A Systematic Literature Review on Robust Swarm Intelligence Algorithms in Search-Based Software Engineering

Hindawi Complexity Volume 2023, Article ID 4577581, 22 pages https://doi.org/10.1155/2023/4577581 Review Article A Systematic Literature Review on Robust Swarm Intelligence Algorithms in Search-Based Software Engineering Alam Zeb ,1 Fakhrud Din ,1 Muhammad Fayaz ,2 Gulzar Mehmood ,3 and Kamal Z. Zamli 4 1 Faculty of Information Technology, Department of Computer Science & IT, University of Malakand, Lower Dir 18800, KPK, Pakistan 2 Department of Computer Science, University of Central Asia, Naryn, Kyrgyzstan 3 Department of Computer Science, IQRA National University, Swat Campus 19220, Peshawar, Pakistan 4 Faculty of Science and Technology, Universitas Airlangga, C Campus JI. Dr. H. Soekamo, Mulyorejo, Surabaya 60115, Indonesia Correspondence should be addressed to Muhammad Fayaz; Received 1 May 2022; Revised 4 September 2022; Accepted 7 October 2022; Published 23 February 2023 Academic Editor: Haitham Abdulmohsin Afan Copyright © 2023 Alam Zeb et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Swarm intelligence algorithms are metaheuristics inspired by the collective behavior of species such as birds, fsh, bees, and ants. Tey are used in many optimization problems due to their simplicity, fexibility, and scalability. Tese algorithms get the desired convergence during the search by balancing the exploration and exploitation processes. Tese metaheuristics have applications in various domains such as global optimization, bioinformatics, power engineering, networking, machine learning, image processing, and environmental applications. Tis paper presents a systematic literature review (SLR) on applications of four swarm intelligence algorithms i.e., grey wolf optimization (GWO), whale optimization algorithms (WOA), Harris hawks optimizer (HHO), and moth-fame optimizer (MFO) in the feld of software engineering. It presents an in-depth study of these metaheuristics’ adoption in the feld of software engineering. Tis SLR is mainly comprised of three phases such as planning, conducting, and reporting. Tis study covers all related studies published from 2014 up to 2022. Te study shows that applications of the selected metaheuristics have been utilized in various felds of software engineering especially software testing, software defect prediction, and software reliability. Te study also points out some of the areas where applications of these swarm intelligence algorithms can be utilized. Tis study may act as a guideline for researchers in improving the current state-of-the-art on generally adopting these metaheuristics in software engineering. 1. Introduction Te goal of optimization is to fnd a solution that either minimizes or maximizes a criterion known as the objective function, ftness function, or cost function. A solution known as a feasible solution or admissible solution is required to solve the objective function under given constraints [1]. An optimization algorithm works iteratively until a specifed number of iterations or when a specifc amount of time is reached to enhance a given criterion [2]. Many real-life optimization problems in engineering, science, and business are complicated and, thus, no exact solutions can be provided in a fair amount of time. To tackle such problems, approximate algorithms are used. Approximate algorithms are further divided into specifc heuristics and metaheuristics. Specifc heuristics being problem-dependent solve specifc problems, whereas metaheuristics are general-purpose strategies used to solve a huge number of optimization problems [3]. Generally, metaheuristics solve instances of hard problems through exploration of the search space where the solution is supposed to be found. Tere are mainly three objectives of metaheuristics: (1) providing a fast solution to a given problem; (2) tackling large problems; and (3) obtaining robust algorithms. Furthermore, metaheuristics are fexible and simple to design as well as implement [3]. 2 Complexity In Search-Based Software Engineering (SBSE), metaheuristic search techniques are applied to optimization problems in software engineering. Te main objective of SBSE is to move problems related to software engineering to machine-based search instead of human-based search. Tis is accomplished by using various techniques of metaheuristics as well as operations research [4]. Near-optimal solutions are acceptable in various engineering disciplines, and software engineering is no exception. It is due to this fact that metaheuristic techniques are applicable to solving software engineering problems. Tus, difcult problems can be solved using such search-based techniques with acceptable solutions where perfect solutions are infeasible or impossible [5]. Various swarm intelligence algorithms such as GWO, WOA, HHO, and MFO algorithms have been applied in numerous domains such as machine learning, networks, engineering optimization, environmental modeling, image processing, power dispatch problems, and medical and bioinformatics [6–9]. Some of the recent nature-inspired metaheuristic algorithms include monarch butterfy optimization (MBO) proposed by Wang et al. [10], slime mould algorithm (SMA) proposed by Li et al. [11], moth search (MS) algorithm proposed by Wang [12], marine predators algorithm (MPA) presented by Faramarzi et al. [13], hunger games search (HGS) presented by Yang et al. [14], and colony predation algorithm (CPA) developed by Tu et al. [15]. Tis paper presents a systematic literature review on applications of four swarm intelligence algorithms, namely, GWO, WOA, HHO, and MFO, in the domain of software engineering. Our motivation comes from the fact that, as per our knowledge, there is no such systematic review available. Also, a review study paper helps researchers with aggregated knowledge on a given topic in one place. Table 1 shows various felds in software engineering that utilize these swarm intelligence algorithms. It also points out various areas in software engineering where these algorithms could be utilized. Te main contributions of this SLR are as follows: (i) Te SLR presents applications of four swarm-based metaheuristics i.e., GWO, WOA, HHO, and MFO in the feld of software engineering. (ii) Te SLR analyzes 46 relevant studies from 2014 to 2022 according to several research questions. (iii) Tis study points out new areas in software engineering where the selected metaheuristics could be utilized thus it identifes further research opportunities in the feld. Many swarm intelligence algorithms are available and are gaining prominence as they ofer reasonable solutions to complex problems. Tese algorithms are mainly inspired by biological systems such as animal-based and insect-based. We have chosen four swarm intelligence algorithms, one from each group, i. e., GWO from animal-based, MFO from insect-base (...truncated)