Advances on cloud services and cloud computing

Abstract

EDITORIALAdvances on cloud services and cloud computing As cloud computing becomes prevalent, many organizations outsource the expensive computing and storage into the cloud servers. This brings many benefits including relief of the burden for storage management, universal data access with independent geographical locations, and avoidance of capital expenditure on hardware, software, and personnel maintenances. Despite the tremendous benefits, the outsource storage inevitably suffers from the security, privacy, trustworthiness, and the access of outsourced data. To address these issues, there has been increased interest to propose and design new algorithms and methodologies for cloud computing and cloud services.This special issue brings together high quality contributions in these research areas. The papers of the special issue have also identified and discussed the challenges in the field that will lead to new solutions and applications for cloud computing and cloud services.The special issue consists of 12 papers organized as follows. In the first paper [1], the authors propose the guidelines of a holistic security model in online collaborative learning through an effective trustworthiness. They consider a parallel processing approach, which can considerably decrease the time of data processing, thus allowing for building relevant trustworthiness models to support learning activities even in real time. The authors used a distributed infrastructure, such as Hadoop and cluster computing, to implement and evaluate the proposed approach for massive processing of log data. The experimental results showed the feasibility of coping with the problem of structuring and processing ill-formatted, heterogeneous, large log files to extract information on trustworthiness indicators and levels from learning groups and ultimately fill a global framework devoted to improve information security in e-learning in real time.In the second paper [2], the authors present a ciphertext-policy attribute-based proxy reencryption scheme for data sharing in public clouds. They proposed a new single-hop unidirectional Ciphertext-Policy Attribute-Based Proxy Re-Encryptio (CP-ABPRE) scheme, which supports attribute-based re-encryption with any monotonic access structure, to tackle the open problem left by the existing CP-ABPRE schemes. The authors showed that the proposed scheme can be proved secure in the random oracle model under the decisional q-parallel bilinear Diffie-Hellman exponentiation assumption.In the third paper [3], the authors propose Ontology as a Service (OaaS), where the ontology tailoring process serves as a service in the cloud. To illustrate OaaS, in this paper, they propose sub-ontology extraction and extension, whereby multiple sub-ontologies are extracted from various source ontologies and then these extracted sub-ontologies are extended to form a complete ontology to be used by the user. The authors use a maximum extraction method to facilitate this. Unified Medical Language System (UMLS) metathesaurus ontology is us...

Self‐adaptation and mutual adaptation for distributed scheduling in benevolent clouds

“…The real strength of the cloud lies in the ability to distribute the workload over multiple nodes using multiple technologies to create a high‐performance, highly scalable, and available platform at a highly affordable price . Some of the applications of cloud computing are online gaming, social networking, and data center . For economic and efficiency reasons, hybrid clouds further integrate private clouds and public clouds into 1 unified environment …”

A scientific workflow management system architecture and its scheduling based on cloud service platform for manufacturing big data analytics

“…And the perfect strategy can make tasks allocated to the suitable resources and check the best execution sequence of parallel tasks. Generally we just research and consider the deterministic abstract workflows, for which the dependencies of tasks, the sizes of each task, the sizes of the output files of each task, the computing capabilities and communication rates of resources are known in advance, but we do not care about the physical locations of resources [2]. What we input to the algorithm is only the abstract mathematic model.…”

“…In the simulation, we utilize the Million Instructions (MI) to measure the lengths of tasks and the GigaBytes (GB) to measure the sizes of output files. The length of tasks is set by the range [10,000, 6,000,000], and the sizes of output files are within the range [2,34]. And the max outdegree of a task is set for 4.…”

“…Under the "Internet of Things" (IoT) ideology, we adopt smart sensors technologies such as RFID technologies and wireless sensor network technologies to collect useful data ubiquitously [1]. Cloud computing which aims at providing the computer storage and computation as services to help the enterprises solve the problem of limited local resources during the process of big data analytics [2]. But many enterprises are still suffering the deficiency of the capabilities of efficient storage, management and utilization of big data, and the techniques of advanced computing and experiment management.…”

“…In modern scientific environment, scientific workflow becomes data-intensive, which can enable the parallel execution to process large-scale data onto distributed resources [3]. Besides, cloud computing which easily provides the distribute hardware and software resources as visual services for scientific researches and engineering applications can also be combined with the scientific workflows [2]. Within cloud environments, we can encapsulate resources and complex collecting, filtering, computing, visualizations, and other processing steps into different services to be orchestrated as a model of scientific workflows to be managed to process the manufacturing big data [6].…”