International Journal of Grid and Utility Computing (20 papers in press)

Regular Issues

• Recommendation system based on space-time user similarity
  by Wei Luo, Zhihao Peng, Ansheng Deng 
  Abstract: With the advent of 5G, the way people get information and the means of information transmission have become more and more important. As the main platform of information transmission, social media not only brings convenience to people's lives, but also generates huge amounts of redundant information because of the speed of information updating. In order to meet the personalised needs of users and enable users to find interesting information in a large volume of data, recommendation systems emerged as the times require. Recommendation systems, as an important tool to help users to filter internet information, play an extremely important role in both academia and industry. The traditional recommendation system assumes that all users are independent. In this paper, in order to improve the prediction accuracy, a recommendation system based on space-time user similarity is proposed. The experimental results on Sina Weibo dataset show that, compared with the traditional collaborative filtering recommendation system based on user similarity, the proposed method has better performance in precision, recall and F-measure evaluation value.
  Keywords: time-based user similarity; space-based user similarity; recommendation system; user preference; collaborative filtering.
  
  
• Design and analysis of novel hybrid load-balancing algorithm for cloud data centres  
  by Ajay Dubey, Vimal Mishra 
  Abstract: In recent the pandemic scenario there is a paradigm shift, from traditional computing to internet-based computing. Now is the time to store and compute the data in the cloud environment. The Cloud Service Providers (CSPs) establish and maintain a huge shared pool of computing resources that provide scalable and on-demand services around the clock without geographical restrictions. The cloud customers are able to access the services and pay according to the accession of resources. When millions of users across the globe connect to the cloud for their storage and computational needs, there might be issues such as delay in services. This problem is associated with load balancing in cloud computing. Hence, there is a need to develop effective load-balancing algorithms. The Novel Hybrid Load Balancing (NHLB) algorithm proposed in this paper manages the load of the virtual machine in the data centre. This paper is focused on certain problems such as optimisation of performance, maximum throughput, minimisation of makespan, and efficient resource use in load balancing. The NHLB algorithm is more efficient than conventional load-balancing algorithms with reduced completion time (makespan) and response time. This algorithm equally distributes the tasks among the virtual machines on the basis of the current state of the virtual machines and the task time required. The paper compares the result of proposed NHLB algorithm with dynamic load-balancing algorithm and honeybee algorithm. The result shows that the proposed algorithm is better than the dynamic and honeybee algorithms.
  Keywords: cloud computing; data centre; load balancing; virtual machine; makespan; performance optimisation.
  
  
• Cloud infrastructure planning considering the impact of maintenance and self-healing routines over cost and dependability attributes  
  by Carlos Melo, Jean Araujo, Jamilson Dantas, Paulo Pereira, Felipe Oliveira, Paulo Maciel 
  Abstract: Cloud computing is the main trend regarding internet service provision. This paradigm, which emerged from distributed computing, gains more adherents every day. For those who provide or aim at providing a service or a private infrastructure, much has to be done, costs related to acquisition and implementation are common, and an alternative to reduce expenses is to outsource maintenance of resources. Outsourcing tends to be a better choice for those who provide small infrastructures than to pay some employees monthly to keep the service life cycle. This paper evaluates infrastructure reliability and the impact of outsourced maintenance over the availability of private infrastructures. Our baseline environments focus on blockchain as a service; however, by modelling both service and maintenance routines, this study can be applied to most cloud services. The maintenance routines evaluated by this paper encompass a set of service level agreements and some particularities related to reactive, preventive, and self-healing methods. The goal is to point out which one has the best cost-benefit for those with small infrastructures, but that still plans to provide services over the internet. Preventive and self-healing repair routines provided a better cost-benefit solution than traditional reactive maintenance routines, but this scenario may change according to the number of available resources that the service provider has.
  Keywords: maintenance; reliability; availability; modelling; cloud Ccmputing; blockchain; container; services; SLA.
  
  
• Data collection in underwater wireless sensor networks: performance evaluation of FBR and epidemic routing protocols for node density and area shape  
  by Elis Kulla 
  Abstract: Data collection in Underwater Wireless Sensor Networks (UWSN) is not a trivial problem, because of unpredictable delays and unstable links between underwater devices. Moreover, when nodes are mobile, continuous connectivity is not guaranteed. Therefore, data collection in UWSN Node scarcity and movement patterns create different environments for data collection in underwater communication. In this paper, we investigate the impact of the area shape and node density in UWSN, by comparing Focused Beam Routing (FBR) and Epidemic Routing (ER) protocols. Furthermore, we also analyse the correlation between different performance metrics. From simulation results we found that when using FBR, delay and delivery probability slightly decrease (2.1%) but the overhead ratio decreases noticeably (46.9%). The correlation between performance metrics is stronger for square area shape, and is not noticeable for deep area shape.
  Keywords: underwater wireless sensor networks; focused beam routing; delay tolerant network; area shape; node density; data collection.
  
  
• Joint end-to-end recognition deep network and data augmentation for industrial mould number recognition  
  by RuiMing Li, ChaoJun Dong, JiaCong Chen, YiKui Zhai 
  Abstract: With the booming manufacturing industry, the significance of mould management is increasing. At present, manual management is gradually eliminated owing to need for a large amount of labour, while the effect of a radiofrequency identification (RFID) system is not ideal, which is limited by the characteristics of the metal, such as rust and erosion. Fortunately, the rise of convolutional neural networks (CNNs) brings down to the solution of mould management from the perspective of images that management by identifying the digital number on the mould. Yet there is no trace of a public database for mould recognition, and there is no special recognition method in this field. To address this problem, this paper first presents a novel data set aiming to support the CNN training. The images in the database are collected in the real scene and finely manually labelled, which can train an effective recognition model and generalise to the actual scenario. Besides, we combined the mainstream text spotter and the data augmentation specifically designed for the real world, and found that it has a considerable effect on mould recognition.
  Keywords: mould recognition database; text spotter; mould recognition; data augmentation.
  
  
• An SMIM algorithm for reduction of energy consumption of virtual machines in a cluster  
  by Dilawaer Duolikun, Tomoya Enokido, Makoto Takizawa 
  Abstract: Applications can take advantage of virtual computation services independently of heterogeneity and locations of servers by using virtual machines in clusters. Here, a virtual machine on an energy-efficient host server has to be selected to perform an application process. In this paper, we newly propose an SMI (Simple Monotonically Increasing) estimation algorithm to estimate the energy consumption of a server to perform application processes and the total execution time of processes on a server. We also propose an SMIM (SMI Migration) algorithm to make a virtual machine migrate from a host server to a guest server to reduce the total energy consumption of the servers by estimating the energy consumption in the SMI algorithm. In the evaluation, we show the energy consumption of servers in a cluster can be reduced in the SMIM algorithm compared with other algorithms.
  Keywords: server selection algorithm; migration of virtual machines; green computing systems; SMI algorithm; SMIM algorithm.
  
  
• FIAC: fine-grained access control mechanism for cloud-based IoT framework  
  by Bhagwat Prasad Chaudhury, Kasturi Dhal, Srikant Patnaik, Ajit Kumar Nayak 
  Abstract: Cloud computing technology provides various computing resources on demand to the user on pay per use basis. The users use the services without the need for establishment and maintenance costs. The technology fails in terms of its usage owing to confidentiality and privacy issues. Access control mechanisms are the tools to prevent unauthorised access to remotely stored data. CipherText Policy Attribute-based Encryption (CPABE) is a widely used tool for facilitating authorised users to access the remotely stored encrypted data with fine-grained access control. In the proposed model FIAC (Fine-Grained Access Control Mechanism for cloud-based IoT Framework ), the access control mechanism is embedded in the cloud-based application to measure and generate a report on the air quality in a city. The major contribution of this work is the design of three algorithms all of which are attribute based: key generation algorithm, encryption and decryption algorithms. Only authorised users can view it to take appropriate action plans. Carbon dioxide concentrations, dust, temperature, and relative humidity are the parameters that we have considered for air quality. To enhance the security of the cloud-based monitoring system, we have embedded a security scheme, all of which are attribute based. Further, the computation time of the model is found to be encouraging so that it can be used in low power devices. The experimental outcomes establish the usability of our model.
  Keywords: air pollution; carbon dioxide concentrations; dust density; internet of things; FIAC; access control.
  
  
• University ranking approach with bibliometrics and augmented social perception data  
  by Kittayaporn Chantaranimi, Rattasit Sukhahuta, Juggapong Natwichai 
  Abstract: Typically, universities aim to achieve a high position in ranking systems for their reputation. However, self-evaluating rankings could be costly because the indicators are not only from bibliometrics, but also the results of over a thousand surveys. In this paper, we propose a novel approach to estimate university rankings based on traditional data, i.e., bibliometrics, and non-traditional data, i.e., Altmetric Attention Score, and Sustainable Development Goals indicators. Our approach estimates subject-areas rankings in Arts & Humanities, Engineering & Technology, Life Sciences & Medicine, Natural Sciences, and Social Sciences & Management. Then, by using Spearman rank-order correlation and overlapping rate, our results are evaluated by comparing with the QS subject ranking. From the result, our approach, particularly the top-10 ranking, performed estimating effectively and then could assist stakeholders in estimating the university's position when the survey is not available.
  Keywords: university ranking; rank similarity; bibliometrics; augmented social perception data; sustainable development goals; Altmetrics.
  
  
• Strongly correlated high utility item-sets recommendation in e-commerce applications using EGUI-tree over data stream  
  by P. Amaranatha Reddy, M. H. M. Krishna Prasad, S. Rao Chintalapudi 
  Abstract: The product recommendation feature helps the customers to select the right items in e-commerce applications. Grounded on a similar kind of customer purchase history previously, it recommends the items to the customers. The customers may or may not choose from the recommended items list. Suppose the customers like and purchase them, it is added advantage to both the buyers and the sellers. In sellers' sales, they get the benefit of an increase; in addition, the buyers too can save their time to search. Thus, a recommendation system must be designed so that the suggested items have High Utility (HU) and strong correlation to meet such requirements. High profit is exhibited by the HU items and the strongly correlated items have more probability of selection. As these measurements have significant roles in the business process, both of those measurements of items needed to be taken care. Here, since up-to-date data in the data stream are got, the stream of purchase transactions is mined using the sliding window technique to extract such item-sets.
  Keywords: high utility item-set mining; recommendation system; utility; correlation; data stream; sliding window.
  DOI: 10.1504/IJGUC.2023.10057772
   
  
• Enhancing fruit farming efficiency through IoT-driven soil moisture analysis and classifier ensemble  
  by Chinmayee Senapati, Swagatika Senapati, Satyaprakash Swain 
  Abstract: Effective soil moisture management is a key to optimising fruit farming productivity. This research work model integrates IoT devices, clustering techniques, PCA and ensemble learning to enhance soil moisture classification for crops like pomegranate, mulberry, mango, grapes, ragi and potato. Among various classifiers tested, the Random Forest model demonstrated superior accuracy. However, a stacked Random Forest-SVM model further improved accuracy to 94.65%. This research underscores the importance of IoT-driven data and machine learning in precision agriculture, demonstrating how advanced techniques can refine soil moisture management. By optimising soil moisture, we directly enhance nutrient availability, root development, and water uptake, leading to better crop yield, quality and sustainability. This approach highlights the synergy between technology and machine learning, advancing sustainable and efficient fruit cultivation
  Keywords: IoT; PCA; NBM; RBF; K-star; RF; SVM; KNN; GNB; DT; clustering.
  DOI: 10.1504/IJGUC.2024.10066245
   
  
• Hybrid optimisation-based reliable routing with traffic management and congestion control  
  by Kaveri Kori, Sridevi Hosmani 
  Abstract: Reliable routing with traffic management in VANET is the most critical aspect that to be solved on an emergency basis. There is a need for a trust computing system adapted to the peculiarities of VANETs to deal with the significant factors of road safety. The purpose of this work is to suggest a new improved trust-based methodology for dependable transmission. To transmit the data packet efficiently, the cluster head selection and optimal routing process takes place. The Fuzzy C-Means Clustering (FCM) technique is employed to select the cluster head, in which the node with high energy is elected as the cluster head. Followed by, a novel hybrid optimization named the Beluga whale-assisted Coati Optimization (BwACO) algorithm is proposed, which is the combination of the Beluga Whale Optimization (BWO) and Coati Optimization Algorithm (COA). The greatest mobility offered using the BWACO approach is -146.386 at the median statistical metric.
  Keywords: Trust-based protocol; Traffic; Energy; Distance; Mobility.
  DOI: 10.1504/IJGUC.2024.10067188
   
  
• Improved data integrity audit scheme based on certificateless public key cryptography and its application in COVID-19 epidemic data management  
  by Xiaoxuan Xu, Yuanyou Cui, Dianhua Tang, Yunfei Cao, Jindan Zhang 
  Abstract: With the continuous development and improvement of cloud computing and big data technology, the cloud audit has gradually become more and more important. For the huge data storage, how to carry out data integrity auditing in the edge environment is very critical. This paper studies and explores the data integrity auditing scheme based on certificateless public key cryptography in the edge environment, analyze the security flaws of the original scheme in the signature generation process, and improves it. The security analysis shows that the new scheme can effectively resist the original attack and three types of adversary in certificateless cryptography, and the computational overhead is reduced in the auditing phase. Finally we discuss its application in COVID-19 epidemic data management.
  Keywords: edge computing; certificateless auditing; data integrity.
  DOI: 10.1504/IJGUC.2024.10067681
   
  
• Optimising DNNs for load forecasting: the power of hyperparameter tuning  
  by Faisal Mehmood Butt, Seong-O Shim, Safa Habibullah, Abdulwahab Ali Almazroi, Lal Hussain, Umair Ahmed Salaria 
  Abstract: This study demonstrates the effectiveness of deep learning for electricity demand forecasting across various timescales Interestingly, the optimal network structures differed depending on the prediction horizon We explored configurations with varying numbers of neurons (similar to adjusting hidden layer units) For one-day forecasts, a double hidden layer network with ReLU and sigmoid activation functions achieved the lowest Mean Absolute Percentage Error (MAPE) of 4 23%, with just 4 neurons in the hidden layers In contrast, a single hidden layer with ReLU activation and 6 neurons yielded the best MAPE of 1 95% for one-week forecasts For longer timescales,more complex architectures were necessary The one-month forecast achieved a MAPE of 2 78% with a double ReLU-sigmoid network and 12 neurons in the hidden layers Even three-month forecasts, a challenging task, were tackled effectively by a double ReLU network with 10 neurons, resulting in a MAPE of 2 75%.
  Keywords: optimisation; ReLU; rectified linear unit; convolutional neural; networks; deep neural networks; signature function; neurons; layers.
  DOI: 10.1504/IJGUC.2024.10067738
   
  
• Low-powered IoT device for monitoring dementia patients using A9G module and ESP32 C3 MCU in cloud environment  
  by Kumar Saurabh, Manish Madhava Tripathi, Satyasundara Mahapatra 
  Abstract: As the growing age of world's population, more and more people are affected by a disease known as dementia that causes gradual mental decline. Owing to this, not only the patient but also the caregiver's personal life and health system faces challenges. To keep the system smart, an intelligent monitoring system is needed. Development of an optimised IoT device with the latest sensors and microcontrollers which stores the data in cloud environment in systematic manner is the basic objective. This research presents a methodology which optimizes a normal monitoring device into advanced IoT system using A9G Module and ESP32 C3 MCU. After the optimisation, the evaluation of the device is based on various qualitative parameters and overall effectiveness in real-world applications. The findings suggest that while simple IoT devices offer indoor monitoring, the proposed device provide comprehensive monitoring which enhance the patient safety and caregiver peace of mind.
  Keywords: dementia; efficient utilisation; IoT based monitoring; patient care; data analysis; privacy; wearable devices.
  DOI: 10.1504/IJGUC.2024.10067985
   
  

Special Issue on: AMLDA 2022 Applied Machine Learning and Data Analytics Applications, Challenges, and Future Directions

• Fuzzy forests for feature selection in high-dimensional survey data: an application to the 2020 US Presidential Election  
  by Sreemanti Dey, R. Michael Alvarez 
  Abstract: An increasingly common methodological issue in the field of social science is high-dimensional and highly correlated datasets that are unamenable to the traditional deductive framework of study. Analysis of candidate choice in the 2020 Presidential Election is one area in which this issue presents itself: in order to test the many theories explaining the outcome of the election, it is necessary to use data such as the 2020 Cooperative Election Study Common Content, with hundreds of highly correlated features. We present the fuzzy forests algorithm, a variant of the popular random forests ensemble method, as an efficient way to reduce the feature space in such cases with minimal bias, while also maintaining predictive performance on par with common algorithms such as random forests and logit. Using fuzzy forests, we isolate the top correlates of candidate choice and find that partisan polarisation was the strongest factor driving the 2020 Presidential Election.
  Keywords: fuzzy forests; machine learning; ensemble methods; dimensionality reduction; American elections; candidate choice; correlation; partisanship; issue voting; Trump; Biden.
  
  
• An efficient intrusion detection system using unsupervised learning AutoEncoder  
  by N.D. Patel, B.M. Mehtre, Rajeev Wankar 
  Abstract: As attacks on the network environment are rapidly becoming more sophisticated and intelligent in recent years, the limitations of the existing signature-based intrusion detection system are becoming more evident. For new attacks such as Advanced Persistent Threat (APT), the signature pattern has a problem of poor generalisation performance. Research on intrusion detection systems based on machine learning is being actively conducted to solve this problem. However, the attack sample is collected less than the normal sample in the actual network environment, so it suffers a class imbalance problem. When a supervised learning-based anomaly detection model is trained with these data, the results are biased toward normal samples. In this paper, AutoEncoder (AE) is used to perform single-class anomaly detection to solve this imbalance problem. The experimental evaluation was conducted using the CIC-IDS2017 dataset, and the performance of the proposed method was compared with supervised models to evaluate the performance
  Keywords: intrusion detection system; advanced persistent threat; CICIDS2017; AutoEncoder; machine learning; data analytics.
  
  
• Optimal attack detection using an enhanced machine learning algorithm  
  by Reddy Saisindhutheja, Gopal K. Shyam, Shanthi Makka 
  Abstract: As computer network and internet technologies advance, the importance of network security is widely acknowledged. Network security continues to be a substantial challenge within the cyberspace network. The Software-as-a-Service (SaaS) layer describes cloud applications where users can connect using web protocols such as hypertext transfer protocol over transport Layer Security. Worms, SPAM, Denial-of-service (DoS) attacks or botnets occur frequently in the networks. In this light, this research intends to introduce a new security platform for SaaS framework, which comprises two major phases: 1) optimal feature selection and (2) classification. Initially, the optimal features are selected from the dataset. Each dataset includes additional features, which further leads to complexity. A novel algorithm named Accelerator updated Rider Optimisation Algorithm (AR-ROA), a modified form of ROA and Deep Belief Network (DBN) based attack detection system is proposed. The optimal features that are selected from AR-ROA are subjected to DBN classification process, in which the presence of attacks is determined. The proposed work is compared against other existing models using a benchmark dataset, and improved results are obtained. The proposed model outperforms other traditional models, in aspects of accuracy (95.3%), specificity (98%), sensitivity (86%), precision (92%), negative predictive value (97%), F1-score (86%), false positive ratio (2%), false negative ratio (10%), false detection ratio (10%), and Matthews correlation coefficient (0.82%).
  Keywords: software-as-a-service framework; security; ROA; optimisation; DBN; attack detection system.
  DOI: 10.1504/IJGUC.2022.10064192
   
  

Special Issue on: Cloud and Fog Computing for Corporate Entrepreneurship in the Digital Era

• Enhanced speculative approach for big data processing using BM-LOA algorithm in cloud environment  
  by Hetal A. Joshiara, Chirag S. Thaker, Sanjay M. Shah, Darshan B. Choksi 
  Abstract: In the event that one of the several tasks is being allocated to an undependable or jam-packed machine, a hugely parallel processing job can be delayed considerably. Hence, the majority of the parallel processing methodologies, namely (MR), have espoused diverse strategies to conquer the issue called the straggler problem. Here, the scheme may speculatively introduce extra copies of a similar task if its development is unnaturally slow when an additional idling resource is present. In the strategies-centred processes, the dead node is exhibited. the (RT) along with backup time of the slow task is assessed. The slow task is rerun with the aid of BM-LOA subsequent to the evaluation. In both heterogeneous and homogeneous environments, the proposed approach is performed. Centred on the performance metrics, the proposed research techniques performance is scrutinised in experimental investigation. Thus, when weighed against the other approaches, the proposed technique achieves superior performance.
  Keywords: modified exponentially weighted moving average; speculative execution strategy; Hadoop supreme rate performance; big data processing; rerun.
  DOI: 10.1504/IJGUC.2022.10063581
   
  
• Importance of big data for analysing models in social networks and improving business  
  by Zhenbo Zang, Honglei Zhang, Hongjun Zhu 
  Abstract: The digital revolution is the build-up of many digital advances, such as the network phenomena transformation. To increase business productiveness and performance, participatory websites that allow active user involvement and the intellectual collection became widely recognized as a value-added tool organization of all sizes. However, it has lacked an emphasis on business and management literature to promote profitable business-to-business (b2b). User knowledge is a way to consider the kinds of search ions that big data can use. The use of social media data for analysing business networks is a promising field of research. Technological advances have made the storage and analysis of large volumes of data commercially feasible. Big information represents many sources of organised, half-structured, and unstructured data in real-time. Big data is a recent phenomenon in the computing field, which involves technological or market study. The management and research of a vast network provide all companies with significant advantages and challenges. The volume of information floating on social networks grows every day and, if processed correctly, provides a rich pool of evidence.
  Keywords: big data; business-to-business; data analytics; text mining; social networks.
  
  
• Study on the economic consequences of enterprise financial sharing model  
  by Yu Yang, Zecheng Yin 
  Abstract: Using enterprise system ideas to examine the business process requirements of firms, the Financial Enterprise Model (FEM) is a demanding program. This major integrates finance, accounting, and other critical business processes. Conventional financial face difficulties due to low economic inclusion, restricted access to capital, lack of data, poor R&D expenditures, underdeveloped distribution channels, and so on. This paper mentions making, consuming, and redistributing goods through collaborative platform networks. These three instances highlight how ICTs (Information and Communication Technologies) can be exploited as a new source of company innovation. The sharing economy model can help social companies solve their market problems since social value can be embedded into their sharing economy cycles. As part of the ICT-based sharing economy, new business models for social entrepreneurship can be developed by employing creative and proactive platforms. Unlike most public organizations, double-bottom-line organizations can create social and economic advantages. There are implications for developing and propagating societal values based on these findings.
  Keywords: finance; economy; enterprise; ICT; social advantage.