International Journal of Grid and Utility Computing (54 papers in press)
Semantic grid resource discovery based on SKOS ontology
by Nabila Chergui, Salim Chikhi, Tahar Kechadi
Abstract: In large-scale distributed environments such as grids, where the number of nodes is constantly increasing and resources have heterogeneous and dynamic nature, an efficient and scalable resource discovery mechanism is required. Hybrid techniques, which are based on nodes clustering, are considered among solutions that guarantee the efficiency and the scalability at the same time. However, lack of a good and effective representation of nodes in a highly heterogeneous environment often results in an irrelevant allocation of nodes to the right group. This is considered to be one of the reasons why queries can fail to find relevant resources. Combining hybrid techniques with semantic web technologies would bring more benefits to the resource discovery process. This paper introduces a semantic clustering of nodes based on their domains of interest to form groups called federations. It presents a new process for constructing federations and a three-layer overlay network, in addition to a mechanism of routing queries to target federations in an efficient and a scalable way. The proposed process is based on a SKOS (Simple Knowledge Organisation System) lightweight ontology that describes domains of applications in the grid. We also conduct extensive simulations to evaluate the performance of the proposed approach and the experimental results demonstrate its efficiency and its ability to scale with the size of the system.
Keywords: resource discovery, grid computing, SKOS, semantic web
Towards a fairer negotiation for dynamic resource allocation in the cloud by relying on trustworthiness
by Mohamed Raouf Habes, Habiba Belleili-Souici
Abstract: The automated negotiation for resource leasing contract in the cloud arises many challenges. Since we incorporate an identity along each participant, future utilities will play a major role in their strategies in a cloud marketplace. In fact, unlike traditional agents strategies, our strategies had to take into account two additional components, namely, future utilities involving the reputation of agents identities as well as transactions fees in a market place destined for dynamic resource allocation for cloud computing. The aim of this paper is to design such strategies that prevent any practice that may cause unfairness to arise owing to some drawbacks in the resource allocation protocol helping some participants to take unfair advantages over others. Comparative experiments with a non-trust model show the effectiveness of our approach in terms of fairness and social welfare.
Keywords: cloud computing; negotiation strategy; multi-agent; trust; electronic marketplace
Multi-level power consumption model and energy-aware server selection algorithm
by Hiroki Kataoka, Shigenari Nakamura, Dilawaer Duolikun, Tomoya Enokido, Makoto Takizawa
Abstract: Electric power consumed by servers has to be reduced in a cluster in order to realise the eco society. We take a macro-level approach to reducing the total electric energy consumption of servers to perform application processes in a server cluster. Servers are now equipped with multi-thread CPUs. In this paper, we discuss a multi-level power consumption (MLPC) model of a server with a multi-thread CPU. Here, the power consumption of a server depends on the number of active cores and active threads, where at least one application process is performed. We also discuss a multi-level computation (MLC) model that gives the expected execution time of a process, which is concurrently performed with other processes on a server. Based on the MLPC model and the MLC model, we newly propose an energy-aware (EA) algorithm to select a server for each process requested by a client in a cluster so as to reduce the total electric energy consumption of the cluster while satisfying deadline requirements of the processes. We evaluate the EA algorithm and show that both the total energy consumption and the average execution time of each process are reduced in the EA algorithm compared with the round-robin and random algorithms.
Keywords: multi-level power consumption model; multi-level computation model; energy-aware server selection algorithm
Cloud-aided online/offline ciphertext-policy attribute-based encryption in the standard model
by Hao Wang, Zhihua Zheng, Yilei Wang
Abstract: Attribute-based encryption (ABE) is a useful encryption method that allows users to encrypt messages based on their attributes. However, the computational cost of ABE system is very high, because it often scales with the number of attributes. This makes ABE difficult to be applied in practice. In this paper, we use cloud computing technology as well as pre-computing technology to solve this problem, and introduce the cloud-aided online/offline ciphertext-policy ABE system. In this system, private key generator (PKG) could pre-compute intermediate keys offline before online key generation phase, encryptors could pre-compute intermediate ciphertexts offline before online encryption phase, and decryptors could call their cloud computing servers to transform the ciphertexts to the partially decrypted ciphertexts before the decryption phase. This greatly reduces the local computational cost of these three phases. It is significant for mobile devices to use ABE. Then, we propose a specific scheme, and prove its adaptive security in the standard model. Finally, we introduce how to deploy our new scheme in the mobile cloud computing environment.
Keywords: cloud-aided; CP-ABE; composite order bilinear groups; dual system encryption; adaptively secure
Big-CAF: a bigraphical-generic cloud architecture framework
by Zakaria Benzadri, Chafia Bouanaka, Faïza Belala
Abstract: Cloud computing is an emerging paradigm that enhances grid computing, but in the sense of the utility computing principle. Hence, it attracts more attention in the worlds of both industry and academia. However, there are still many obstacles that are slowing down its adoption and growth. An important and challenging issue in this area is how to associate a clear semantic to cloud architecture concepts. Based on BRS (Bigraphical Reactive Systems) theory, the paper presents a formal framework that provides mathematical definitions of main elements involved in a cloud architecture, specifying its static structure and dynamic evolution. As a consequence of this formalisation, a practical implementation based on Maude and its model checker, is proposed to ensure the correctness of cloud systems.
Keywords: cloud computing; cloud-layered architecture; virtualisation; formal methods; bigraphical reactive systems
Cost-effective task scheduling using hybrid approach in cloud
by Jyoti Thaman, Manpreet Singh
Abstract: Cloud computing is a recent computing paradigm where IT services are provided and delivered over the internet on demand. Increasing concerns about energy costs has not spared the cloud environment from considering energy-conscious solutions. The scheduling problem for cloud computing is widely studied as the applications tasks could be mapped to the available resources to achieve better results. Best effort results from traditional heuristic approaches depend on the tasks and resources at disposal. Particle Swarm Optimisation (PSO), an iteration-based meta-heuristic, performs searches over the solution space and converges in a minimum number of iterations. A variant of PSO has been proposed for various parameters, such as makespan, energy consumption, usage and economic efficiency. The performance of the proposed algorithm has been evaluated and it has been proved to be stable and to find near-optimal solutions.
Keywords: makespan; energy efficiency; cost-aware; heuristic; scheduling;.
Integrated quality of user experience and quality of service approach to service selection in internet of services
by Senthil Murugan Balakrishnan, Arun Kumar Sangaiah
Abstract: Service Oriented Architecture (SOA) approaches are now becoming applicable to embedded devices that feature embedded processing and communication, enabling hosting of services on resource-rich fixed machines to wireless resource-constrained devices and on any physical object boosted with communication capability, thereby creating an environment called Internet of Services (IoS). Discovery and retrieval of the ideal service in such a developed environment comprising a resource-constrained hosting platform is really a challenging job. In this paper, we propose the middleware for IoS that addresses the issue of selecting the best set of non-composite web services that have maximum utility value over Quality of Services (QoS) and Quality of User Experience (QoUE) attributes. The novelty behind the proposed middleware development is in achieving a high degree configuration and customisability by selecting a subset of middleware functionality depending on the need. Particularly, the middleware addresses the service selection issue in the context of non-composite services by pre-evaluating services on the basis of user experience factors. However, the conventional service selection approaches do not consider QoUE and QoS as a whole; they expect the user to provide a set of QoS constraints. This paper addresses the selection issue based on the service usage record and proposes the evaluation routines for QoUE and a decision-making model for QoS. The selection approach is based on the multi-attribute utility theory decision-making method. The experimental results are encouraging, and the middleware performs well when compared with the constraint model and linear programming models.
Keywords: internet of things; service selection; middleware; quality of user experience; quality of service.
An improved multi-instance multi-label learning algorithm based on representative instances selection and label correlations
by Chanjuan Liu, Tongtong Chen, Hailin Zou, Xinmiao Ding, Yuling Wang
Abstract: Multi-instance multi-label learning (MIML) has been successfully used in image and text classification problems. It is noteworthy that few of the previous studies consider the pattern-label relations. Inevitably, there are some useless instances in a bag which will reduce the accuracy of the annotation. In this paper, we focus on this problem. Firstly, an instance selection method via joint-norms constraint is employed to eliminate the useless instances and select the representative instances by modelling the instance correlation. Then, bags are mapped to these representative instances. Finally, the classifier is trained by an optimisation algorithm based on label correlations. Experimental results on image dataset, text datasets and birdsong audio dataset show that the proposed algorithm significantly improves the performance of the MIML classifier compared with the state-of-the-art methods.
Keywords: multi-instance multi-label learning; representative instances selection; joint-norms constraint; label correlations.
A self-adaptive structuring for large-scale P2P grid environment: design and experimental analysis
by Bassirou Gueye, Olivier Flauzac, Cyril Rabat, Ibrahima Niang
Abstract: Resource management is a key issue for large-scale grid environments. In particular, the resource discovery mechanism uses high impacts on the efficiency of resource sharing and cooperative computing. Meanwhile, the increasing size of resources and users in large-scale distributed systems has led to a scalability problem. The self-organising, fault tolerant and decentralised nature of P2P technology, which helps to reduce the management cost of grid infrastructure, is a good basis for resolving both aforementioned problems. In this context, we propose a self-adaptive structuring model for large-scale P2P grid environments. The proposed specification, called P2P4GS, is generic i.e. not linked to a particular P2P architecture. Our structuring approach is completely distributed, and requires only local knowledge about neighbouring nodes to implement a node virtual community. Indeed, given a nodes connection and based on its neighbours, we dynamically create virtual communities or clusters. A particular node called ISP (Information System Proxy) acts as a service directory within each cluster. In addition, to provide an efficient lookup mechanism in our system, we propose to build a spanning tree constituted by the set of ISPs. An experimental validation, through simulation, shows that our approach ensures a high scalability in terms of clusters distribution and communication cost.
Keywords: peer-to-peer systems; grid services; resources management; modelisation; distributed algorithms; clustering; spanning tree; Oversim simulator
Mining top-k approximate closed patterns in an imprecise database
by Yu Xiaomei, Hong Wang, Xiangwei Zheng
Abstract: Over the last few years, the growth of data is exponential, leading to colossal amounts of information being produced by computational systems. Meanwhile, the data in real-life applications are usually incomplete and imprecise, which poses big challenges for researchers to obtain exact and valid analytical results with traditional frequent pattern mining methods. Since the potential faults can break the original characteristics of data patterns into multiple small fragments, it is impossible to recover the long true patterns from these fragments. To explore the huge amount of imprecise data by means of frequent pattern mining, we propose a service-oriented model that enables a new way of service provisioning based on users' QoS (quality of service) requirements.The novel model is developed to solve the problem of mining top-k approximate closed patterns in imprecise databases and will be further applied to diagnosis and treatment of potential patients in online medical applications. We test the novel model in an imprecise medical database and the experimental results show that the new model can successfully improve the health services for online customers.
Keywords: data mining; approximate frequent pattern; frequent closed pattern; clustering; equivalence class; health service.
A comparison between TOSCA and OpenStack Hot through cloud patterns composition
by Antonio Esposito, Beaniamino Di Martino, Giuseppina Cretella
Abstract: Cloud computing is driving formidable change in the technology industry and transforming how enterprises do their business in Europe and around the world. However, despite its spreading, cloud computing is still affected by issues that still need to be addressed carefully. Among these, challenges related to the portability and interoperability of cloud applications often arise, since there is not enough support for the development and deployment of cloud-oriented software on heterogeneous platforms and frameworks. In such a situation, cloud patterns can provide optimal solutions to several issues arising in the cloud computing scenery. However, the lack of a common standard for their representation and of tools to immediately exploit them in developing applications can hamper their usefulness. In this paper we compare two workflow description languages, namely TOSCA and OpenStack HOT, trying to assess their capability to describe cloud patterns and provide the support needed to fully exploit the solutions they provide.
Keywords: cloud computing; cloud patterns; TOSCA; OpenStack HOT; interoperability; portability.
Morphosemantic strategies for the automatic enrichment of Italian lexical databases in the medical domain
by Flora Amato, Antonino Mazzeo, Annibale Elia, Alessandro Maisto, Serena Pelosi
Abstract: Because of the importance of the information conveyed by the clinical documents, and owing to the large quantity of raw texts produced in the healthcare system, the extraction and the management of meaningful data, starting from real text occurrences is a determinant challenge in the NLP research field. In this paper, we approach a corpus of 5000 medical diagnoses with sophisticated linguistic and computational devices, which are able to access the semantic dimension of words and sentences contained in it. Our morphosemantic method is grounded on a list of neoclassical formative elements pertaining to the medical domain which has been used for the automatic creation and population of medical lexical resources. The outcomes of this work are automatically built electronic dictionaries and thesauri and an annotated corpus for the NLP in the medical domain.
Keywords: medical thesauri population; information extraction; natural language processing.
A microcontroller multicore in FPGAs: detailed architecture and case studies of embedded critical applications
by Cesar Penteado, Edward Moreno, Fabio D. Pereira
Abstract: This paper presents the concept and preliminary tests in FPGA of architecture for a flexible multicore microcontroller. It is aimed to intermediate complexity embedded applications. A previous exact characterisation of the microcontroller model and its target applications is a costly-time task, and depends mostly on the engineers' and programmers' experiences. The proposed architecture can aid the development of new applications, since it allows the selection of hardware components useful to the project during the development of the application. We have designed a prototype in FPGA and it is working with seven CPUs. We have used two applications (PID controllers and AES security algorithm), and our results allow the idea and utility of multicore microcontrollers to be validated.
Keywords: multicore; microcontroller; embedded system; FPGA and VHDL; soft processor; critical applications
Performance analysis of Linux containers for high performance computing applications
by David Beserra, Edward Moreno, Patricia Endo, Jymmy Barreto, Stênio Fernandes, Djamel Sadok
Abstract: Although cloud infrastructures can be used as High Performance Computing (HPC) platforms, many issues from virtualisation overhead have kept them almost unrelated. However, with advent of container-based virtualisation, this scenario acquires new perspectives because this technique promises to decrease the virtualisation overhead, achieving a near-native performance. In this work, we analyse the performance of a container-based virtualisation solution - Linux Container (LXC) - against a hypervisor-based virtualisation solution - KVM - under HPC activities. For our experiments, we consider CPU and (network and inter-process) communication performance. Results show that hypervisor type can impact distinctly in performance according to resource used by HPC application.
Keywords: cloud computing; HPC; LXC; container-based virtualisation; performance evaluation.
Enhanced cuckoo search algorithm for virtual machine placement in cloud data centres
by Esha Barlaskar, Jayanta Singh, Biju Issac
Abstract: In order to enhance resource utilisation and power efficiency in cloud data centres it is important to perform Virtual Machine (VM) placement in an optimal manner. VM placement uses the method of mapping virtual machines to physical machines (PM). Cloud computing researchers have recently introduced various metaheuristic algorithms for VM placement considering the optimised energy consumption. However, these algorithms do not meet the optimal energy consumption requirements. This paper proposes an Enhanced Cuckoo Search (ECS) algorithm to address the issues with VM placement focusing on the energy consumption. The performance of the proposed algorithm is evaluated using three different workloads in CloudSim tool. The evaluation process includes comparison of the proposed algorithm against the existing Genetic Algorithm (GA,) Optimised Firefly Search Algorithm (OFS), and Ant Colony (AC) algorithm. The comparision results illustrate that the proposed ECS algorithm consumes less energy than the participant algorithms while maintaining a steady performance for SLA and VM migration. The ECS algorithm consumes around 25% less energy than GA, 27% less than OFS, and 26% less than AC.
Keywords: virtual machine placement; metaheuristic algorithms; enhanced cuckoo search algorithm; cloud computing
SER performance optimisation of AF cooperative communication system based on directional antenna
by Ruilian Tan, Zhe Li, Xi Su
Abstract: Aiming at the cooperative communication system with directional antenna, this paper has studied the SER (Symbol Error Rate) performance under AF (Amplify-and-Forward protocol). The model of AF cooperative communication system using directional antenna is first established to deduce the closed-form expression of SER in this model, as well as the upper limit of the SER. Then, the OPA (Optimum Power Allocation) is also analysed with the purpose of minimising the SER. Combining specific simulation numerical values, the SER performance of established model is thoroughly researched. Simulation results demonstrate that the system SER decreases by adopting a cooperative communication system with directional transmitting and directional receiving. Each nodes directional gain, channel quality and power allocation method all have great influence on the systems overall performance. The OPA is also proved to be superior to EPA (Equal Power Allocation).
Keywords: cooperative communication; directional antenna; amplify and forward; symbol error rate.
MT-DIPS: a new data duplication integrity protection scheme for multi-tenants sharing storage in SaaS
by Lin Li, Yongxin Zhang, Yanhui Ding
Abstract: In SaaS, the data sharing storage mode and tenant isolation requirement present a new challenge to traditional remote data duplication protection schemes. This paper aims at the new requirement of tenant data duplication protection in SaaS, and presents a tuple sampling based tenant duplication protection mechanism MTDIPS (Duplication Integrity Protection Scheme for Multi-tenants). Instead of data block sampling, MT-DIPS accommodates the data isolation requirement of different tenants by sampling tenants' physical data tuples. Through periodical random sampling, MTDIPS reduces the complexity on the service provider side of verification object construction and eliminates resource waste. Analysis and the experimental results show that if the damage rate of tenant data tuples is about 1%, the random sampling data number is about 5% of the total number of tuples. MT-DIPS makes use of homomorphism labels with auxiliary authentication structure to allow trusted third party verification without disclosing tenant data to relieve the verification burden on tenants' client sides.
Keywords: SaaS; multi-tenant; duplication; integrity authentication; cloud computing.
An autonomic monitoring framework for IaaS cloud applications
by Rocco Aversa, Luca Tasquier, Giuseppe Sanges
Abstract: Within the cloud application lifecycle, cloud monitoring takes an important role because the fulfillment of the requirements has to be continuously controlled in order to avoid saturation or underuse of cloud resources and to check the compliance of the signed service level agreements with the real performance of the infrastructure. In fact, to ensure scalability and dependability, the user's applications are often distributed on several computational resources, such as virtual machines, storages and so on, and the customer often is able to retrieve information about the cloud infrastructure only by acquiring monitoring services provided by the same vendor that is offering the cloud resources. In this work, we present a complete framework that covers all the monitoring activities that take place within a cloud application lifecycle, introducing autonomic monitoring facilities that, exploiting the agent technology capabilities, allow the monitoring infrastructure to automatically adapt itself to the execution environment. These facilities provide both robustness and performance control to the framework, reducing the invasiveness of the monitoring and allowing, only if necessary, a deeper analysis of the measured data.
Keywords: cloud autonomic monitoring; mobile agents; IaaS cloud; service level agreement.
A comparison of two fuzzy-based systems considering node security in MANET clusters
by Mijeta Alinci, Takaaki Inaba, Donald Elmazi, Evjola Spaho, Vladi Kolici, Leonard Barolli
Abstract: A Mobile Ad hoc Network (MANET) is a multi-hop wireless network in which the mobile nodes are dynamic in nature. The network has a limited bandwidth and minimum battery power. Owing to this challenging environment the mobile nodes can be grouped into clusters to achieve better stability and scalability. Grouping the mobile nodes is called clustering, in which a leader node is elected to manage the entire network. In this paper, first we introduce various approaches for clustering focus on different performance metrics. Then, we show some clustering schemes. Finally, we present and compare two fuzzy-based systems (called F2SMC1 and F2SMC2) for improving the security of cluster nodes in MANETs. We compare the performances of F2SMC1 and F2SMC2 and show that the F2SMC2 is more complex than F2SMC1, but is more reliable and secure.
Keywords: MANET; clustering; cluster head; algorithms.
An improved image classification based on K-means clustering and BoW model
by YongLang Liu, Zhong Cai, JiTao Zhang
Abstract: Image classification constitutes an important issue in large-scale image data process systems based on clusters. In this context, a significant number of relying BoW model and SVM methods have been proposed for image fusion systems. Some works classified these methods into generative mode and discriminative mode. Very few works deal with a classifier based on the fusion of these modes when building an image classification system. In this paper, we propose a revised algorithm based on weighted visual dictionary of K-means cluster. First, it uses SIFT and Laplace spectrum features to cluster objects respectively to get local characteristics of low dimension images (sub visual dictionary); then it clusters low dimension characteristics to get the super visual dictionaries of two features; finally, we get the final visual dictionary although most of these features have been proposed for a balance role through weighting of the parent visual dictionaries. Experimental results show that the algorithm and this model are efficient in describing image information and can provide image classification performance. It is widely used in unmanned navigation and machine vision and other fields.
Keywords: image classification; visual dictionary; K-means; BoW model.
Users priority focused resource provisioning over cloud computing infrastructure
by Madhumathi Ramasamy, Radhakrishnan Rathinavel
Abstract: Resource provisioning is the process of activating a bundle of an allocated quantity of resources to bear the user requests. The scheduling algorithm plays a vital role in effective use of resources, though resource allocation fails to accomplish the satisfaction of a few users where the user is prioritised and then the resource is allocated. To overcome this failure, Drip irrigation-based Resource Allocation (DRA) scheduling algorithm is proposed to exploit the user priority on the basis of the reputation of the users who frequently request the resources. This algorithm works in three stages, namely resource selection, resource matching and drip-based resource allocation, where the unallocated requests are divided into smaller sub-tasks and the limited available resources are allocated to them. Simulation results demonstrate that the DRA scheduling mechanism is effective in satisfying the users diverse requirements by considering their priority. It also performs better in terms of resource use rate compared to First Come First Serve (FCFS) and UFeed algorithms from the cloud providers perspective, and satisfies more users in the cloud system.
Keywords: cloud computing; resource allocation; scheduling; virtual machines; priority; users.
Extensible markup language keywords search based on security access control
by MeiJuan Wang, Jian Wang, KeJun Guo
Abstract: With increasing rate of storing and sharing information in the cloud by users, data storage brings new challenges to the Extensible Markup Language (XML) database in big data environments. The efficient retrieval of data with protection and privacy issues for accessing mass data in the cloud has become more and more important. Most existing research about XML data query and retrieval focuses on efficiency or establishing the index, and so on. However, these methods or algorithms do not take into account the data and data structure for their own safety issues. Furthermore, traditional access control rules read XML document nodes in a dynamic environment, and relevant dynamic query-based keyword research data security and privacy protection requirements are not many. In order to improve the search efficiency with security condition, this paper discusses how to generate the sub-tree of matching keywords that the user can access by the access control rules for the users role. The corresponding algorithm is proposed to achieve safe and efficient keywords search.
Keywords: XML access control; keywords search; SLCA.
A hybrid heuristic resource allocation model for computational grid for optimal energy usage
by Deo Vidyarthi, Achal Kaushik
Abstract: Computational grid helps in faster execution of compute-intensive jobs. The resource allocation for the job execution in computational grid demands a lot of characteristic parameters to be optimised, but in the process the green aspect is ignored. Reducing the energy consumption in computational grid is a major recent research issue. The conventional systems, which offer energy-efficient scheduling strategies, ignore other quality of service parameters while scheduling the jobs. The proposed work tries to optimise the energy for resource allocation and at the same time makes no compromise on other related characteristic parameters. A hybrid model, which uses genetic algorithm and graph theory concept, has been proposed for this purpose. In this model, an energy-saving mechanism is implemented using a dynamic threshold method followed by genetic algorithm to further consolidate the saving. Eventually, a graph theory concept of minimum spanning tree is applied. The performance of the proposed model has been studied by its simulation. The result reveals the benefits achieved with the proposed model for optimal energy usage with resource allocation in the grid.
Keywords: computational grid; green energy; resource allocation; genetic algorithm; minimum spanning tree; quality of service.
Improved quantisation mechanisms in impulse radio ultra wideband systems based on compressed sensing
by Yunhe Li, Qinyu Zhang Zhang, Shaohua Wu
Abstract: To reduce the impact of quantization noise during low-rate sampling of Impulse Radio Ultra-Wideband (IR-UWB) under compressed sensing framework, and in the meantime considering the equal information carrying of compressed measurements and characteristics of Gaussian distribution, three modified quantisation mechanisms are designed in this study: overload uniform quantisation, non-uniform quantisation, and overload non-uniform quantisation. Besides the influencing elements of overload factor in overload mechanism are mentioned to obtain the optimisation scheme close to the optimal overload by fitting. The simulation results show that all the three modified mechanisms, especially the overload non-uniform quantisation mechanism, have resulted in great improvement in performance when compared with the uniform quantisation, Whats more, the performance of the overload uniform quantisation mechanism featured with low complexity is better than that of the non-uniform quantisation mechanism with high complexity, thus providing a practical quantisation method for the IR-UWB system under the compressed sensing framework
Keywords: compressed sensing; impulse radio ultra-wideband; quantisation mechanism; quantisation noise; overload interval factor.
Performance analysis of two WMN architectures by WMN-GA simulation system considering different distributions and transmission rates
by Keita Matsuo, Miralda Cuka, Takaaki Inaba, Tetsuya Oda, Leonard Barolli, Admir Barolli
Abstract: In this paper, we evaluate the performance of two Wireless Mesh Network (WMN) architectures considering throughput, delay and fairness index metrics. For simulations, we used a Genetic Algorithm (GA) based simulation system (called WMNGA) and ns-3. We compare the performance for the two architectures considering normal and uniform distributions, different transmission rates and OLSR protocol. The simulation results show that the throughput and delay are increased, but the fairness index is decreased as the transmission rate increases. The throughput of the hybrid WMN is higher than that of the I/B WMN, but the delay of the I/B WMN is higher than that of the hybrid WMN for normal distribution. The fairness index of the normal distribution is higher than that of the uniform distribution.
Keywords: genetic algorithms; wireless mesh networks; NS-3; network architecture; OLSR; SGC; NCMC; normal distribution; uniform distribution; transmission rate.
An ontology-based cloud infrastructure service discovery and selection system
by Manoranjan Parhi, Binod Kumar Pattanayak, Manas Ranjan Patra
Abstract: In recent years, owing to the global economic downturn, many organisations have resorted to downsizing their Information Technology (IT) expenses by adopting innovative computing models, such as cloud computing, which allows business houses to reduce their fixed IT costs by promising a greener, scalable, cost-effective alternative to use their IT resources. A growing number of pay-per-use cloud services are now available on the web in the form of Software as a Service (SaaS), Infrastructure as a Service (IaaS) and Platform as a Service (PaaS). With the increase in the number of services, there has also been an increase in demand and adoption of cloud services, making cloud service identification and discovery a challenging task. This is due to varied service descriptions, non-standardised naming conventions, heterogeneity in type and features of cloud services. Thus, selecting an appropriate cloud service according to consumer requirements is a daunting task, especially for applications that use a composition of different cloud services. In this paper, we have designed an ontology-based cloud infrastructure service discovery and selection system that defines functional and non-functional concepts, attributes and relations of infrastructure services. We have shown how the system enables one to discover appropriate services optimally as requested by consumers.
Keywords: cloud computing; cloud service discovery and selection; infrastructure as a service; cloud ontology.
Certificateless multi-signcryption scheme in standard model
by Xuguang Wu
Abstract: A signcryption scheme can realise the security objectives of encryption and signature simultaneously, which has lower computational cost and communication overhead than the sign-then-encrypt approach. To adapt multi-user settings and solve the key escrow problem of ID-based multi-signcryption schemes, this paper defines the formal model of certificatless multi-signcryption scheme and proposes a certificateless multi-signcryption scheme in the standard model. The scheme is proved secure against adaptive chosen ciphertext attacks and adaptive chosen message attacks under decisional bilinear Diffie-Hellman assumption and computational Diffie-Hellman assumption, respectively.
Keywords: signcryption; multi-signcryption; certificatless encryption.
Per-service security SLAs for cloud security management: model and implementation
by Valentina Casola, Alessandra De Benedictis, Jolanda Modic, Massimiliano Rak, Umberto Villano
Abstract: In the cloud computing context, Service Level Agreements (SLAs) tailored to specific Cloud Service Customers (CSCs) seem to be still a utopia, and things are even worse as regards the security terms to be guaranteed. In fact, existing cloud SLAs focus on only a few service terms, and Cloud Service Providers (CSPs) mainly provide uniform guarantees for all offered services and for all customers, regardless of any particular service characteristics or of customer-specific needs. In order to expand their business volume, CSPs are currently starting to explore alternative approaches, based on the adoption of a CSC-based per-service security SLA model. This paper presents a framework that enables the adoption of a per-service SLA model, supporting the automatic implementation of cloud security SLAs tailored to the needs of each customer for specific service instances. In particular, the process and the software architecture for per-service SLA implementation are shown. A case study application, related to the provisioning of a secure web container service, is presented and discussed, to demonstrate the feasibility and effectiveness of the proposed solution.
Keywords: cloud security; per-service SLA; security service level agreement.
A (multi) GPU iterative reconstruction algorithm based on Hessian penalty term for sparse MRI
by Salvatore Cuomo, Pasquale De Michele, Francesco Piccialli
Abstract: A recent trend in the Magnetic Resonance Imaging (MRI) research field is to design and adopt machines that are able to acquire undersampled clinical data, reducing the time for which the patient is lying in the body scanner. Unfortunately, the missing information in these undersampled acquired datasets leads to artifacts in the reconstructed image, therefore computationally expensive image reconstruction techniques are required. In this paper, we present an iterative regularisation strategy with a second order derivative penalty term for the reconstruction of undersampled image datasets. Moreover, we compare this approach with other constrained minimisation methods, resulting in improved accuracy. Finally, an implementation on a massively parallel architecture environment, a multi Graphics Processing Unit (GPU) system, of the proposed iterative algorithm is presented. The resulting performance gives clinically feasible reconstruction run times, speed-up and improvements in terms of reconstruction accuracy of the undersampled MRI images.
Keywords: compressed sensing; MRI iterative reconstruction; numerical regularisation; graphics processing unit; parallel and scientific computing.
Labelled evolutionary Petri nets/genetic algorithm based approach for workflow scheduling in cloud computing
by Manel Femmam, Okba Kazar, Laid Kahloul, Mohamed El-kabir Fareh
Abstract: Nowadays, more evolutionary algorithms for workflow scheduling in cloud computing are proposed. Most of those algorithms focused on the effectiveness, discarding the issue of flexibility. Research on Petri nets addresses the issue of flexibility; many extensions have been proposed to facilitate the modelling of complex systems. Typical extensions are the addition of "colour", "time" and "hierarchy". By mapping scheduling problems into Petri nets, we are able to use standard Petri net theory. In this case, the scheduling problem can be reduced to finding an optimal sequence of transitions leading from an initial marking to a final one. To find the optimal scheduling, we propose a new approach based on a recent proposed formalism Evolutionary Petri Net (EPN), which is an extension of Petri net, enriched with two genetic operators, crossover and mutation. The objectives of our research are to minimise the workflow application completion time (makespan) as well as the amount cost incurred by using cloud resources. Some numerical experiments are carried out to demonstrate the usefulness of our algorithm.
Keywords: workflow scheduling; cloud computing; petri nets; genetic algorithm.
An improved SMURF scheme for cleaning RFID data
by He Xu
Abstract: With the increasing usage of internet of things devices, our daily life is facing Big Data. RFID technology enables the reading over a long distance, provides high storage capacity and is widely used in the internet of things environmental supply chain management for object tracking and tracing. With the expansion of the RFID technology application areas, the demand for reliability of business data is increasingly important. In order to fulfil the needs of upper applications, data cleaning is essential and directly affects the correctness and completeness of the business data, so it needs to filter and handle RFID data. The traditional statistical smoothing for unreliable RFID data (SMURF) algorithm dynamically adjusts the size of a window according to tags average reading rate of sliding window during the process of data cleaning. To some extent, SMURF overcomes the disadvantages of fixed sliding window size; however, the SMURF algorithm is only aimed at constant speed data flow in ideal situations. In this paper, we overcome the shortcomings of the SMURF algorithm, and propose a SMURF scheme improved in two aspects. The first one is based on dynamic tags, and the second one is the RFID data cleaning framework, which considers the influence of data redundancy. The experiments verify that the improved scheme is reasonable in dynamic settings of sliding window, and the accuracy of cleaning effect is improved as well.
Keywords: RFID; data cleaning; internet of things; sliding window.
Adaptive co-operation in parallel memetic algorithms for rich vehicle routing problems
by Jakub Nalepa, Miroslaw Blocho
Abstract: Designing and implementing co-operation schemes for parallel algorithms has become a very important task recently. The scheme, which defines the co-operation topology, frequency and strategies for handling transferred solutions, has a tremendous influence on the algorithm search capabilities, and can help to balance the exploration and exploitation of the vast solution space. In this paper, we present both static and dynamic schemes: the former are selected before the algorithm execution, whereas the latter are dynamically updated on the fly to better respond to the optimisation progress. To understand the impact of such co-operation approaches, we applied them in the parallel memetic algorithms for solving rich routing problems, and performed an extensive experimental study using well-known benchmark sets. This experimental analysis is backed with the appropriate statistical tests to verify the importance of the retrieved results.
Keywords: co-operation; parallel algorithm; memetic algorithm; rich routing problem; VRPTW; PDPTW.
Cloud computing based on agent technology, superrecursive algorithms and DNA
by Rao Mikkilineni, Mark Burgin
Abstract: Agents and agent systems are becoming more and more important in the development of a variety of fields, such as ubiquitous computing, ambient intelligence, autonomous computing, data analytics, machine learning, intelligent systems and intelligent robotics. In this paper, we examine interactions of theoretical computer science with computer and network technologies, analysing how agent technology is presented in mathematical models of computation. We demonstrate how these models are used in the novel distributed intelligent managed element (DIME) network architecture (DNA), which extends the conventional computational model of information processing networks, allowing improvement of the efficiency and resilience of computational processes. Two implementations of DNA described in the paper illustrate how the application of agent technology radically improves the current cloud computing state of the art. The first example demonstrates the live migration of a database from a laptop to a cloud without losing transactions and without using containers or moving virtual machine images. The second example demonstrates the implementation of cloud agnostic computing over a network of public and private clouds, where live computing process workflows are moved from one cloud to another without losing transactions. Both these implementations demonstrate the power of scientific thought for dramatically extending the current state of the art of cloud computing practice.
Keywords: cloud computing; agent technology; inductive Turing machine; grid computing; DIME network architecture; intelligent systems; super-recursive algorithms.
Attendance management system using selfies and signatures
by Jun Iio
Abstract: There have been many proposals to optimise student attendance management in higher education. However, each method has pros and cons and we have not yet found a perfect solution. In this study, a novel framework for attendance management is proposed that consists of a mobile device and a web application. During lectures, students participating in the lecture can register their attendance on the mobile device with their selfie or their signature. After the lecture is finished, the registration data are sent to the database and they are added to the 'RollSheet'. This paper reports an overview of this system and the results of an evaluation after a trial period, which was conducted in the second semester of the 2015 fiscal year.
Keywords: attendance management system; selfie photograph; hand-writing signature; mobile device; web application.
Trust modelling for opportunistic cloud services
by Eric Kuada
Abstract: This paper presents a model for the concept of trust and a trust management system for opportunistic cloud services platforms. Results from applying the systematic review methodology to review trust-related studies in cloud computing revealed that the concept of trust is used loosely without any formal specification in cloud computing discussions and trust engineering in general. Formal definition and a model of the concept of trust is, however, essential in the design of trust management systems. The paper therefore presents a model for the formal specification of the concept of trust. A trust management system for opportunistic cloud services is also presented. The applicability of the trust model and the trust management system is demonstrated for cloud computing by applying it to software as a service and infrastructure as a service usage scenarios in the context of opportunistic cloud services environments.
Keywords: opportunistic cloud services; trust engineering; trust in cloud computing; trust modeling; trust management system; pseudo service level agreements.
Efficient cache replacement policy for minimising error rate in L2-STT-MRAM caches
by Rashidah F. Olanrewaju, Burhan Ul Islam Khan, A. Raouf Khan, Mashkuri Yaacob, Md Moktarul Alam
Abstract: In recent times, various challenges have been encountered in the design and development of Static-RAM (SRAM) caches, which consequently has led to a design where memory cell technologies are converted into on-chip embedded caches. The current research statistics for cache designing reveals that Spin Torque Transfer Magnetic RAMs, preferably termed as STT-MRAMs, have become one of the most promising technologies in the field of memory chip design, gaining a lot of attention from researchers owing to their dynamic direct map and data access policies for reducing the average cost, i.e. both time and energy optimisation. Though STT-MRAMs possess high density, less power rating and non-volatility, increasing rates of WRITE failures and READ disturbances highly affect the reliability of STT-MRAM caches. Besides workload behaviours, process variations directly affect these failure/disturbance rates. Furthermore, it can be seen that cache replacement algorithms play a significant part in minimising the Error Rate (ER) induced by WRITE operations. In this paper, the vulnerability of STT-MRAM caches has been investigated to examine the effect of workloads as well as process variations for characterising the reliability of STT-MRAM caches. The current study is intended to analyse and evaluate an existing efficient cache replacement policy, namely Least Error Rate (LER), which uses Hamming Distance (HD) computations to reduce the Write Error Rate (WER) of L2-STT-MRAM caches with acceptable overheads. The performance analysis of the algorithm ensures its effectiveness in reducing the WER and cost overheads compared with the conventional LRU technique implemented on SRAM cells.
Keywords: cache replacement algorithm; field assisted STT-MRAM; error rate; L2 caches.
An infrastructure model for smart cities based on big data
by Eliza Helena Areias Gomes, Mario Antonio Ribeiro Dantas, Douglas D. J. De Macedo, Carlos Roberto De Rolt, Julio Dias, Luca Foschini
Abstract: The spread of projects focused on smart cities has grown in recent years. With this, the massive amount of data generated in these initiatives creates a degree of complexity in how to manage all this information. In attention to solve this problem, several approaches have been developed in recent years. In this paper, we propose an infrastructure model for big data for a smart city project. The goal of this model is to present the stages for the processing of data in the steps of extraction, storage, processing and visualisation, as well as the types of tool needed for each phase. To
implement our proposed model, we used the ParticipACT Brazil, a project based in smart cities. This project uses different databases to compose its big data and uses this data to solve urban problems. We observe that our model provides a structured vision of the software to be used in big data server of ParticipACT Brazil.
Keywords: big data; smart city; big data tools.
A negotiation-based dynamic pricing heuristic in cloud computing
by Gaurav Baranwal, Dinesh Kumar, Zahid Raza, Deo Prakash Vidyarthi
Abstract: Over the years, cloud computing has emerged as a good business platform for IT related services. In the cloud, prices of computing resource act as a lever to control the use of the resources. That is the reason, when the number of cloud customers started increasing, why cloud service providers started offering resources with various pricing schemes to attract customers. This work proposes a negotiation-based heuristic for dynamic pricing that considers the behaviour of both the service provider and the customer and tries to optimally satisfy both for pricing. Both customer and provider are reluctant to reveal information about their utility to each other. The designed utility function for the provider considers payment offered by the customer and the opinion of the provider about the customer. Similarly, the utility function for the customer considers the price offered by the provider and the opinion of the customer about the provider. This will encourage both to offer their true value. Performance study indicates that the proposed method performs well and is a potential candidate for its implementation in a real cloud.
Keywords: cloud pricing; negotiations; cloud market; cloud agent; trustability.
Playing in traffic: an investigation of low-cost, non-invasive traffic sensors for street lighting luminaire deployment
by Karl Mohring, Trina Myers, Ian Atkinson
Abstract: Real-time traffic monitoring is essential to the development of smart cities as well as its potential for energy savings. However, real-time traffic monitoring is a task that requires sophisticated and expensive hardware. Owing to the prohibitive cost of specialised sensors, accurate traffic counts are typically limited to intersections where traffic information is used for signalling purposes. The sparse arrangement of traffic detection points does not provide adequate information for intelligent lighting applications, such as adaptive dimming. This paper investigates the low-cost and off-the-shelf sensors to be installed inside street lighting luminaires for traffic sensing. A luminaire-mounted sensor test-bed installed on a moderately busy road trialled three non-invasive presence-detection sensors: Passive Infrared (PIR), Sonar (UVD) and lidar. The proof-of-concept study revealed that a HC-SR501 PIR motion detector could count traffic with 73% accuracy at a low cost and may be suitable for intelligent lighting applications if accuracy can be further improved.
Keywords: commodity; internet of things; vehicle detection; sensors; smart cities; wireless sensor networks.
Real-time web-cast system by multihop WebRTC communications
by Daiki Ito, Michitoshi Niibori, Masaru Kamada
Abstract: A software system is developed for casting the screen images and voices from a host PC to the client web browsers on many other PCs in real time. This system is intended to be used in the classrooms. Students have only to bring their own PCs and connect to the teachers host PC by a web browser via a wireless network to see and listen to the teaching materials presented on the host PC. Then the client web-browsers are organised in the shape of a binary tree along which the video and audio data are relayed in the multihop fashion by the Web Real-time Communication (WebRTC) protocol. This structure of binary multihop relay is adopted in order not to burden the host PC with communications load. A test has shown that voice and the motion pictures in a rather small size of 320 x 240 pixels on a teachers PC have been presented at the rate of five frames per second without any noticeable delays on the web browsers running on 38 client devices for students under a local WiFi network. To host more client devices, we have to lower the frame rate as slow as the slide show of still pictures.
Keywords: real-time web-cast system; bring your own device; WebSocket; web real-time communication.
Dynamic migration of virtual machines to reduce energy consumption in a cluster
by Dilawaer Duolikun, Tomoya Enokido, Makoto Takizawa
Abstract: Virtual machines are widely used to support applications with virtual service in server clusters. Here, a virtual machine can migrate from a host server to a guest server. In this paper, we consider a cluster where virtual machines are dynamically created and dropped depending on the number of processes. We propose a dynamic virtual machine migration (DVMM) algorithm to reduce the total electric energy consumption of servers. If an applications issues a process to a cluster, the most energy-efficient host server is first selected and then the process is performed on a virtual machine of the server. Then, a virtual machine migrates from a host server to a guest server so that total electric energy consumption of the servers can be reduced. In the evaluation, we show the total electric energy consumption and active time of servers and the average execution time of processes can be reduced in the DVMM algorithm.
Keywords: energy-efficient computation; virtual machine; power consumption model; energy-aware dynamic migration of virtual machines.
Energy-efficient placement of virtual machines in cloud datacentres, based on fuzzy decision making
by Leili Salimian, Faramarz Safi-Esfahani
Abstract: Placement of virtual machines (VMs) on physical nodes as a sub-problem of dynamic VM consolidation has been driven mainly by energy efficiency and performance objectives. However, owing to varying workloads in VMs, placement of the VMs can cause a violation in the Service Level Agreement (SLA). In this paper, the VM placement is regarded as a bin packing problem, and a fuzzy energy-aware algorithm is proposed to estimate the host resource usage. The estimated resource usage is used to find the most energy-efficient host to reallocate the VMs. The fuzzy algorithm generates rules and membership functions dynamically to adapt to workload changes. The main objective of the proposed algorithm is to optimise the energy-performance trade-off. The effectiveness of the proposed algorithm is evaluated through simulations on the random and real-world PlanetLab workloads. Simulation results demonstrate that the proposed algorithm reduces the energy consumption, while it provides a high level of adherence to the SLAs.
Keywords: dynamic VM consolidation; CPU usage; VM placement; fuzzy decision making.
An identity-based cryptographic scheme for cloud storage applications
by Manel Medhioub, Mohamed Hamdi
Abstract: The use of remote storage systems is gaining an expanding interest, namely the cloud storage based services. In fact, one of the factors that led to the popularity of cloud computing is the availability of storage resources provided at a reduced cost. However, when outsourcing the data to a third party, security issues become critical concerns, especially confidentiality, integrity, authentication, anonymity and resiliency. Based on this challenge, this work provides a new approach to ensure authentication in cloud storage applications. ID-based cryptosystems (IBC) have many advantages over certificate-based systems, such as simplification of key management. This paper proposes an original ID-based authentication approach in which the cloud tenant is assigned the IBCPrivate Key Generator (PKG) function. Consequently, it can issue public elements for its users, and can keep confidential the resulting IBC secrets. Moreover, in our scheme, the public key infrastructure is still in usage to establish trust relationships between the PKGs.
Keywords: cloud storage; authentication; identity-based cryptography; security; Dropbox.
COBRA-HPA: a block generating tool to perform hybrid program analysis
by Thomas Huybrechts, Yorick De Bock, Haoxuan Li, Peter Hellinckx
Abstract: The Worst-Case Execution Time (WCET) of a task is an important value in real-time systems. This metric is used by the scheduler in order to schedule all tasks before their deadlines. However, the code and hardware architecture have a significant impact on the execution time and thus the WCET. Therefore, different analysis methodologies exist to determine the WCET, each with their own advantages and/or disadvantages. In this paper, a hybrid approach is proposed that combines the strengths of two common analysis techniques. This hybrid methodology tackles the problem that can be described as 'the gap between a machine and a human in solving problems'. The two-layer hybrid model splits the code of tasks into so-called basic blocks. The WCET can be determined by performing execution time measurements on each block and statically combining those results. The COBRA-HPA framework presented in this paper is developed to facilitate the creation of hybrid block models and automate the measurements/analysis process. Additionally, an elaborated discussion on the implementation and performance of the framework is given. In conclusion, the results of the COBRA-HPA framework show a significant reduction in analysis effort while keeping sound WCET predictions for the hybrid method compared with the static and measurement-based approach.
Keywords: worst-case execution time; WCET; hybrid analysis methodology; COde Behaviour fRAmework; COBRA; basic block generator.
The big data mining forecasting model based on combination of improved manifold learning and deep learning
by Xiurong Chen, Yixiang Tian
Abstract: For the most important dilemma in big data processing that extensive redundant information and useful information mix with each other, which makes these big data difficult to be effectively used to establish prediction models, in our work we combine the manifold learning dimension reduction algorithm LLE with deep learning feature extraction algorithm CDBN as the input of RBF, constructing a mixed-feature RBF forecast model. As for depending too much on the local domain, which is not easy to determine in the LLE algorithm, we used the idea of mapping by kernel function of KECA to transfer original global nonlinear problem into global linear one under the high-dimensional kernel feature space to solve, removing the redundant information more accurately and reducing data complexity. As for the difficulty in confirming network structure and the lack of supervision in learning process of CDBN, we used the kernel entropy information computed in KECA to determine the number of network layers and supervise the learning process, which makes it more effective to extract deep features to explore the essential characteristics of big data information. In the empirical part we chose the foreign exchange rate time series to conduct research, the results show that the improved KELE can reduce dimensionality of sample data effectively which makes we obtain the more optimised and reasonable representation of original data, providing an assurance for further learning and understanding of big data. And the improved KECDBN can extract distributed features of data more effectively. Then improve the prediction accuracy of the mixed-feature RBF forecast model based on KELE and KECRBM.
Keywords: locally liner embedding; continuous deep belief network; kernel entropy component analysis; kernel entropy liner embedding; kernel entropy continuous deep belief network.
Cost-aware hybrid cloud scheduling of parameter sweep calculations using predictive algorithms
by Stig Bosmans, Glenn Maricaux, Filip Van Der Schueren, Peter Hellinckx
Abstract: This paper investigates various techniques for scheduling parameter sweep
calculations cost efficiently in a hybrid cloud environment. The combination of both
a private and public cloud environment integrates the advantages of being cost effective and having virtually unlimited scaling capabilities at the same time. To make an accurate estimate for the required resources, multiple prediction techniques are discussed. The estimation can be used to create an efficient scheduler which respects both deadline and cost. These findings have been implemented and tested in a Java-based cloud framework that operates on Amazon EC2 and OpenNebula. Also, we present a theoretical model to further optimise the cost by leveraging the Amazon Spot Market.
Keywords: parameter sweep; cloud computing; Amazon AWS EC2; predictive algorithms; OpenNebula; machine learning; Amazon spot market.
Impact of software architecture on execution time: a power window TACLeBench case study
by Haoxuan Li, Paul De Meulenaere, Siegfried Mercelis, Peter Hellinckx
Abstract: Timing analysis is used to extract the timing properties of a system. Various timing analysis techniques and tools have been developed over the past decades. However, changes in hardware platform and software architecture introduced new challenges in timing analysis techniques. In our research, we aim to develop a hybrid approach to provide safe and precise timing analysis results. In this approach, we will divide the original code into smaller code blocks, then construct a timing model based on the information acquired by measuring the execution time of every individual block. This process can introduce changes in the software architecture. In this paper we use a multi-component benchmark to investigate the impact of software architecture on the timing behaviour of a system.
Keywords: WCET; timing analysis; hybrid timing analysis; power window; embedded systems; TACLEBench; COBRA block generator.
Accountability management for multi-tenant cloud services
by Fatma Masmoudi, Mohamed Sellami, Monia Loulou, Ahmed Hadj Kacem
Abstract: The widespread adoption of multi-tenancy in the Software as a Service delivery model triggers several data protection issues that could decrease the tenants' trustworthiness. In this context, accountability can be used to strengthen the trust of tenants in the cloud through providing the reassurance of the processing of personal data hosted in the cloud according to their requirements. In this paper, we propose an approach for the accountability management of multi-tenant cloud services allowing: compliance checking of services's behaviours with defined accountability requirements based on monitoring rules, accountability-violation detection otherwise, and post-violation analysis based on evidences. A tool-suite is developed and integrated into a middleware to implement our proposal. Finally, experiments we have carried out show the efficiency of our approach relying on some criteria.
Keywords: cloud computing; accountability; multi-tenancy; monitoring; accountability violation.
A big data approach for multi-experiment data management
by Silvio Pardi, Guido Russo
Abstract: Data sharing among similar experiments is limited by the usage of ad hoc directory structures, data and metadata naming as well as by the variety of data access protocols used in different computing model. The Open Data and Big Data paradigms provide the context to overcome the current heterogeneity problems. In this work, we present a study for a Global Storage Ecosystem designed to manage large and distributed datasets, in the context of physics experiments. The proposed environment is entirely based on the open protocols HTTP/WebDav, together with modern data searching technologies, according to the Big Data paradigm. More specifically, the main goal is to aggregate multiple storage areas exported with open protocols and to simplify the operations of data retrieval, thanks to a set of engine-search-like tools, based on Elasticsearch and Apache Lucene library. This platform offers to physicists an effective instrument to simplify the multi-experiment data analysis, by enabling data searching, without knowing a priori the directory format or the data itself. As a proof of concept, we realised a prototype over the ReCaS Supercomputing infrastructure, by aggregating and indexing the files stored in a set of already existing storage systems.
Keywords: big data; data federation.
Special Issue on: Recent Developments in Parallel, Distributed and Grid Computing for Big Data
GPU accelerated video super-resolution using transformed spatio-temporal exemplars
by Chaitanya Pavan Tanay Kondapalli, Srikanth Khanna, Chandrasekaran Venkatachalam, Pallav Kumar Baruah, Kartheek Diwakar Pingali, Sai Hareesh Anamandra
Abstract: Super-resolution (SR) is the method of obtaining high resolution (HR) image
or image sequence from one or more low-resolution (LR) images of a scene. Super-resolution has been an active area of research in recent years owing to its applications to defence, satellite imaging, video surveillance and medical diagnostics. In a broad sense, SR techniques can be classified into external database driven and internal database driven approaches. The training phase in the first approach is computationally intensive as it learns the LR-HR patch relationships from huge datasets, and the test procedure is relatively fast. In the second approach, the super-resolved image is directly constructed from the available LR image, eliminating the need for any learning phase but the testing phase is computationally intensive. Recently, Huang et al. (2015) proposed a transformed self-exemplar internal database technique which takes advantage of the fractal nature in an image by expanding patch search space using geometric variations. This method fails if there is no patch redundancy within and across image scales and also if there is a failure in detecting vanishing points (VP), which are used to determine perspective transformation between LR image and its subsampled form. In this paper, we expand the patch search space by taking advantage of the temporal dimension of image frames in the scene video and also use an efficient VP detection technique by Lezama et al. (2014). We are thereby able to successfully super-resolve even the failure cases of Huang et al. (2015) and achieve an overall improvement in PSNR. We also focused on reducing the computation time by exploiting the embarrassingly parallel nature of the algorithm. We achieved a speedup of 6 on multi-core, up to 11 on GPU, and around 16 on hybrid platform of multi-core and GPU by parallelising the proposed algorithm. Using our hybrid implementation, we achieved 32x super-resolution factor in limited time. We also demonstrate superior results for the proposed method compared with current state-of-the-art SR methods.
Keywords: super-resolution; self-exemplar; perspective geometry; temporal dimension; vanishing point; GPU; multicore.
Energy-efficient fuzzy-based approach for dynamic virtual machine consolidation
by Anita Choudhary, Mahesh Chandra Govil, Girdhari Singh, Lalit K. Awasthi, Emmanuel S. Pilli
Abstract: In cloud environment the overload leads to performance degradation and Service Level Agreement (SLA) violation while underload results in inefficient use of resources and needless energy consumption. Dynamic Virtual Machine (VM) consolidation is considered as an effective solution to deal with both overload and underload problems. However, dynamic VM consolidation is not a trivial solution as it can also lead to violation of negotiated SLA owing to runtime overheads in VM migration. Further, dynamic VM consolidation approaches need to answer many questions, such as (i) when to migrate a VM? (ii) which VM is to be migrated? and (iii) where to migrate the selected VM? In this work, efforts are made to develop a comprehensive approach to achieve better solutions to such problems. In the proposed approach, future forecasting methods for host overload detection are explored; a fuzzy logic based VM selection approach that enhances the performance of VM selection strategy is developed; and a VM placement algorithm based on destination CPU use is also developed. The performance evaluation of the proposed approaches is carried out on CloudSim toolkit using PlanetLab data set. The simulation results have exhibited significant improvement in the number of VM migrations, energy consumption, and SLA violations.
Keywords: cloud computing; virtual machines; dynamic virtual machine consolidation; exponential smoothing; fuzzy logic.
A distributed framework for cyber-physical cloud systems in collaborative engineering
by Stanislao Grazioso, Mateusz Gospodarczyk, Mario Selvaggio, Giuseppe Di Gironimo
Abstract: Distributed cyber-physical systems play a significant role in enhancing group decision-making processes, as in collaborative engineering. In this work, we develop a distributed framework to allow the use of collaborative approaches in group decision-making problems. We use the fuzzy analytic hierarchy process, a multiple criteria decision-making method, as the algorithm for the selection process. The architecture of the framework makes use of open-source utilities. The information components of the distributed framework act in response to the feedback provided by humans. Cloud infrastructures are used for data storage and remote computation. The motivation behind this work is to make possible the implementation of group decision-making in real scenarios. Two illustrative examples show the feasibility of the approach in different application fields. The main outcome is the achievement of a time reduction for the selection and evaluation process.
Keywords: distributed systems; cyber-physical systems; web services; group decision making; fuzzy AHP; product design and development.
Special Issue on: Resource Provisioning in Cloud Computing
A power saver scheduling algorithm using DVFS and DNS techniques in cloud computing datacentres
by Saleh Atiewi, Salman Yussof, Mohd Ezanee, Mutasem Zalloum
Abstract: Cloud computing is a fascinating and profitable area in modern distributed computing. Aside from providing millions of users with the means to use offered services through their own computers, terminals, and mobile devices, cloud computing presents an environment with low cost, simple user interface, and low power consumption by employing server virtualisation in its offered services (e.g., infrastructure as a service). The pool of virtual machines found in a cloud computing datacentre (DC) must run through an efficient task scheduling algorithm to achieve resource usage and good quality of service, thus ensuring the positive effect of low energy consumption in the cloud computing environment. In this paper, we present an energy-efficient scheduling algorithm for a cloud computing DC using the dynamic voltage frequency scaling technique. The proposed scheduling algorithm can efficiently reduce the energy consumption for executing jobs by increasing resource usage. GreenCloud simulator is used to simulate our algorithm. Experimental results show that, compared with other algorithms, our algorithm can increase server usage, reduce energy consumption, and reduce execution time.
Keywords: DVFS; DNS; virtual machine; datacentre; cloud computing; power consumption.
Towards providing middleware-level proactive resource reorganisation for elastic HPC applications in the cloud
by Rodrigo Righi, Cristiano Costa, Vinicius Facco, Igor Fontana, Mauricio Pillon, Marco Zanatta
Abstract: Elasticity is one of the most important features of cloud computing, referring to the ability to add or remove resources according to the needs of the application or service. Particularly for High Performance Computing (HPC), elasticity can provide a better use of resources and also a reduction in the execution time of applications. Today, we observe the emergence of proactive initiatives to handle the elasticity and HPC duet, but they present at least one problem related to the need of a previous user experience, large processing time, completion of parameters or design for a specific infrastructure and workload setting. Concerning the aforesaid context, this article presents ProElastic, a lightweight model that uses proactive elasticity to drive resource reorganisation decisions for HPC applications. Using ARIMA-based time series and analysing the mean time to launch virtual machines, ProElastic anticipates under- and over-loaded situations, triggering elasticity actions beforehand to address them. Our idea is to explore both performance and adaptivity at middleware level in an effortless way at user perspective, who does not need to either complete elasticity parameters or rewrite the HPC application. Based on ProElastic, we developed a prototype that was evaluated with a master-slave iterative application and compared against reactive-based elasticity and non-elastic approaches. The results showed performance gains and a competitive cost (application time multiplied by consumed resources) in favour of ProElastic when confronted with these two last approaches.
Keywords: cloud elasticity; proactive optimisation; performance; resource management; adaptivity.