Article: From legacy to intelligence: the future of cloud load balancing Journal: International Journal of Cloud Computing (IJCC) 2026 Vol.15 No.2 pp.119 - 180 Abstract: The inherent cloud dynamicity and resource-intensive nature mandates development of vigorous load balancers to ensure optimal resource utilisation. This paper aims to provide insight into state-of-the-art techniques falling under static, dynamic and nature-inspired load balancing environment paving way to explore machine learning-based solutions. It expounds how predictive capabilities of ML algorithms can yield informed decisions in dynamically allocating resources. Numerous LB models that can attain better solutions and various performance metrics have been analysed. The paper envisions future trends in developing ML-based solutions by entailing manifold research gaps. The novelty of survey originates from cohesive analysis of static, dynamic and nature-inspired LB strategies coupled with emerging ML-based approaches, performance-metric-based comparative assessment and discernment of open research challenges. By orchestrating LB indebted by advancements through ML approaches, this paper serves as collectible resource for researchers by yielding nuanced understanding for developing roadmap in developing efficacious cloud load balancers. Inderscience Publishers - linking academia, business and industry through research

Title: From legacy to intelligence: the future of cloud load balancing

Authors: Neha Miglani; Chander Diwaker

Addresses: Department of Computer Science and Engineering, University Institute of Engineering and Technology (UIET), Kurukshetra University, Kurukshetra, India ' Department of Computer Science and Engineering, University Institute of Engineering and Technology (UIET), Kurukshetra University, Kurukshetra, India

Abstract: The inherent cloud dynamicity and resource-intensive nature mandates development of vigorous load balancers to ensure optimal resource utilisation. This paper aims to provide insight into state-of-the-art techniques falling under static, dynamic and nature-inspired load balancing environment paving way to explore machine learning-based solutions. It expounds how predictive capabilities of ML algorithms can yield informed decisions in dynamically allocating resources. Numerous LB models that can attain better solutions and various performance metrics have been analysed. The paper envisions future trends in developing ML-based solutions by entailing manifold research gaps. The novelty of survey originates from cohesive analysis of static, dynamic and nature-inspired LB strategies coupled with emerging ML-based approaches, performance-metric-based comparative assessment and discernment of open research challenges. By orchestrating LB indebted by advancements through ML approaches, this paper serves as collectible resource for researchers by yielding nuanced understanding for developing roadmap in developing efficacious cloud load balancers.

Keywords: cloud computing; load balancing; LB; virtual machine; VM; task scheduling; resource utilisation; energy consumption; makespan; scalability; machine learning; accuracy.

DOI: 10.1504/IJCC.2026.154435

International Journal of Cloud Computing, 2026 Vol.15 No.2, pp.119 - 180

Received: 01 Apr 2025
Accepted: 28 Aug 2025
Published online: 29 Jun 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article

Title: From legacy to intelligence: the future of cloud load balancing

Keep up-to-date