A computational model for ranking cloud service providers using hypergraph based techniques

doi:10.1016/j.future.2016.08.014

Future Generation Computer Systems

Volume 68, March 2017, Pages 14-30

https://doi.org/10.1016/j.future.2016.08.014 Get rights and content

Highlights

•
This paper presents a Hypergraph based Computational Model (HGCM) for ranking CSPs.
•
HGCM uses hypergraph based technique (Minimum Distance-Helly Property (MDHP)) to evaluate the CSPs.
•
MDHP exploits the relation between the Service Measurement Index (SMI) metrics to select appropriate CSPs based on CUs requirement.
•
HGCM uses arithmetic residue and Expectation–Maximization (EM) algorithms to impute missing values.
•
Complexity of HGCM has been reduced through exploitation of the Helly property.

Abstract

In a cloud marketplace, the existence of wide range of Cloud Service Providers (CSPs) makes it hard for the Cloud Users (CUs) to find an appropriate CSP based on their requirements. The design of a suitable service selection framework helps the users in the selection of a suitable CSP, while motivating the CSPs to satisfy the assured Service Level Agreement (SLA) and enhance the Quality of Service (QoS). Existing service selection models employ random assignment of weights to the QoS attributes, replacement of missing data by random values, etc. which results in an inaccurate ranking of the CSPs. Moreover, these models have high computational overhead. In this study, a novel cloud service selection architecture, Hypergraph based Computational Model (HGCM) and Minimum Distance-Helly Property (MDHP) algorithm have been proposed for ranking the cloud service providers. Helly property of the hypergraph had been used to assign weights to the attributes and reduce the complexity of the ranking model, while arithmetic residue and Expectation–Maximization (EM) algorithms were used to impute missing values. Experimental results provided by MDHP under different case studies (dataset used by various research communities and synthetic dataset) confirms the ranking algorithm to be scalable and computationally attractive.

Introduction

‘Cloud computing’—an internet-based technology has changed the way through which the computing resources were accessed and utilized by the users [1], [2]. Based on the user’s requirements, the Cloud Service Provider (CSP) models the requested resources and offers them in the form of cloud services. Computing resources can be modeled using the three cloud service models namely, Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS) [3]. IaaS abstracts the physical hardware (server, network, etc.) in the form of virtual servers and storages, thereby provide the Cloud Users (CUs) with an environment to deploy, run and monitor them (e.g. Amazon Web Services (AWS), Google Compute Engine (GCE), Windows Azure, etc.). PaaS provides a platform on top of the abstracted hardware to develop cloud applications (e.g. Google App Engine, Apprenda, etc.). SaaS provides the entire application as a service, enabling the CUs to overlook the suspicions about the infrastructure, platform and application installation (e.g. Google Apps, Citrix GoToMeeting, Cisco WebEx, etc.). From the CUs viewpoint, IaaS offers greater flexibility and minimum application automation, in comparison with the other two service delivery models [4].

Support provided by the Information Technology (IT) infrastructure becomes mandatory for organizations up and down the scale. For large organizations, setting up and maintaining an IT infrastructure (servers, network cables, storage, cooling infrastructure, etc.) might be as easy as pie. Small and medium enterprises might perceive it as a ‘hard nut to crack’, since it requires skilled manpower (developers, network administrators, system administrators, etc.) and high capital investment on IT infrastructure. Cloud computing aims to deliver appropriate services based on Quality of Service (QoS) requirements at competitive costs to the CUs in different organizations [4]. CUs can access these services from anywhere, at any time on a ‘Pay as You Go’ fashion. Therefore, it is unnecessary for small and medium sized organizations to invest on hardware and manpower for delivering business services. To summarize, cloud computing offers overwhelming benefits (flexibility, disaster recovery, software updates, no capital investment, etc.) to a variety of business organizations, by providing liberation from the hitches in the task of setting up an IT infrastructure and enabling them to concentrate on an innovative way to enhance their business service values [3]. In order to exploit the various benefits offered by cloud computing, various organizations have started developing their own applications on cloud infrastructures. These merits in turn attract many organizations to move on to the cloud in order to provide their business solutions for a large-scale user community. However, moving an existing business model to cloud involves numerous challenges with respect to the unique requirements and characteristics of different applications.

Conventionally, computing resources have been purchased or leased from the data centers and the users were billed irrespective of their usage. Emergence of cloud computing enables the users to access computing as a basic utility similar to water, electricity, gas, etc. where the consumers pay only for the resource(s) utilized [4]. Evolution of this technology produces a wide range of public cloud services which varies with respect to their features, performance, and pricing levels. Nevertheless, identification of an appropriate CSP who can satisfy their QoS requirements becomes harder on the CUs end as there exists a trade-off between various functional and non-functional requirements. Hence, it is important for the CUs to evaluate and find a suitable CSP for a service request rather than discovering multiple CSPs.

Security and trust management are the two major thrust areas for future research in the field of cloud computing. Trust Management System (TMS) helps the CUs to find an appropriate CSP with expected QoS [5]. Trust of a service assessed through the existing trust evaluation systems insinuates the security and QoS level offered by various CSPs [6]. In general, TMS consists of components such as cloud service discovery, trust metrics selection and measurement, trust assessment, trust evolution and trust based ranking model to assess the trustworthiness of any CSP (Fig. 1) [7]. Out of these five components, ranking models which have been used to rank the CSPs, play a vital role in the entire life cycle of TMS. This paper, emphasizes the importance of service selection mechanism (ranking models) designed to prioritize CSPs based on their likelihood to the CUs requirements.

Selection of CSPs who match with the maximum set of functional and non-functional requirements requested by the CUs is a decision problem. This problem is similar to Multi-Criteria Decision-Making (MCDM), as complex decision making process involves multiple attributes and interdependent relationships among them [8]. In the present study, a novel cloud service selection architecture with Hypergraph based Computational Model (HGCM) and Minimum Distance-Helly Property (MDHP) ranking algorithm have been proposed to address the problem of service selection. The proposed MDHP ranking algorithm ranks and selects the most suitable CSPs from the available pool of CSPs. Further, Cloud Service Measurement Index Consortium—Service Measurement Index (CSMIC—SMI) has been used as standard metric [9] to assess different CSPs based on the CUs requirements. CSMIC—SMI was launched by Carnegie Mellon University as a standard measure to evaluate any service based on the user’s requirements.

In this paper, Section 2 describes the CSMIC—SMI, novel cloud service selection architecture and quality model for cloud services. Section 3 enlightens the basics of hypergraph with its properties, HGCM, MDHP ranking algorithm along with its complexity analysis and missing data imputation techniques. Section 4 deals with the performance analysis of the MDHP algorithm using various case studies. Section 5 concludes the paper with future works.

Section snippets

Service Measurement Index (SMI)

Traditional High Performance Computing (HPC) metrics and benchmarks, focusing on performance and cost [10] cannot be applied to cloud environment due to its distributed and dynamic nature. This prompted many standard bodies to frame benchmark tools like Information and Communication Technology Service Quality (ICTSQ), ISO/IEC 9126, Application Performance Index (APDEX), eSourcing Capability Model—Client Organizations (eSCM-CL) and SMI to evaluate different services [11]. SMI is a hierarchical

Hypergraph preliminaries

Let $Y = {y_{1}, y_{2}, \dots, y_{n}}$ be a finite set. A hypergraph $Y$ is a family $H = {E_{1}, E_{2}, \dots, E_{n}}$ of subsets of $Y$ such that [35], [36] $E_{i} \neq ϕ$ $⋃ E_{i} = Y; i = 1, 2, \dots, m .$ The elements $y_{1}, y_{2}, \dots, y_{n}$ of $Y$ are called vertices and the sets $E_{1}, E_{2}, \dots, E_{m}$ are called the hyperedges of the hypergraphs. A simple graph is a simple hypergraph, each of whose edges have a cardinality 2. A simple hypergraph or Sperner family is a hypergraph $H = {E_{1}, E_{2}, \dots, E_{m}}$ such that $E_{i} \subset E_{j}$ implies $i = j$ . Hypergraph provides exciting facilities to represent multiple

Case study—ranking CSPs using hypergraph based computational model

Various case studies presented in this paper mainly focus on the SMI metrics. Service indices of the SMI metrics were calculated using the QoS data of the various CSPs (public and private cloud set up).

Conclusion and future work

Cloud computing spans across various organizations to satisfy the computational demands put forth by diversified users. Tremendous growth in the area of cloud computing had led to increase in the number of CSPs, service offerings and user expectations. Hence, finding an appropriate CSP fulfilling the QoS requirement of the CUs, remains a challenge. To overcome this issue, a CU needs to identify and measure some standard set of performance metrics relevant to their application. Therefore, CSMIC

Acknowledgments

The first and third author thank the Department of Science and Technology, India for INSPIRE Fellowship (Grant No: DST/INSPIRE Fellowship/2013/963) and Fund for Improvement of S&T Infrastructure in Universities and Higher Educational Institutions (SR/FST/ETI-349/2013) for their financial support. The second author thanks the Department of Science and Technology—Fund for Improvement of S&T Infrastructure in Universities and Higher Educational Institutions Government of India (SR/FST/MSI-107/2015

Glossary

Highlights

HGCM: Hypergraph based Computational Model
MDHP: Minimum Distance-Helly Property
CSP(s): Cloud Service Provider(s)
CU(s): Cloud User(s)
SMI: Service Measurement Index
EM: Expectation–Maximization

Abstract

SLA: Service Level Agreement
GIS: Geographic Information System

Keywords

TMS: Trust Management System

Introduction

IaaS: Infrastructure as a Service
PaaS: Platform as a Service
SaaS: Software as a Service
AWS: Amazon Web Services
GCE: Google Compute Engine
QoS: Quality of Service
MCDM: Multi-Criteria Decision-Making
CSMIC: Cloud Service Measurement Index Consortium
HPC: High

Nivethitha Somu is a Full time Ph.D. Scholar at School of Computing, SASTRA University, Thanjavur, INDIA. She is a member of Ramanujan Mathematical Society (Mem. No: 1198). She received her Master’s degree in Science from Anna University, Chennai, INDIA in 2011. She also received her Master’s degree in Technology from SASTRA University, Thanjavur, INDIA in 2013. Her current research interests include trust management, energy aware workload consolidation and secure live migration in cloud.

References (48)

R. Buyya et al.
Cloud computing and emerging IT platforms: vision, hype, and reality for delivering computing as the 5th utility
Future Gener. Comput. Syst.
(2009)
Jagpreet Sidhu et al.
Compliance based trustworthiness calculation mechanism in cloud environment
Procedia Comput. Sci.
(2014)
S.K. Garg et al.
A framework for ranking of cloud computing services
Future Gener. Comput. Syst.
(2013)
O.S. Vaidya et al.
Analytic hierarchy process: an overview of applications
European J. Oper. Res.
(2006)
R. Dharmarajan et al.
A hypergraph-based algorithm for image restoration from salt and pepper noise
AEU-Int. J. Electron. Commun.
(2010)
Alain Bretto et al.
Hypergraph imaging: an overview
Pattern Recognit.
(2002)
Alain Bretto et al.
An efficient algorithm for Helly property recognition in a linear hypergraph
Electron. Notes Theor. Comput. Sci.
(2001)
Kirthivasan Kannan et al.
Root mean square filter for noisy images based on hyper graph model
Image Vis. Comput.
(2010)
A. Sivakumar et al.
A novel feature selection technique for number classification problem using PNN—A plausible scheme for boiler flue gas analysis
Sensors Actuators B
(2009)
Frank John Krautheim
Building trust into utility cloud computing
(2010)

Felix N. Njeh

Cloud computing: An evaluation of the cloud computing adoption and use model

(2014)

Peter Mell et al.

The NIST Definition of Cloud Computing

(2011)

Mohamed Firdhous et al.

Applying bees algorithm for trust management in cloud computing

D. Marudhadevi et al.

A trust evaluation model for cloud computing using service level agreement

Comput. J.

(2014)

Milan Zeleny et al.

Multiple Criteria Decision Making

(1973)

CSMIC Consortium, Service Measurement Index Version 1.0, Carnegie Mellon University Silicon Valley Moffett Field, CA...

C. Binnig, D. Kossmann, T. Kraska, S. Loesing, How is the weather tomorrow?: towards a benchmark for the cloud, in:...

P.M.A.C. Costa

Evaluating cloud services using multicriteria decision analysis

(2013)

V.Neelaya Dhatchayani et al.

Trust aware identity management for cloud computing

Int. J. Inf. Comm. Technol.

(2014)

Robert John Hennan, John G. Roane, Security monitoring tool for computer network. US Patent 7,904,456, issued March 8,...

Wolfgang. Barth

Nagios: System and Network Monitoring

(2008)

Lie Qu et al.

Cloud service selection based on contextual subjective assessment and objective assessment

Sushil Singh et al.

Trust evaluation in cloud based on friends and third party’s recommendations

Cited by (61)

A decision framework with nonlinear preferences and unknown weight information for cloud vendor selection
2023, Expert Systems with Applications
Cloud vendor selection (CVS) is a complex decision-making problem, which actively adheres to human behavior/cognition. The complex nature of the problem is due to personal biases/hesitation, trade-offs among attributes, uncertainty in rating, and the nonlinear relationship among cloud vendors and associated attributes. In recent times, researchers started paying more attention to user/expert behavior, which led to non-linear decision-making. Most of the extant decision models for CVS considered the linear form of decision-making, which is not realistic due to expert opinions' complexity and dynamism. Motivated by the claim, in this paper, a non-linear decision approach is put forward for CVS. Likert scale rating is adopted for rating cloud vendors based on some attributes, which are transformed to polynomial space from the linear fuzzy space. After this, weights of attributes are determined by using CRITIC in the non-linear space. Following this, cloud vendors are ranked in a personalized fashion using the proposed algorithm that encompasses the WASPAS procedure and rank fusion schemes. Finally, a case study is exemplified to validate the usefulness of the decision approach. Comparison and sensitivity analysis showcases the efficacy and robustness of the developed approach.
A Fuzzy-Multi Attribute Decision Making approach for efficient service selection in cloud environments
2022, Expert Systems with Applications
Citation Excerpt :
From (Alam et al., 2018), it is clear that state-of-the-art approaches for solving the CSS problem in the light of objective-subjective based trust assessment falls into the logic, description, Multi Attribute Decision Making (MADM), optimization, QoS prediction, reputation models, and recommendation system (Ding et al., 2018). However, the complex and dynamic relationship between the QoS attributes and the service quality necessitates continuous research developments in the design of efficient cloud service selection approaches based on recent advancements and technologies (Nivethitha et al., 2019; Somu, Gauthama Raman et al., 2018; Somu et al., 2017). In this way, recent research works have explored the benefits of artificial intelligence techniques like Artificial Neural Networks (ANN) and its variants, Naives Bayes, decision trees, Random Forest (RF), Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), etc. for accurate CSS through modeling the non-linear complex relationships in the QoS data (Marrone, 2015; Mohanty et al., 2010; Nguyen & Bai, 2018; Somu, Gauthama et al., 2018; Zhang et al., 2016).
In general, objective and subjective QoS assessment data forms the major data source for service evaluations. Especially, in the most dynamic service-oriented computing ecosystem like cloud environments, such assessment data is subjected to high uncertainty due to the dynamics in the number of cloud services and related QoS values. In such cases, the accuracy and reliability of the most efficient cloud service selection model remain questionable. Further, the complex interdependencies among the multiple QoS parameters and the trustworthiness of the cloud services complicate the process of handling uncertainty during cloud service selection. This work presents Fuzzy-Multi Attribute Decision Making (FMADM), a hybrid trust prediction model to provide accurate trust-based cloud service evaluations while efficiently handling the uncertainty in the cloud service assessment data. FMADM employs (i) Picture Fuzzy Sets (PFSs): to capture inconsistency, uncertainty, and ambiguity in the QoS data, (ii) Naïve Bayes: to recompute the weights of the QoS attributes, (iii) Measurement of Alternatives and Ranking according to COmpromise Solution (MARCOS): to model the non-linear relationship between the QoS values of cloud services and their corresponding trust result, and (iv) Random Forest Classifier (RFC): to predict the trust value of cloud services. The performance of FMADM was evaluated using Quality Web Service (QWS) v1.0 dataset under different scenarios using various quality metrics.
Evaluation of cloud vendors from probabilistic linguistic information with unknown/partial weight values
2020, Applied Soft Computing Journal
Citation Excerpt :
Three commonly provided services are platform, software, and infrastructure, which in general is termed as X-aaS, where X is as mentioned above. Google compute engine, and Windows Azure are examples of IaaS, GoToMeeting, and Cisco WebEx are examples of SaaS, and Google App Engine and Apprenda are examples of PaaS [7]. Buyya et al. [4] claimed that cloud services could be acquired as primary sources of life viz., water, and air, by IT sectors for effective management of resources.
As IT industries grow at a faster pace, cloud technology becomes inevitable. Attracted by the scope, many cloud vendors (CVs) arise. A rational/systematic selection is an urge to tackle the scalability of CVs. To circumvent the issue, in this paper, a framework is proposed for CV selection under with probabilistic linguistic term sets (PLTSs). The PLTS is a flexible structure that allows partial ignorance of occurring probabilities. Initially, attributes’ weights are calculated using a programming model, which uses partial information effectively. Later, decision-makers’ (DMs’) weights are computed by integrating evidence theory with Bayes approximation. Preferences from DMs are aggregated by proposing a two-way operator, which aggregates linguistic preferences using the rule-based method and occurring probabilities using Maclaurin symmetric mean. Moreover, CVs are ranked by using an integrated PROMETHEE–Borda method under the PLTS. Further, to test the validity of the framework, a case study on CV selection is presented for a small-scale company. Finally, the advantages and limitations of the proposed framework are investigated by comparison with other methods and the results infer that (i) the proposed framework is 63.67% robust even after adequate changes are made to the alternatives; (ii) the proposed framework is 87.67% robust even after adequate changes are made to the attributes; (iii) from partial adequacy test, the robustness is determined as 77.67% and 92.33%; and (iv) from the broadness test, the proposed framework produces an average deviation of 9% among their rank values, which is better than the extant models that produce an average deviation close to 7.8%.
IBGSS: An Improved Binary Gravitational Search Algorithm based search strategy for QoS and ranking prediction in cloud environments
2020, Applied Soft Computing Journal
Citation Excerpt :
The competitive cloud marketplace has enabled a wide range of CSPs to offer functionally similar cloud services at different performance and price which makes the identification of optimal service, an open challenge for the cloud users [2]. The existence of a multitude of cloud services with similar functionalities poses a significant challenge in the selection of suitable services to build high quality applications [3–7]. For example, consider a service composition problem, where a consumer needs to construct an e-commerce application which consists of five unique services, namely e-commerce, manufacture, payment, logistic, and storage.
Quality of Service (QoS) value prediction and QoS ranking prediction have their significance in optimal service selection and service composition problems. QoS based service ranking prediction is an NP-Complete problem which examines the order of ranked service sequence with respect to the unique QoS requirements. To address the NP-Complete problem, greedy and optimization-based strategies such as CloudRank and PSO have been widely employed in service oriented environments. However, they pose several challenges with respect to the similarity measure based QoS prediction, trap at local optima, and near optimal solution. Hence, this paper presents Improved Binary Gravitational Search Strategy (IBGSS), an optimization based search strategy to address the challenges in the state-of-the-art QoS value prediction and service ranking prediction techniques. IBGSS employs improved cosine similarity measure, and Newton–Raphson inspired Binary Gravitational Search Algorithm (NR-BGSA) for accurate QoS value prediction and optimal service ranking prediction respectively. The effectiveness of IBGSS over the state-of-the-art QoS value prediction and ranking prediction techniques was validated using two real world QoS datasets, namely WSDream#1 and web service QoS dataset in terms of various statistical measures (Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Average Precision Correlation (APC)).
A Decision Aiding Framework for Two-Way Selection of Cloud Vendors in Medical Centers With Generalized Orthopair Fuzzy Information
2023, IEEE Transactions on Engineering Management
Investigating Bio Cloud-Oriented University Tax Accounting Specifications and Advancements in University AccountingInformationSystems
2023, Journal of Commercial Biotechnology

View all citing articles on Scopus

Kannan Kirthivasan is a Professor in the Department of Mathematics, SARTRA University, Thanjavur, INDIA. He obtained his Bachelor’s and Master’s degrees from the University of Madras, India, in 1980 and 1982, respectively. He also received his Bachelor’s and Master’s degrees in Education from Madurai Kamaraj University, India, in 1984 and 1986 respectively. He obtained his M.Phil degree in Mathematics from Regional Engineering College, Tiruchirapalli, India, in 1988. He was conferred Ph.D. in Mathematics in the area of Computational Fluid Dynamics by Alagappa University, Karaikudi, India, in 2000. He has been in Academia for the past 25 years. His specific areas of interest include Combinatorial Optimization, Artificial Neural Networks and Hypergraph-based Image Processing.

Shankar Sriram V.S. is an Associate Professor in School of Computing, SASTRA University, Thanjavur, Tamil Nadu, INDIA. He received his Bachelor’s degree in Science from Madurai Kamraj University, Madurai, INDIA in 1997. He obtained his Master’s degree in Computer Applications from Madurai Kamraj University, Madurai, INDIA in 2000. He also received his Master’s degree in Engineering from Thapar University, Punjab, INDIA in 2004. He was conferred Ph.D. in Information and Network Security from Birla Institute of Technology, Mesra, INDIA in 2010. He has been in the Academia for the past 15 years. His current area of research includes Information and network security, Cryptography, MANETS, Steganography and Cloud computing.

View full text

A computational model for ranking cloud service providers using hypergraph based techniques

Highlights

Abstract

Introduction

Section snippets

Service Measurement Index (SMI)

Hypergraph preliminaries

Case study—ranking CSPs using hypergraph based computational model

Conclusion and future work

Acknowledgments

Glossary

Highlights

Abstract

Keywords

Introduction

Future Gener. Comput. Syst.

Procedia Comput. Sci.

Future Gener. Comput. Syst.

European J. Oper. Res.

AEU-Int. J. Electron. Commun.

Pattern Recognit.

Electron. Notes Theor. Comput. Sci.

Image Vis. Comput.

Sensors Actuators B

Building trust into utility cloud computing

Cloud computing: An evaluation of the cloud computing adoption and use model

The NIST Definition of Cloud Computing

Applying bees algorithm for trust management in cloud computing

A trust evaluation model for cloud computing using service level agreement

Comput. J.

Multiple Criteria Decision Making

Evaluating cloud services using multicriteria decision analysis

Trust aware identity management for cloud computing

Int. J. Inf. Comm. Technol.

Nagios: System and Network Monitoring

Cloud service selection based on contextual subjective assessment and objective assessment

Trust evaluation in cloud based on friends and third party’s recommendations