International Journal of Web and Grid Services (19 papers in press)
An Overview of Compute First Networking
by Liang Tian, Mingzhe Yang, Shangguang Wang
Abstract: Edge computing has become an important innovative business model in the 5G era, especially its low latency feature, which is considered to be not available in traditional solutions. However, the collaboration of computing power between edge computing and cloud computing in the central position has become a new technical problem. On the basis of studying the compute distribution and scheduling requirements of edge computing, this paper introduces a compute network scheme based on the deep fusion of cloud, edge and network, compute first networking. Firstly, we introduce the basic concept of compute first networking. Then, we mainly discuss the framework and key technology of compute first networking. After that, we present some applications with respect to compute first networking. Finally, we discuss the challenges and opportunities in the area of compute first networking.
Keywords: Compute first networking; mobile edge computing; cloud computing; 5G.
Virtual Vehicle based on Incremental Learning for Navigation Service
by Tao Lei, Lei Yang, Zhizhong Shi, Shangguang Wang
Abstract: In internet of vehicles (IoV),drivers always use navigation systems to plan out routes and optimally navigate real-time road congestion.However, the navigation problem cannot be solved nicely by the present approach due to the emphasis put on one point of view on this problem only. The intelligent transportation systems do not consider the drivers preferences adequately, and driverless cars do not consider real-time traffic conditions. To solve this dispersion, in this paper, we first configure an image of driver and vehicle, named virtual vehicle, to replace driver making some decisions in IoV. Then, we propose an incremental learning approach for virtual vehicles based on negative correlation learning algorithm, called divided negative correlation learning algorithm, to obtain drivers preference knowledge. In the proposed algorithm, the trained ensemble is divided into three parts, where the first part is trained on a new data set, and the second part is used to retrain the old data set. And in the end of the proposed algorithm, we combine the three parts to form a new ensemble. Finally, the experimental results show that virtual vehicles can obtain drivers preference knowledge effectively.
Keywords: internet of vehicles; navigation systems; virtual vehicle.
Automatic Incremental Recomposition Algorithm for QoS-aware Internet-of-Things Service Composition
by Hyunyoung Kil, Wonhong Nam
Abstract: In the near future, the Internet-of-Things (IoT) technology will improve dramatically our daily life as a new pervasive computing paradigm. For the IoT computing, various devices and wireless networks are the hardware infrastructure, and Service-Oriented Architecture (SOA) is a valuable software system that allows heterogeneous devices to interoperate each other. Even though IoT researchers have tackled a number of challenges for service composition, the orchestration techniques on IoT are rarely studied yet. Given a set of IoT services and a goal, the QoS-aware IoT service composition problem constructs a composite IoT service with the optimal accumulated QoS value, which satisfies the given goal specification. However, in the IoT environment, frequent changes happen inherently---for instance, temporary machine down, heavy system workload, and network failure. If the solution we have constructed is not valid anymore due to the changes, we should solve a new problem again. In this paper, we propose a novel incremental recomposition algorithm, which does not solve the new composition problem from scratch but explores only the changed space. In the experiment, our incremental recomposition algorithm can deal with the composition problem much faster than the original algorithm to solve from scratch.
Keywords: Incremental algorithm; Recomposition; Internet-of-Things (IoT); Service composition; QoS optimization.
A Generative Adversarial Network Enabled Learning Scheme for Power Grid Vulnerability Analysis
by Ying Liu, Tao Ye, Zhixiang Zeng, Yingbin Zhang, Guoshi Wang, Ning Chen, Cunli Mao, Xiaohui Yuan
Abstract: Real measurements of power grids are usually limited for research and modeling of extreme events such as the impact of typhoons due to confidentiality concerns. To overcome the dearth of valuable, trustworthy data, this paper proposes an active learning method based on the generative adversarial network. To obtain informative examples, the falsely classified examples together with examples that are correctly classified with low confidence are used to train a GAN for producing synthetic examples to reinforce the learning. The new power grid examples are selected according to the likelihood of the true data distribution. An evaluation was conducted with data acquired by the China Southern Power Grid in Hainan. Most significantly, the performance of detecting the occurrence of a power grid fault under the impact of typhoons is greatly improved. It was demonstrated that the proposed method improved the performance of predicting power grid fault in extreme events by 8.9%. Using the modulated GAN network, the synthetic data closely follows the distribution of the real data as indicated by large p-values. Our method takes minutes to complete training a model, which enables an efficient response to disasters with modern computing facilities such as edge computing.
Keywords: Power Grid; Generative Adversarial Network; Classification; Rare Events.
Voronoi Tesselation Based Load-Balanced Multi-Objective Priority-Based Heuristic Optimization for Multi-Cell Region Coverage with UAVs
by Kemal Kilic, Orhan Gemikonakli, Leonardo Mostarda
Abstract: The existing communication infrastructure can be disrupted by unexpected events or can be extended for temporary events. In such cases, UAVs can be deployed as mobile base stations to provide efficient and economic communication. This task requires the optimization of several conflicting objectives. Our work focused on the temporal region coverage with UAVs. The proposed multi-objective priority-based optimization framework utilizes an evolutionary heuristic algorithm with a scoring scheme and gives priorities to objectives to achieve such a task. Our previous single Base Station (BS) based optimization framework is extended by considering Voronoi Tesselation of the coverage region based on the existing nearby BSs. This scheme provides "cells" where the Virtual Base Stations (VBSs) are placed and helps UAVs to receive the best signal from the related center point of the cell, load balancing of the required bandwidth for the whole mission region.
Keywords: Region coverage; Multi-objective optimization; UAVs; Heuristic Evolutionary Algorithms; Voronoi Tesselation.
A Hybrid Pre-joined Service Network in Graph Database and Memory for Service Composition
by Jing Li, Ming Zhang, Ming Zhu, Lizhen Cui, Yuhong Yan, Bowen Song
Abstract: To make full use of the large space and persistency provided by a database and the fast processing speed of main memory, we propose a Hybrid Pre-joined Service Network method combining in-graph-database and in-memory calculations. This method firstly stores services information and combinations in a graph database. Then, for a users request, it retrieves candidate services from the database and constructs a service network in memory. After that, services are picked and composed to fulfill an optimal solution in memory. We assess the performance of our method by conducting experiments and comparing the approach with other database-based methods. Experiment results indicate the efficiency of this method in comparison, and it can always find solutions and lead to higher users satisfaction.
Keywords: Service Composition; QoS; Graph Database.
A study on data sharing system using ACP-ABE- SE in cloud environment
by YongWoon Hwang, Im Yeong Lee
Abstract: With the development of cloud computing, users can store their data externally and it is convenient to share with other users. However, in a network environment connected to an external cloud, various security threats such as spoofing attacks and collusion attacks may occur, and an attacker's data may be leaked to the external. Therefore, security of data shared in the cloud environment is essential. Among various security technologies, CP-ABE provides data encryption, decryption, and access control. So far, research on CP-ABE methods has continued, but the existing CP-ABE method has a weak security method and an inefficient method. In particular, third parties may violate the privacy of users by inferring the attributes of users accessing data with the access structure contained in the ciphertext. In addition, the cloud server is inefficient because it discovers stored cryptographic statements after users have decrypted them when requesting them, and because the server can identify the decrypted data, confidentiality of the data is not guaranteed. In order to solve this, anonymous CP-ABE methods are being studied to anonymize the access structure of cryptographic statements, and studies are being conducted to apply searchable encryptions to CP-ABE methods. However, problems that can be considered in CP-ABE methods such as increasing ciphertext size and outsourcing servers occur. In this paper, we propose a new ACP-ABE-SE data sharing system that combines anonymous CP-ABE and searchable encryption for secure data sharing in the cloud. In the proposed scheme, a fake access structure is created to protect the privacy of users accessing the ciphertext. In addition, searchable encryption is used to efficiently search for the desired ciphertext among the ciphertext stored on the cloud server. Therefore, it aims to share data safely and efficiently in a cloud environment.
Keywords: Cloud security; ciphertext-policy attribute-based encryption; Attribute-based encryption; Anonymous; Searchable Encryption;.
An Egalitarian Approach of Scheduling Time Restricted Tasks in Mobile Crowdsourcing for Double Auction Environment
by Jaya Mukhopadhyay, Vikash Kumar Singh, Sajal Mukhopadhyay, Anita Pal, Meghana M. D
Abstract: Crowdsourcing with the intelligent agent felicitated with portable smart devices is becoming increasingly popular in recent years as it has opened up meeting an extensive list of real life problems such as supervising pollution control, giving information about the damaged road and so on. In literature this isrnpopularly known as Mobile crowdsourcing (MCS) or participatory sensing (PS).rnHow to motivate the task executors, has been a challenge in MCS environmentrnwhen the tasks are available to be performed. To mitigate this issue several auction based schemes are proposed where agents are motivated by providing incentives. In this paper we have addressed this motivation issue in a double auction environment when the tasks are time restricted (each task has a start time and a finish time) and may be overlapped. Here, we have taken an egalitarian approach so that a balanced allocation of tasks can be established to the participating agents (task executors). In our approach, first the tasks that are imparted by the task providers, are partitioned into several slots in a non-overlapping manner and then allocated to the participating agents through double auction. It is proved that our mechanism is polynomial time solvable (such as to ensure that it can be scaled with varying input sizes) and abide by the economic robustness (i.e. supported with truthful, individually rational, budget balanced properties). It is also exhibited via simulation that our proposed mechanism will perform better when the agents (both task executors and task providers) misreport their valuations.
Keywords: Participatory sensing; Strategic; Truthful; Auction; Scheduling.
An integrated blockchain network with energy router based trading strategy for optimal energy management
by Yunfei Du, Zia Ullah, Xianggen Yin, Jinmu Lai, Zhen Wang
Abstract: The new emerging idea of energy router and energy blockchain network (EBN) transactions gaining remarkable attention in the smart grid and advanced power system due to a variety of applications and innovative solutions for efficient energy trading. However, the implementation of the energy blockchain network (EBN) transactions raises various challenges such as lack of power loss consideration and congestion management, insufficient computing, the storage capacity of trading nodes, and higher transaction costs due to weak central organization low performance of smart meters. This paper introduces a new energy trading approach, focused on the energy router and EBN together, to effectively counter energy trading challenges and perform optimal energy trading. The proposed systematic design of energy trading has divided into four steps which execute (i) forming trading pairs through a two-stage Stackelberg game model, (ii) checking security by using the characteristics of directional power flow of energy router, (iii) congestion management through the minimum loss multi-path power transmission (MLMPT) scheme, (iv) forming smart contracts, and settling transactions. The performance of the proposed framework is verified using 13 nodes EBN implementation, where the results obtained demonstrate the accurate energy trading, balanced trading prices, even in congestion management conditions. Despite congestion prices and trading pairs, the MLMPT scheme in the proposed design also minimizes the power loss to protect sellers' interests.
Keywords: Energy Blockchain Network; Energy Router; Power Transactions; Energy Trading; Stackelberg Master-Slave Game Mode; Congestion Management.
A Sequence-to-Sequence Traffic Predictor on Software-Defined Networking
by Wenchuan Yang, Rui Hua, Qiuhan Zhao
Abstract: Network traffic prediction is very important for load balancing and network planning. Under the current network architecture, it is difficult to realize the collection, prediction and centralized management of traffic. This paper proposes an attention-based traffic predictor (ATP) model to achieve traffic prediction in a software-defined network (SDN) environment. To improve the accuracy and efficiency of a prediction, improvements are made from three aspects: data, model, and evaluation optimization. First, during the traffic data acquisition phase, to reduce the resource consumption of the request caused by acquiring realtime network traffic information and to maintain the accuracy of the prediction, a combination of lower sampling frequency and data augmentation is adopted. Second, based on the long correlation and self-similar characteristics of network traffic, a sequence-to-sequence model with attention (Seq2Seq+Attention) is selected for network traffic prediction. Finally, the traditional mean squared error (MSE) evaluation method sets the same weight for samples of different values, which is not suitable for network traffic prediction. Therefore, this paper proposes an improved weighted MSE evaluation method. Experiments show that the proposed method can maintain the prediction accuracy while reducing the sampling frequency by 50%. The weighted MSE evaluation method can improve the accuracy by 5.37% compared with the original MSE evaluation method. On the basis of highly accurate traffic forecasts, it is possible to further realize intelligent control of network traffic, improve network utilization, reduce network delay, and improve the efficiency of intelligent services.
Keywords: Software defined networking; Network traffic prediction; Data augmentation; Seq2Seq+Attention; Weighted MSE.
Machine Learning Applications for Fog Computing in IoT: A Survey
by Mitra Mousavi, Javad Rezazadeh, Omid Ameri Sianaki
Abstract: Today, Internet of Things (IoT) has become an important paradigm. Everyday increasing number of IoT applications and services emerge. Smart devices connected by the IoT generate significant amounts of data. Analysis IoT sensor data using machine learning algorithms is a key to achieve useful information for prediction, classification, data association and data conceptualisation. Offloading input data to cloud servers leads to increased communication costs. Undertaking Data analytics at the network edge using fog computing enables the rapid processing of incoming data for real-time response. In this paper, we examine the results of using different machine learning algorithms on fog nodes based on existing research. These results are low latency, high accuracy and low bandwith. Also, this work presents the current fog computing architecture which consists of different layers that distribute computing, storage, control and networking and finally we investigate the challenges and open issues related to the deployment of machine learning on fog nodes.
Keywords: Internet of Things (IoT); Fog computing; Machine Learning; Fog-based machine learning.
A Novel Elliptic Curve Cryptography Based System for Smart Grid Communication
by Ajay Kumar, Abhishek Kumar, Kunjal Shah, Suyel Namasudra, Seifedine Kadry
Abstract: Smart Grid describes an electrical grid that has integrated with a fully computerized two-way communication network.
Smart grid (SG) communication has recently experienced attention regarding distributed electric power transmission frameworks. Because of the intricate nature of the intelligent grid and different safekeeping prerequisites, structuring a reasonable validation system is a perplexing task. Therefore, validation and data protection of these devices, including lightweight operations with trivial computations, play an exceptional job in fruitful coordination of SG technologies. In this inquiry, a validation system dependent on Elliptic Curve Cryptography (ECC) for securing the smart grid is proposed. The formal verification of this procedure is implemented utilizing the ProVerif tool along with BurrowsAbadiNeedham logic (BAN logic), which affirms its security strength within sight of a conceivable trespasser. The proposed system is free from all security attacks. The informal security analysis, performance analysis, and assessment of the proposed approach with several other existing systems demonstrate that it is powerful enough in terms of operation cost, storage cost and transmission cost, efficient and stout against numerous security attacks.
Keywords: Validation; Elliptic Curve Cryptography; ProVerif tool; Burrows Abadi Needham logic; Security Analysis.
Attack Resistance based Topology Robustness of Scale free Internet of Things for Smart Cities
by Talha Naeem Qureshi, Nadeem Javaid, Ahmad Almogren, Zain Abubaker, Hisham Almajed, Irfan Mohiuddin
Abstract: Increase in growth of Internet of Things (IoT) devices leads to an exponential increase in IoT applications. This exponential growth of devices increases the complexity of IoT network. Increase in network complexity intensifies the risks against topology robustness. An IoT network acts as a core enabler for converting conventional cities to smart cities. These devices produce large amount of data related: nodes' geographic location, connected neighbors, etc. Therefore, improving topology robustness of the IoT networks against targeted and malicious attacks is a prime issue. Four algorithms: Enhanced Angle Sum Operation EASO-ROSE, Enhanced ROSE, Adaptive Genetic Algorithm (AGA) and Cluster Adaptive Genetic Algorithm (CAGA) are proposed to cater the topology robustness issue for IoT enabled smart cities. In addition, the proposed solutions keep the nodes' initial degree distribution unchanged by maintaining the scale-free nature of the topology. Enhanced ROSE and EASO-ROSE significantly improve the topology robustness by calculating nodes' degree difference along with rearranging the surrounding angles according to the highest degree node. CAGA and AGA also significantly improve the topology robustness by using adaptive probabilities of crossover and mutation that guide algorithm to converge towards global optimal solution. Extensive simulations verify that proposed algorithms outperform the ROSE and simulating annealing. Moreover, the Enhanced ROSE and EASO-ROSE are compared with ROSE and simulating annealing. Furthermore, CAGA and AGA algorithms are compared with simulating annealing and hill climbing. Enhanced ROSE, EASO-ROSE, CAGA and AGA perform 61.3%, 48.3%, 45.5% and 34.95%, respectively better as compared to simulating annealing.
Keywords: Internet of things; topology robustness; malicious attacks; data driven; scale-free.
An Adaptive Enhanced Differential Evolution Strategies for Topology Robustness in Internet of Things
by Talha Naeem Qureshi, Nadeem Javaid, Ahmad Almogren, Asad Ullah Khan, Hisham Almajed, Irfan Mohiuddin
Abstract: Internet of Things (IoT) is the backbone of any automation process and a pivot point to lay down the base of smart cities. To effectively increase the robustness of the IoT network without changing the degree distribution of nodes is still a challenging issue. To tackle this problem, we have proposed two algorithms, such as Enhanced Differential Evolution (EDE) and Adaptive EDE (AEDE). Initially, we have generated scale-free topologies according to the characteristics and behavior of IoT networks in real world. Geographic information of IoT sensors are gathered from big data server using a mechanism that saves the IoT sensors from computational overhead of algorithms. Proposed algorithms effectively improve the robustness of the IoT network without changing the degree distribution of nodes. The algorithms are capable to converge the results towards the global optima from solution space along with the fast convergence speed. The AEDE dynamically changes the probabilities of multiple operations of the EDE along with the changing environment. Also, it maintains the balance between the diversity of solution space and the convergence speed through adaptive probabilities. The proposed algorithms show better performance in terms of increasing network's robustness as compared to previous schemes. To validate the proposed concept, both proposed algorithms are compared with well known previous algorithms including the Genetic Algorithm (GA), the Simulating Annealing (SA) and the Hill climbing Algorithm (HA). The proposed algorithms are also compared with benchmark schemes in large scale networks with the changing number of nodes and edge densities as well. The EDE performs 7.13%, 31.6% and 41.8% better as compared to GA, SA and HA, respectively.
Keywords: Data driven; Internet of things; Malicious attacks; Scale-free; Topology robustness.
Special Issue on: Security for Cloud Computing
Searchable Symmetric Encryption Based on the
Inner Product for Cloud Storage
by Jun Yang, Shujuan Li, Xiaodan Yan, Baihui Zhang, Baojiang Cui
Abstract: Searchable encryption enables the data owner to store their own data after
encrypting them in the cloud. Searchable encryption also allows the client to search over
the data without leaking any information about it. In this paper, we rst introduce a
searchable symmetric encryption scheme based on the inner product: it is more ecient
to compute the inner product of two vectors. In our construction, the parties can be Data
Owners, Clients or the Cloud Server. The three parties communicate with each other
through the inner product to achieve the goal that the client can search the data in the
cloud without leaking any information on the data the owner stored in the cloud. We then
perform a security analysis and performance evaluation, which show that our algorithm
and construction are secure and ecient.
Keywords: Searchable Encryption; Searchable Symmetric Encryption; Inner Product;
the Cloud Server; Security.
Special Issue on: Web of Things (WoT) and its Intelligent Data Processing Services
QoS-Prioritized Media Delivery with Adaptive Data Throughput in IoT-Based Home Networks
by Chih-Lin Hu, Liang-Xing Kuo, Yung-Hui Chen, Thitinan Tantidham, Pattanasak Mongkolwat
Abstract: To maintain network efficiency of a home network, it is crucial to moderately distribute limited data throughput and bandwidth resources to various home devices and services that require different network resource provisions. This paper accounts for QoS differentiation and fairness of network resource allocation which are essential to the design of an efficient IoT-based home media delivery mechanism against network traffic dynamics. Accordingly, our study in this paper proposes a QoS-prioritized media delivery mechanism with adaptive data throughput. This mechanism design includes several functions: assigning weights of relative importance to home service types, adjusting media quality of home services, reducing data workload against traffic congestion, and fairly distributing free bandwidth to prioritized home services. Furthermore, our study fulfills a simplified proof-of-concept implementation in a small-scaled IoT-based home network. Practical experiments can generate real measure data for performance examination. Results show that the proposed mechanism is able to sustain not only QoS differentiation like data throughput and delay, but also fair user satisfaction with comfortable media playing quality.
Keywords: media streaming; quality of servie (QoS); service differentiation; network resource allocation; home networks; smart home; internet of things (IoT).
Association between Alcohol Consumption and
by Jianqiang Li, Xi Xu, Yan Pei, Jason C. Hung, Weiliang Qiu
Abstract: Both Telomere length and alcohol consumption have an important impact on biological age and carcinogenesis. Researchers have conducted many efforts on this subject to study the relationship between alcohol consumption and telomere length. However, there is no agreement has been reached on this issue. In our study, a meta-analysis is performed and relevant investigation results from previous literature are integrated. 21 articles published between 2000 and 2016, which comprise 27 analyses with a total sample size 35,891, have met our eligibility criteria. A significant relationship between alcohol consumption and telomere length is found (Fishers combined p-value = 3.52E-8 and Liptaks weighted p-value = 8.24E-3). Whether the relationship between alcohol consumption and telomere length is significant also varies with study type (cohort, case-control, or cross-sectional) and study population (Europe, Asia, American, or Australia). It is deduced by combined evidence that alcohol consumption is associated with telomere length. In the future, the consistent quantifications of alcohol consumption and telomere length will benefit further aggregation of the evidence from varies studies.
Keywords: Age; Alcoholism; Cancer; Meta-Analysis; Mitotic Clock; Telomere Length.
Data analysis of simulated WoT-based anti-crime scenario
by Chih-Chi Kuang, Kuei Min Wang, Lin Hui, Chuan-Yu Chang, Kuang Hui Chiu
Abstract: Police work is characterized by high risk and high cost. When a police task has no real-time information, the opaque situation could seriously compromise the success of the task. With the ap-pearance of the Internet of Things (IoT), it has become possible to create the smart things we need and to make them more effective. In addition, the Web of Things (WoT) has emerged recently. It enables us to integrate IoT more easily so that it can be applied to develop a specific system with a wireless sensor network (WSN) and platforms to perform specific tasks. This study proposes a methodology for examining the effectiveness of the WoT in support of the police work of freeing hostages held by terrorists. The proposed WoT-based police rescue squad force (RSF) concept was modeled and simu-lated using the developed Monte Carlo Simulation. The current and WoT-based RSF were the alternatives analyzed by simulation. The simulation results and t tests showed that there is a significant difference between the current and WoT-based RSF. The information gained from the simulation can support the police authoritys decision makers in upgrading the specific police information equipment with less risk, less cost and high effectiveness. The limitation is that the information from the study can only be used in the situation of hostage rescue tasks, i.e. it is inappropriate to apply it to other police work.
Keywords: WoT; IoT; RSF; Simulation; Terrorist; UAV.
A Quadratic Fusion Estimating Model Based on the Clustering Kernel for Real-Time Data in Web of Things
by Chao Li, Zhenjiang Zhang, Yingsi Zhao, Peng Zhang, Bo Shen
Abstract: Real-time data processing is a very important part of data processing in the Web of things (WoT). The devices in WoT collect data and provide real-time information. The accuracy of the collected data is critical to provide valid results. Many existing methods are devoted to modifying filter algorithms. However, little attention is devoted to the inner relationship of data and data accuracy. In the present study, a quadratic filter model based on the clustering kernel is presented. First, the common filter method is used. Second, the clustering algorithm is adopted to deliver the clustering result. The attractor of the class is gained to the clustering kernel. Finally, the quadratic filter is processed according to the clustering kernel. The simulations show that the proposed model can increase the data accuracy.
Keywords: Quadratic Estimating; Clustering Kernel; Web of Things; Fusion Estimating; Real-Time Data.