Improving the Processing Speed of Task Scheduling in Cloud

Computing Using the Resource Aware Scheduling Algorithm over the

Max-Min Algorithm

P. Sushma Priya and S. John Justin Thanjaraj

Department of Computer Science and Engineering, Saveetha Institute of Medical and Technical Sciences,

Saveetha University, Chennai, Tamil Nadu, India

Keywords: Cloud Sim, Grid Sim, Max-Min Algorithm, Min-Min Algorithm, Novel Resource Aware Scheduling,

Processing Speed, Resources, Cloud Computing.

Abstract: The study aimed to bolster the efficiency of task scheduling in cloud computing through the novel Resource

Aware Scheduling Algorithm (RASA) and contrasted its performance with the Max-Min Algorithm. Cloud

computing, with its decentralised nature, distributes internet-based resources, handling a diverse array of

demands. For this research, data for task scheduling was sourced from the Cloudsim Tool. The cloud

infrastructure was analysed, designed, and implemented to test both RASA and the Max-Min Algorithm,

utilising 120 samples in two distinct groups. The processing speed outcomes, analysed via IBM SPSS, showed

RASA to be notably superior. The independent sample t-test had a significance of 0.007(p < 0.05),

highlighting the distinctiveness of the algorithms. Impressively, RASA's processing speed outstripped that of

the Max-Min Algorithm.

1 INTRODUCTION

The processing speed of task scheduling improves by

organising incoming requests in a particular way,

leading to a reduction in execution time (Nakum,

Ramakrishna, and Lathigara 2014; Peng and Wolter

2019). Cloud computing offers numerous advantages,

including quality of service (QoS), job scheduling,

data storage, productivity improvements, and robust

security measures. Effective job scheduling and

system load management have become pivotal issues

in the cloud computing domain (Heidari and Buyya

2019). These challenges can be adeptly tackled

through the deployment of appropriate task

scheduling algorithms. The Max-Min algorithm, for

instance, functions within the cloud computing

framework by estimating the expected completion

time of tasks based on the extant resources. Task

scheduling aims to diminish execution times and

costs, thereby optimising performance. Analyses

indicate that task scheduling can enhance processing

speed (Kakaraparthi and Karthick 2022; AS, Vickram

et al. 2013). This boosts the populace's quality of life

by providing enhanced security and flexibility. Using

the same dataset, the performance of the resource-

aware scheduling algorithm is compared to a

traditional scheduling method, specifically the max-

min algorithm. Gridsim is utilised as a simulation

framework for modelling and simulating distributed

systems. This comparison unequivocally shows the

superior performance of the new resource-aware

scheduling algorithm over the max-min algorithm

(Bandaranayake et al. 2020). Cloud computing

applications encompass online data storage and

services like Gmail (Sen 2017).

Over 100 articles have been showcased on

ResearchGate. The Dynamic Adaptive Particle

Swarm Optimisation (DAPSO) technique, designed

to enhance the original PSO algorithm's functionality,

reduces a task set's makespan and augments resource

utilisation (Rosić et al. 2022). Due to its user-

friendliness and efficacy across numerous

applications, Particle Swarm Optimisation (PSO) has

seen rising popularity. This study's aim is to assess

cloud computing's impact on individuals' quality of

life. A novel task scheduling algorithm is introduced

to address historical challenges associated with the

min-min and max-min algorithms (Aref, Kadum, and

Kadum 2022; Vijayan et al. 2022). The RAS

(Resource Aware Scheduling Algorithm) algorithm

alternately applies the Max-Min and Min-Min tactics

for task allocation to resources. Within the Gridsim

Priya, P. and Thanjaraj, S.

Improving the Processing Speed of Task Scheduling in Cloud Computing Using the Resource Aware Scheduling Algorithm over the Max-Min Algorithm.

DOI: 10.5220/0012544200003739

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Artiﬁcial Intelligence for Internet of Things: Accelerating Innovation in Industry and Consumer Electronics (AI4IoT 2023), pages 589-594

ISBN: 978-989-758-661-3

589

Figure 1: Bar chart showing the comparison of mean Processing Speed and standard errors.

framework, the Resource Aware Scheduling

Algorithm (RASA) is employed. Practical Gridsim

demonstrations highlight the merits of using RASA to

schedule independent tasks within grid environments.

The ACO algorithm capitalises on the generic

algorithm's global search function to pinpoint the

optimal solution, subsequently translating it into the

ACO's initial pheromone. Simulation results suggest

this algorithm outpaces genetic algorithms and ACO

under comparable conditions, outshining them in

larger-scale contexts. It's a potent job scheduling

strategy for cloud computing (Aref, Kadum, and

Kadum 2022; Guo 2017). The Improved Max-Min, a

refined max-min method, was conceived by this

algorithm. Rather than choosing the largest task, this

method selects tasks of median or close sizes, which

are then delegated to the resource with the shortest

completion time. Consequently, this algorithm

diminishes the overall makespan whilst ensuring

workload is evenly distributed across resources (Aref,

Kadum, and Kadum 2022; Guo 2017; Ming and Li

2012).

Previous task scheduling algorithms have

suffered from slow processing speeds. An innovative

algorithm known as RASA utilises a blend of Max-

Min and Min-Min tactics in an alternating fashion to

distribute tasks amongst available resources. There's

a discernible research gap due to the scant focus on

energy-efficient scheduling algorithms and diverse

workload patterns. Moreover, there exists a demand

for scheduling algorithms adept at handling multiple

objectives, including accelerating response times,

conserving energy, and guaranteeing equitable

resource allocation. The study's primary goal is to

craft the RASA Algorithm to boost processing speed.

2 MATERIALS AND METHODS

The research presented in this article was undertaken

at the cloud computing laboratory within the

Department of Computer Science and Engineering at

the Saveetha Institute of Medical and Technical

Sciences in Chennai. Initial assessments were made

using pre-tests with a G-power set at 80%, a

significance threshold of 0.05%, and a 95%

confidence interval (Zhang et al. 2021).

The study comprised two groups, each containing

120 participants, making a total of 240 participants.

Group 1 applied the RASA Algorithm, whereas

Group 2 employed the Max-Min Algorithm. The

experimental setup consisted of the Eclipse IDE for

Java, SPSS version 26.0.1, and a laptop equipped

with 8GB RAM, an Intel Gen i5 Processor, and a 4GB

Graphics card. The study accommodated various

types of data, encompassing documents, images, and

videos. Additionally, Gridsim was utilised, a tool

designed for modelling and simulating elements in

parallel and distributed computing systems. This tool

covered a range of entities, including resources,

applications, users, and resource brokers or

schedulers (Udayasankaran and Thangaraj 2023).

2.1 Group 1: Novel Resource Aware

Scheduling Algorithm

The Novel Resource Aware Scheduling Algorithm

(RASA) is a cutting-edge approach in cloud

computing, cleverly oscillating between the Max-Min

and Min-Min strategies to proficiently allocate tasks

among the available resources. By taking into account

AI4IoT 2023 - First International Conference on Artiﬁcial Intelligence for Internet of things (AI4IOT): Accelerating Innovation in Industry

and Consumer Electronics

590

aspects such as the size of the task and the time a

resource takes to complete it, RASA strives to

maximise processing speed. The goal is to curtail the

overall duration tasks remain in the system, ensuring

a balanced workload distribution, thereby elevating

the efficacy of task execution in cloud ecosystems.

The practical efficacy of RASA, especially when

orchestrating standalone tasks within grid systems,

has been substantiated through real-world Gridsim

demonstrations.

To implement this, one should follow the below steps:

Step 1: Procure and install the necessary software

packages.

Step 2: Integrate your dataset into the chosen coding

environment.

Step 3: Extract the requisite files from Gridsim.

Step 4: Initiate the execution of the program.

Step 5: Incorporate the Resource Aware Scheduling

Algorithm (RASA).

Step 6: Ascertain the parameters needed for the model

to ensure an optimal fit.

Step 7: Delve deeper into the analysis, especially

focusing on gauging processing speed metrics.

2.2 Group 2: Max-Min Algorithm

In the realm of distributed systems, this algorithm is

frequently utilised. For each task, the anticipated

completion time is determined considering the

available resources. The algorithm allocates tasks to

these resources based on their projected completion

times, with a central objective of achieving an even

distribution of workload across the resources. The

primary aim is to amplify overall efficiency by

curtailing the completion time of the most substantial

tasks, thereby diminishing the collective processing

duration.

Step 1: Set out the objective function.

Step 2: Initialise with starting values.

Step 3: Ascertain the maximum value.

Step 4: Identify the minimum value.

Step 5: Compute the intermediary values.

Step 6: Undergo repeated iterations.

Step 7: Distribute the resources appropriately.

3 STATISTICAL ANALYSIS

The study utilised IBM SPSS version 26.0.1, a

statistical analysis software, to evaluate the data

(George and Mallery 2021). After ensuring the

datasets were normalised, they were structured into

arrays. The most appropriate number of clusters was

pinpointed and then assessed in conjunction with the

extant algorithms. It was observed that as the number

of tasks allocated to the RASA algorithm rose, there

was a decline in the error rate, leading to an

enhancement in processing speed. Within the dataset,

the task size acted as an independent variable, whilst

both task bandwidth and size were treated as

dependent variables. The independent T-Test was

employed to scrutinise the findings of the research.

4 RESULTS

Figure 1 presents a bar chart comparing the

Processing Speed of two algorithms across varying

sample sizes. The x-axis designates the algorithms,

and the y-axis delineates the Processing Speed.

Evidently, the Novel RASA algorithm surpasses the

Max-Min Algorithm in this aspect.

Table 1 provides a breakdown of statistical

outcomes based on all iteration variables. The RASA

Algorithm consistently outshines the Max-Min

Algorithm. The standard deviation for RASA stands

at 1.3984, while for the Max-Min Algorithm it is

4.62161.

Table 1: Number of epochs taken for the RASA and Max-

Min Algorithms are 20. Mean value for Group 1 is 3.7430

and Group 2 is 8.4090.

ALGORITHM

(Number

of Epochs)

Mean

Standard

Deviation

Standard

Mean

Error

PROCESSING

SPEED (MIPS)

RASA

3.7430

1.39834

1.44219

PROCESSING

SPEED (MIPS)

MAX-MIN

8.4090

4.62161

1.46148

The standard error mean for RASA is recorded at

0.44219 and it's 1.46148 for the Max-Min Algorithm.

Subsequent findings from a random sample test can

be perused in Table 2. Subjecting the dataset to an

Independent Sample T-Test at a 95% confidence

interval reveals the Novel Resource Aware

Scheduling Algorithm's distinct edge over the Max-

Min Algorithm.

Resource Aware Scheduling Algorithm. Resource

Aware Scheduling Algorithm is better than Max-Min

in terms of mean processing speed and standard

deviation. X-Axis: Resource Aware Scheduling

Algorithm vs Max-Min Algorithm Y-Axis: Mean

Processing Speed of detection ± 2 SD.

When comparing processing speeds between

RASA and the Max-Min algorithm, both exhibit a

mean difference of 4.6 and a similar disparity in

standard deviation. The Max-Min Algorithm's 95%

Improving the Processing Speed of Task Scheduling in Cloud Computing Using the Resource Aware Scheduling Algorithm over the

Max-Min Algorithm

591

Table 2: Independent sample T-Test is applied for dataset fixing confidence intervals as 95%. The independent sample t test

significance shown after the test is 0.007 (p<0.05) (Resource Aware Scheduling Algorithms appear to perform better than

Max-Min Algorithm).

Levene’s test for equality of variables

sig

Processing speed

(MIPS)

Equal variances assumed

21.781

.000

Processing speed

(MIPS)

Equal variance not assumed

Sig(2-tailed)

Processing speed

(MIPS)

Equal variances assumed

0.07

Processing speed

(MIPS)

Equal variance not assumed

0.07

T-test for equality of Means

Mean Difference

Std.error Difference

Processing speed

(MIPS)

Equal variances assumed

-4.66

1.52

Processing speed

(MIPS)

Equal variance not assumed

-4.66

1.52

95% Confidence Interval of Difference

Lower

Upper

Processing speed

(MIPS)

Equal variances assumed

-7.87

-1.45808

Processing speed

(MIPS)

Equal variance not assumed

-4.66

-1.29113

confidence interval showcases a range reaching

7.873. Following the analysis, the significance level

of the independent sample t-test emerges as 0.007 (p

< 0.05), pointing to a pronounced difference in

variance, which strengthens RASA's standing. In

terms of processing speed, a comparative

examination between the two algorithms has been

undertaken. The algorithms have also undergone

distinct evaluations based on bandwidth parameters.

It is conclusive that the Novel RASA significantly

surpasses the Max-Min Algorithm in performance.

5 DISCUSSION

The insights gleaned from this study confirm that the

Resource Aware Scheduling Algorithm is more

proficient than the Max-Min Algorithm, especially in

terms of processing speed. The independent sample t-

test's significance post-analysis stands at 0.007 (p <

0.05), underscoring the Novel Resource Aware

Scheduling Algorithm's marked superiority over the

Max-Min Algorithm. Parallel findings corroborate

the advancements in task scheduling's processing

speed. For instance, an article details the novel

resource-aware scheduling algorithm underpinning

task scheduling (Buakum and Wisittipanich 2022).

The evaluation juxtaposes the optimized TSA with

other algorithms such as Existing Min-Min, Existing

Max-Min, RASA, Improved Max-Min, and

Enhanced Max-Min, employing Java 7 for the

simulation (Aref, Kadum, and Kadum 2022). The

time each task is anticipated to take, based on

available resources, is meticulously calculated

(Gandomi et al. 2020). Within this framework, the

scheduler's responsibility is to methodically organise

all tasks within the earmarked meta-tasks in light of

the resources on hand (Kamalam and Sentamilselvan

2022). As the volume of tasks and samples swells,

processing speed follows suit, while the error rate

inversely contracts. The propounded algorithm

judiciously deploys one of these strategies, contingent

on the context and computed parameters, thereby

amalgamating the merits of other seasoned task

scheduling methodologies (Aref, Kadum, and Kadum

2022). Cloud computing not only offers cost-

effectiveness, particularly beneficial to small and

medium businesses, but also augments the quality of

life. RASA endeavours to strike a balance, mitigating

AI4IoT 2023 - First International Conference on Artiﬁcial Intelligence for Internet of things (AI4IOT): Accelerating Innovation in Industry

and Consumer Electronics

592

the pitfalls of both approaches while capitalising on

their strengths (Bandaranayake et al. 2020).

Factors such as makespan, load balancing, and

cloud computing — which inherently deals with an

expansive array of resources and requests — are

pivotal determinants in task scheduling (Etminani and

Naghibzadeh 2007; Tang 2018). Deploying a dataset

abundant with tasks could bolster the assertion of

time efficiency, all the while preserving an elevated

processing speed. For future ventures, transplanting

the proposed method directly into a genuine cloud

computing milieu (via cloud sim) for myriad

empirical evaluations, including scalability,

resilience, and accessibility, could further refine task

scheduling (Gandomi et al. 2020).

6 CONCLUSION

In the article under consideration, the study

underscores the efficiency of the Novel Resource

Aware Scheduling Algorithm, which boasts a

processing speed of 3.7430 MIPS, outpacing the

Max-Min Algorithm that operates at 8.4090 MIPS.

The empirical results derived from the Gridsim

simulator unequivocally establish RASA's

dominance over the Max-Min algorithm, especially

when deployed within extensive distributed systems.

REFERENCES

Aref, Ismael Salih, Juliet Kadum, and Amaal Kadum.

(2022). “Optimization of Max-Min and Min-Min Task

Scheduling Algorithms Using G.A in Cloud

Computing.” 2022 5th International Conference on

Engineering Technology and Its Applications

(IICETA).

https://doi.org/10.1109/iiceta54559.2022.9888542.

AS, Vickram, Raja Das, Srinivas MS, Kamini A. Rao, and

Sridharan TB, (2013). "Prediction of Zn concentration

in human seminal plasma of Normospermia samples by

Artificial Neural Networks (ANN)." Journal of assisted

reproduction and genetics 30: 453-459.

Bandaranayake, K. M. S. U., K. M. S. Bandaranayake, K.

P. N. Jayasena, and B. T. G. Kumara. (2020). “An

Efficient Task Scheduling Algorithm Using Total

Resource Execution Time Aware Algorithm in Cloud

Computing.” 2020 IEEE International Conference on

Smart Cloud (SmartCloud).

https://doi.org/10.1109/smartcloud49737.2020.00015.

Buakum, Dollaya, and Warisa Wisittipanich. (2022).

“Selective Strategy Differential Evolution for

Stochastic Internal Task Scheduling Problem in Cross-

Docking Terminals.” Computational Intelligence and

Neuroscience 2022 (November): 1398448.

Etminani, K., and M. Naghibzadeh. 2007. “A Min-Min

Max-Min Selective Algorihtm for Grid Task

Scheduling.” (2007) 3rd IEEE/IFIP International

Conference in Central Asia on Internet.

https://doi.org/10.1109/canet.2007.4401694.

Gandomi, Amir H., Ali Emrouznejad, Mo M. Jamshidi,

Kalyanmoy Deb, and Iman Rahimi. (2020).

Evolutionary Computation in Scheduling. John Wiley

& Sons.

George, Darren, and Paul Mallery. (2021). IBM SPSS

Statistics 27 Step by Step: A Simple Guide and

Reference. Routledge.

Guo, Qiang.(2017). “Task Scheduling Based on Ant

Colony Optimization in Cloud Environment.” AIP

Conference Proceedings.

https://doi.org/10.1063/1.4981635.

Heidari, Safiollah, and Rajkumar Buyya. (2019). “Quality

of Service (QoS)-Driven Resource Provisioning for

Large-Scale Graph Processing in Cloud Computing

Environments: Graph Processing-as-a-Service

(GPaaS).” Future Generation Computer Systems.

https://doi.org/10.1016/j.future.2019.02.048

Kakaraparthi, Aditya, and V. Karthick. (2022). “A Secure

and Cost-Effective Platform for Employee

Management System Using Lightweight Standalone

Framework over Diffie Hellman’s Key Exchange

Algorithm.” ECS Transactions 107 (1): 13663–74.

Kishore Kumar, M. Aeri, A. Grover, J. Agarwal, P. Kumar,

and T. Raghu, “Secured supply chain management

system for fisheries through IoT,” Meas. Sensors, vol.

25, no. August 2022, p. 100632, 2023, doi:

10.1016/j.measen.2022.100632.

Kamalam, G. K., and K. Sentamilselvan. (2022). “SLA-

Based Group Tasks Max-Min (GTMax-Min)

Algorithm for Task Scheduling in Multi-Cloud

Environments.” Operationalizing Multi-Cloud

Environments. https://doi.org/10.1007/978-3-030-

74402-1_6.

Ming, Gao, and Hao Li. (2012). “An Improved Algorithm

Based on Max-Min for Cloud Task Scheduling.”

Recent Advances in Computer Science and Information

Engineering. https://doi.org/10.1007/978-3-642-

25789-6_32.

Nakum, Sunilkumar, C. Ramakrishna, and Amit Lathigara.

2014. “Reliable RASA Scheduling Algorithm for Grid

Environment.” Proceedings of IEEE International

Conference on Computer Communication and Systems

ICCCS14. https://doi.org/10.1109/icccs.2014.7068183.

Peng, Guang, and Katinka Wolter. (2019). “Efficient Task

Scheduling in Cloud Computing Using an Improved

Particle Swarm Optimization Algorithm.” Proceedings

of the 9th International Conference on Cloud

Computing and Services Science.

https://doi.org/10.5220/0007674400580067.

Rosić, Maja, Miloš Sedak, Mirjana Simić, and Predrag

Pejović. (2022). “Chaos-Enhanced Adaptive Hybrid

Butterfly Particle Swarm Optimization Algorithm for

Passive Target Localization.” Sensors 22 (15).

https://doi.org/10.3390/s22155739.

Improving the Processing Speed of Task Scheduling in Cloud Computing Using the Resource Aware Scheduling Algorithm over the

Max-Min Algorithm

593

Sen, Jaydip. (2017). Cloud Computing: Architecture and

Applications. BoD – Books on Demand.

Tang, Ling. (2018). “Load Balancing Optimization in

Cloud Computing Based on Task Scheduling.” 2018

International Conference on Virtual Reality and

Intelligent Systems (ICVRIS).

https://doi.org/10.1109/icvris.2018.00036.

Udayasankaran, P., and S. John Justin Thangaraj. (2023).

“Energy Efficient Resource Utilization and Load

Balancing in Virtual Machines Using Prediction

Algorithms.” International Journal of Cognitive

Computing in Engineering 4 (June): 127–34.

Vijayan, D. S., Mohan, A., Nivetha, C., Sivakumar, V.,

Devarajan, P., Paulmakesh, A., & Arvindan, S. (2022).

Treatment of pharma effluent using anaerobic packed

bed reactor. Journal of Environmental and Public

Health, 2022.

V. P. Parandhaman, (2023)"An Automated Efficient and

Robust Scheme in Payment Protocol Using the Internet

of Things," Eighth International Conference on Science

Technology Engineering and Mathematics

(ICONSTEM), Chennai, India, 2023, pp. 1-5, doi:

10.1109/ICONSTEM56934.2023.10142797.

AI4IoT 2023 - First International Conference on Artiﬁcial Intelligence for Internet of things (AI4IOT): Accelerating Innovation in Industry

and Consumer Electronics

594