sciences, SoftwareX, Neurocomputing, Operations Research Perspectives, etc. In 2020 he was designated by prestigious Stanford University list as top 2% researchers in the world.

http://eng.singidunum.ac.rs/profile/nbacanindzakula