Systematic Literature Review and Meta-Analysis Journal Homepage: http://slr-m.com/index.php/home Symbiotic Organisms Search Optimisation Algorithm in Cloud Computing: A Nature-inspired Meta-heuristic Suleiman Sa’ad Modibbo Adama University of Technology Yola, Nigeria, suleimanu@mautech.edu.ng Muhammed Abdullah Universiti Putra Malaysia, Malaysia, abdullah@upm. edu.my Azizol Abdullah Universiti Putra Malaysia, Malaysia, azizol@upm.edu.my Fahrul Hakim Ayob Universiti Putra Malaysia, Malaysia, fahrul@upm.edu.my Abstract In the past few years nature-inspired algorithms are experiencing rapid growth where most optimisation problems in different domains are addressed using it. As a result of this development come the issue of handling a complex optimisation problem within short period remains very difficult. Symbiotic organisms search (SOS) algorithm is one of the nature-inspired metaheuristics that mimics the symbiotic association of organisms in an ecosystem. This paper proposes to investigate symbiotic organisms search algorithms used in handling various optimisation problems in different fields to bring out strengths and weaknesses of the existing algorithms as well as to point out future directions for the upcoming studies in the domain. To achieve that, studies done in optimisation problems using symbiotic organisms search from 2014 – 2020 that are obtained from some databases (Scopus, ScienceDirect, IEEE Xplore, ACM) were surveyed; where review of various issues related to SOS such as diversity of solution search space, variants, scalability, and applications of the SOS. Finally, future research directions in the area were recommended. Key words: 1. cloud computing, nature-inspired algorithms, optimisation, symbiotic organisms search Introduction Optimisation techniques were used to solve complex problem in many fields of study such as IT, Engineering, Management and Telecommunications. Those techniques include heuristic and metaheuristic, the former is found to be inefficient due to nature as well as evolution of the emerging optimisation problems (Abdullahi et al., 2016). Metaheuristic techniques proved to be promising because of its ability to handle a more complex problem due its explorative and exploitive nature. Furthermore, there are many nature-inspired algorithms used in solving optimisation problems, such as Genetic Algorithm (GA), Particle Swarm Optimisation (PSO), Ant Colony Optimisation (ACO), Symbiotic organisms search (SOS) and their variants. These algorithms have shown to be promising in tackling optimisation problems in different fields of study, but SOS outperforms other optimization techniques because of its simplicity in terms of mathematical operation of the three phases, parameter-less features (Abdullahi et al., 2017; Abdullahi & Ngadi, 2016; Choe et al., 2018; Do et al., 2018). Moreover, SOS algorithm been one of such nature-inspired techniques that mimics the relationship between organisms in an ecosystem, it exhibits three phases of the relationship which are; mutualism, commensalism and parasitism (Abdullahi et al., 2019; Cheng & Prayogo, 2014; Gharehchopogh et al., 2019). Despite the achievements of the SOS algorithm and its variants there are still some issues related to entrapment into local optima due to insufficient exploitation at the local search space of the algorithm (Do et al., 2018). Hence, the realisation of high computational cost and convergence rate. In order to overcome some of the issues raised, Systematic Literature Review and Meta-Analysis, 2021 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 there is the need to diversify the local solution search space at the mutualism phase of the SOS algorithm. The remaining part of the paper is as follows: Section 2 explains on the related literature, Section 3 expounds on the structure of the SOS algorithm, different variants and improved version of the SOS algorithm is explained in Section 4, Section 5 elaborates on the applications of SOS, presentation of discussions is in Section 6 and finally Section 7 conclusion and future works. 2. METHODOLOGY For this systematic search, a search strategy was developed to identify relevant literature. This search strategy was basically tailored to the following databases: Scopus, ScienceDirect, IEEE Xplore and ACM. And the search terms used were “Symbiotic Organisms Search Algorithm” OR “Optimis(z)ation” OR “Cloud Computing” OR “Task Scheduling” OR “Heuristics”. All the searches spanned from 2014 until 2020 and included journal articles published in English language only. Further, the selection criteria were based on the PRISMA Statement (Moher et al., 2009). The search mainly focused on mapping the existing literature on usage of symbiotic organism search algorithm in the fields of Computer Science; Engineering; Mathematics; Energy; Business, Management and Accounting; Decision Science; Materials Science; Physical and Astronomy; Social Sciences; and Chemical Engineering. Then the search was narrowed to the subject area of task scheduling optimisation problems in cloud computing environment. The research span was from 2014 – 2020. All articles before 2014 were excluded from the search. The search was mainly focused on experimental journal articles, thus review articles, conference papers and research reports were excluded. Therefore, based on the exclusion criteria mention above 22,445 research articles were excluded from a total of 23,231 articles that were extracted at this stage. Consequently, this study is based only on original research articles where experiments were conducted. Hence, for maintenance of quality of the review, all duplicate publications were thoroughly checked to remove them. Abstracts of the articles were as well reviewed deeply for analysis and purification of the articles to ensure the quality and related academic literature included in the review process. A careful evaluation of each research article was carried out at a later state. The next exclusion criteria were to limit the articles published in English language only. There were two articles (Scopus) in non-English language and were excluded from the study. Moreover, after the filtration of duplicate records, 97 more articles were also removed from the study. Then after assessing each article based on the inclusion and exclusion criteria 110 were selected. Figure 1 shows the literature inclusion and exclusion at every stage (PRISMA statement). a. b. c. d. The articles must be original papers, while review papers, conference papers, book series, books, reports, and case studies were excluded. The articles must in English language and from the IT and Computer Science. The extracted articles were published between 2014 to 2020 from Scopus, ScienceDirect, IEEE Xplore and ACM databases. The extracted papers were from all over the globe. 2 Identification Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 Records identified through database searching (n = 23,231) Addition records identified through other sources (n = 0) Screening Records after duplicates removed (n = 23,231) Records screened (n = 788) Records excluded (n = 22,443) #Full-text articles excluded with reasons (n = 22,443) Full-text articles assessed for eligibility (n =788) #No access to full text (n = 3,164) Eligibility #Non-English articles (n = 02) #Review/Survey (09), Editorial (178), Introduction (129), Conference proceedings (104), Opinion (31), Section (9), Erratum (3), Books/book series (6), Early access article (3) Studies included qualitative assessment (n = 211) #Full text excluded with reason (n = 97) Included #Duplicate (25) Studies included in data extraction (n= 110) Figure 1: Flow of information through different stages of the SLR (Moher et al., 2009). In the data extraction state, 110 articles were selected, and the characteristics extracted were as follows: 3 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 3. 3. RESULTS AND INTERPRETATIONS In order to interpret the results, descriptive analysis and literature classification were employed. Based on the data obtained, descriptive analysis was done by year of publications, journal wise over years, also by the number of citations of the articles, number of citations wise, methodology, environment, whether it experimental paper or not, databases wise and finally by the subject areas or fields of study. As it is shown in Figures 2 to 8. In Figure 2, it illustrates the various research articles distribution over the years under review. Where it shows that the research domain is really growing especially from 2016 onward as such there is prospect in the field. Figure 2: Research article distributions over the years The journal wise distribution displayed in Figure 3, shows a form of consistency with some journals having more than 2000 articles under average - in the domain of nature-inspired algorithms specifically using the symbiotic organisms search strategy. This is an indication for rapid acceptability of the study area in recent time, which really becomes a motivational factor for making researchers to venture in this line. The fact remain every researcher needs his/her paper to be published, hence, the potentiality of the field. Figure 3: Journal wise distributions of the articles Also, journal wise distribution as shown in Figure 3, it can be seen that there are many publications in the area of optimisation using SOS algorithm as we have at least 200 publications in almost all the journals across the year under review. As such the area is really a viable one in terms of research. 4 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 Figure 4: Number of citations wise distributions of the articles Figure 4 depicted the number of citations of the related articles in the field of study under review. Furthermore, it is illustrating how the area is so much cited within the period in question. Thus, the field is worth delving into to conduct research in it. Figure 5: Field of study wise distributions of the articles over years Field of study wise distributions of the various articles over the years in question is been shown in Figure 5. Here we can vividly deduce that over 2,500 publications in area like antenna array, capacitated vehicle routing, data mining, economic load dispatch, engineering applications, parallel machine scheduling and so on. Furthermore, it is more glaring in the areas of cloud computing, energy, power/power systems where we have more than 12,000 publications in those. Consequently, the need for further research in nature-inspired algorithm using the SOS algorithm. 5 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 Figure 6: Experimental studies distributions of the articles Figure 6 depicts the form in which the research was conducted which is experimental. And indicates that predominantly works are done experimentally in nature-inspired algorithms through utilisation of SOS algorithm. As such it is recommended that those partaking into research in this field should endeavour to adopt experimental methodology as shown in the Figure 6, more 95 percent of previous works done used that. Figure 7: Environment of the studies distributions over years Figure 7 is showing the environment of the studies distribution over the period under review. Where it is indicating that simulation is the highest with over 130,000 publications followed by Mathematical model. Therefore, it is apparent the dominance of implementation via simulation in the domain of optimisation using meta-heuristic algorithms such as SOS. 6 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 Figure 8: Paper types of the studies distributions over years Figure 8 is the paper types of the studies distributions over years under consideration. Based on the Figure, we can see that experimental articles and its variances significantly outweighs all other ones. Hence, signifying a tremendous usage of that methodology in the domain of optimisation in nature-inspired algorithms and SOS algorithm seems to be at the front burner. 4. 4. DISCUSSION Based on the results obtained and as shown in section 3 that is Result Section, it can be deduced that the area of task scheduling in cloud computing environment using nature-inspired algorithms is a promising one. As such many studies are under-taken not only the field of computer science, but it also cut across so many fields such as Engineering, Marketing, Science, Technology and so on. 5. 5. FUTURE AGENDA The main commendations for SLRs that are like this, comprise expanding the search filters to transverse more databases. Consequently, this could avail additional related works. Precisely, this study reviewed only full-text online articles. Accordingly, the findings are limited to the studies included. The future research agenda will be based on relaxing the search criteria to incorporate other scholastic databases for further comparative results. In addition, future studies could relax the search filters to include more books, conference papers, and so on. Noteworthy, to build on or replicate the reviewed studies, future research should explore other intelligent bibliographic techniques for in-depth analysis of the existing literature. 6. REFERENCES Abdullahi, M., & Ngadi, M. A. (2016). Hybrid symbiotic organisms search optimization algorithm for scheduling of tasks on cloud computing environment. PLoS ONE, 11(6), 1–29. https://doi.org/10.1371/journal.pone.0158229 Abdullahi, M., Ngadi, M. A., & Abdulhamid, S. M. (2016). Symbiotic Organism Search optimization based task scheduling in cloud computing environment. Future Generation Computer Systems, 56, 640–650. https://doi.org/10.1016/j.future.2015.08.006 Abdullahi, M., Ngadi, M. A., & Dishing, S. I. (2017). Chaotic Symbiotic Organisms Search for Task Scheduling Optimization on Cloud Computing Environment. 1–4. Abdullahi, M., Ngadi, M. A., Dishing, S. I., Abdulhamid, S. M., & Ahmad, B. I. eel. (2019). An efficient symbiotic organisms search algorithm with chaotic optimization strategy for multi-objective task scheduling problems in cloud computing environment. Journal of Network and Computer Applications, 133, 60–74. https://doi.org/10.1016/j.jnca.2019.02.005 Cheng, M. Y., & Prayogo, D. (2014). Symbiotic Organisms Search: A new metaheuristic optimization algorithm. Computers and Structures, 139, 98–112. https://doi.org/10.1016/j.compstruc.2014.03.007 7 Systematic Literature Review and Meta-Analysis, Vol 3, No. 1 Choe, S., Li, B., Ri, I., Paek, C., Rim, J., & Yun, S. (2018). Improved Hybrid Symbiotic Organism Search Task-Scheduling Algorithm for Cloud Computing. KSII TRANSACTIONS ON INTERNET AND INFORMATION SYSTEMS, 12(8), 3516– 3541. Do, D. T. T., Lee, D., & Lee, J. (2018). Material optimization of functionally graded plates using deep neural network and modified symbiotic organisms search for eigenvalue problems. Composites Part B: Engineering, 159(June 2018), 300–326. https://doi.org/10.1016/j.compositesb.2018.09.087 Gharehchopogh, F. S., Shayanfar, H., & Gholizadeh, H. (2019). A comprehensive survey on symbiotic organisms search algorithms. In Artificial Intelligence Review (Issue 0123456789). Springer Netherlands. https://doi.org/10.1007/s10462-01909733-4 Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred Reporting Items for Systematic Reviews and MetaAnalyses: The PRISMA Statement (Reprinted from Annals of Internal Medicine). Physical Therapy, 89(9), 873–880. https://doi.org/10.1371/journal.pmed.1000097 8