Search

With the development of deep neural networks, the smart edge-cloud scenario is expected to meet the diverse and personalized requirements of users. However, the exiting edge-cloud architecture requires the deployment of manually designed neural networks. Providing adaptive network architecture in this environment, especially for heterogeneous edge devices, becomes the major challenge. To this end, we design the customized network architecture with neural architecture search for adaptively neural network search. Specifically, our architecture can be identified in three parts: cloud layer, equipped with huge amounts of computing resources, conducts multi-objective architecture search on a proxy dataset; the edge layer aims to train the optimal architecture based on the target dataset from scratch, which makes full use of the computing resource of the edge server; and the user layer deploys the multi-objective model in diverse devices for inference in real time. Furthermore, we deploy an industrial sensor monitoring scenario as a case study to search for the temporal convolutional network, demonstrating the effectiveness of the proposed architecture. Experimental results show the feasibility of the proposed architecture. [ABSTRACT FROM AUTHOR]

The parameter server architecture has shown promising performance advantages when handling deep learning (DL) applications. One crucial issue in this regard is the presence of stragglers, which significantly retards DL training progress. Previous solutions for solving stragglers may not fully exploit the computation resource of the cluster as evidenced by our experiments, especially in the heterogeneous environment. This motivates us to design a heterogeneity-aware parameter server paradigm that addresses stragglers and accelerates DL training from the perspective of computation parallelism. We introduce a novel methodology named straggler projection to give a comprehensive inspection of stragglers and reveal practical guidelines to solve this problem in two aspects: (1) controlling each worker's training speed via elastic training parallelism control and (2) transferring blocked tasks from stragglers to pioneers to fully utilize the computation resource. Following these guidelines, we propose the abstraction of parallelism as an infrastructure and design the Elastic-Parallelism Synchronous Parallel (EPSP) algorithm to handle distributed training and parameter synchronization, supporting both enforced- and slack-synchronization schemes. The whole idea has been implemented into a prototype called Falcon which effectively accelerates the DL training speed with the presence of stragglers. Evaluation under various benchmarks with baseline comparison demonstrates the superiority of our system. Specifically, Falcon reduces the training convergence time, by up to 61.83, 55.19, 38.92, and 23.68 percent shorter than FlexRR, Sync-opt, ConSGD, and DynSGD, respectively. [ABSTRACT FROM AUTHOR]

Deploying blockchain in IoT is an effective way to address traditional security issues. However, existing approaches have two major limitations: since the blockchain itself is subject to attacks, including selfish mining, double spending, and distributed denial of service attacks, IoT smart devices are also vulnerable once hackers successfully invade blockchain systems; due to the heterogeneity and resource limitations of IoT devices, the deployment of the existing blockchain systems in the IoT scenario cannot reflect strong adaptability and meet IoT service requirements. In this article, we introduce Spacechain, a secure and high-performance blockchain system with three-dimensional ledger architecture, to enable blockchain open in IoT. Specifically, we first design a three-dimensional architecture with novel data structures to deal with the heterogeneity and scalability of IoT networks. Then, we propose the Three-Dimensional Greedy Heaviest-Observed Sub-Tree (3D-GHOST) consensus mechanism for Spacechain to improve security and network performance. Additionally, we conduct detailed security analysis and extensive experimental verification to demonstrate the performance of Spacechain. [ABSTRACT FROM AUTHOR]

With the rapid development of society and the economy and the increasing seriousness of environmental problems, renewable energy and high-quality energy services in low-carbon communities have become popular research topics. However, a large number of volatile distributed generation power systems in the community are connected to the grid. It is difficult to stabilize and efficiently interact with fragmented and isolated energy management systems, and it is difficult to meet energy management needs in terms of low-carbon emissions, stability, and intelligence. Therefore, by considering operation costs, pollution control costs, energy stability, and plug-in hybrid electric vehicles, this article proposes a regional energy supply model called community energy Internet and builds a low-carbon community energy adaptive management model for smart services. Then, to address energy supply instability, an adaptive feedback control mechanism developed based on model predictive control is introduced to adapt to the changing environment. Finally, a long short-term memory-recurrent neural network-based Tabu search is introduced to prevent the multiobjective particle swarm optimization algorithm from easily falling into a local optimum. The simulation results show that the proposed model can effectively realize the optimal allocation of energy, which solves the problem of fragmented energy islands caused by distributed power access. This method has quality of service benefits for users, such as cost, time, and stability, and realizes wide interconnections, high intelligence, and low-carbon efficiency of community energy management. [ABSTRACT FROM AUTHOR]

With the fast-growing demands for efficient and scalable content distribution, information-centric networking (ICN) can be a promising candidate for future networks. ICN promises direct retrieval of the content using its unique, persistent, and location-independent name. In ICN, all nodes work together to scale up content delivery, but security issues caused by malicious behaviors of ICN nodes cannot be avoided. To this end, in this article we develop a trust-enhanced blockchain based ICN (BICN) architecture for content delivery. In our proposed architecture, the whole process of content delivery is first traced in an implicitly trusted way by exploiting the excellent properties of the blockchain, so that the malicious ICN nodes can be located. Second, we leverage transactions to record the mapping between human-readable name and self-certifying name in BICN, which supports convenient and trusted alteration for users' requirements. Moreover, we present a case study in BICN based smart grid for energy data delivery, where we perform security analysis and conduct experiments. Numerical results show the superior performance of our proposal. Finally, we present open issues for future work. [ABSTRACT FROM AUTHOR]

Rapid economic development has been observed worldwide, which has caused environmental problems to worsen. Thus, the Energy Internet (EI), which accesses renewable energy and provides high-quality power services, has recently become a hot issue. As a subnet of the EI, an energy local area network (ELAN) consists of renewable power generation equipment, controllable distributed power generation equipment, storage systems, electric vehicles, and a large number of loads. Energy management is required for economic, environmental, and safety considerations. This paper proposes an energy management optimization model that addresses ELAN operations and includes pollution treatment fees; this model provides intelligent control of the charging and discharging of plug-in hybrid electric vehicles (PHEVs). This model achieves a nonlinear energy management optimization for an ELAN. To promote optimal performance, an improved comprehensive learning particle swarm optimization (CLPSO) algorithm is presented; it combines Tabu Search (TS) and CLPSO to avoid local optima. To verify the performance of our model, two experimental scenarios are built. The simulation results show that our energy management optimization model fulfills the optimal allocation of energy and that the PHEV intelligent charging/discharging strategy promotes economic benefits for the network. [ABSTRACT FROM AUTHOR]

The 5G system has been recognized as the most promising technology to provide high-quality network services. As a huge number of networking and computing equipments that generate big data are integrated into the 5G system, energy efficiency becomes the major challenge in building a green 5G system. In this article, we propose a software-defined green 5G system for big data, which consists of three planes: the control plane, the data plane and the energy plane. The data plane contains networking and computing equipments, which can be powered by both traditional grid and renewable energy sources in the energy plane. The control plane monitors the system status and configures the corresponding equipments to achieve energy efficiency and quality-of-service. Furthermore, to reduce the overhead of this software- defined architecture, we investigate a FRS to eliminate redundant system monitoring information. To integrate features in software-defined architecture, we propose an AIFS to mine latent rules among features. Simulation results indicate that our proposals achieve higher efficiency in the green 5G system. [ABSTRACT FROM AUTHOR]

With the popularity of smart devices, multi-access edge computing (MEC) has become the mainstream of dealing with big data in heterogeneous Internet of Things (H-IoT). MEC makes full use of the computing power of edge nodes, which greatly reduces the computing pressure of data centers, and brings great convenience to the storage and processing of big data. However, it is easy to become the object of hacker attacks due to the lack of centralized management of distributed nodes. Once these nodes are compromised, a series of privacy issues can happen. In this article, we first overview the architecture of MEC for H-IoT. The MEC covers three-level advanced functional entities, including moblie edge (ME) system-level, ME host-level and ME network- level. Second, we draw our attention to the privacy issues in the MEC, especially in data aggregation and data mining. In addition, we consider machine learning privacy preserving as a case study in the application of MEC. Simulation results are shown to reveal the feasibility of the proposed method. Finally, we propose open issues for future work. [ABSTRACT FROM AUTHOR]

When there exist not enough historical defect data for building an accurate prediction model, semisupervised defect prediction (SSDP) and cross-project defect prediction (CPDP) are two feasible solutions. Existing CPDP methods assume that the available source data are well labeled. However, due to expensive human efforts for labeling a large amount of defect data, usually, we can only utilize the suitable unlabeled source data. We call CPDP in this scenario as cross-project semisupervised defect prediction (CSDP). Although some within-project semisupervised defect prediction (WSDP) methods have been developed in recent years, there still exists much room for improvement on prediction performance. In this paper, we aim to provide a unified and effective solution for both CSDP and WSDP problems. We introduce the semisupervised dictionary learning technique and propose a cost-sensitive kernelized semisupervised dictionary learning (CKSDL) approach. CKSDL can make full use of the limited labeled defect data and a large amount of unlabeled data in the kernel space. In addition, CKSDL considers the misclassification costs in the dictionary learning process. Extensive experiments on 16 projects indicate that CKSDL outperforms state-of-the-art WSDP methods, using unlabeled cross-project defect data can help improve the WSDP performance, and CKSDL generally obtains significantly better prediction performance than related SSDP methods in the CSDP scenario. [ABSTRACT FROM AUTHOR]

Vehicle-to-grid (V2G) networks are expected to balance the supply and demand in smart grid by reducing the peak-to-average ratio of power grid load curve. We are entering the era of wireless communication, where we can enjoy various advantages such as lower cost, lower battery consumption, and lower access latency. We believe advanced wireless communication techniques have great potential to further promote the economical deployment of V2G network and lower energy consumption. In this article, we address the green V2G network for efficient energy management. However, we still face many challenging issues even if we exploit the promising wireless communication technique in green V2G networks. For example, it becomes more and more challenging to achieve the efficiency and economy of renewable energy resource allocation due to the increasing number of electric vehicles and limited capacity of local aggregators (LAGs). To address the issues, we consider a software-defined green V2G network for energy management, which consists of three planes: management plane, control plane, and data plane. Specifically, the control plane is aimed at guiding both data flow and energy flow to implement an efficient and economic strategy for energy scheduling, while the data plane collects the information through LAGs for the customized services in the management plane. Additionally, we present an energy management scheme of charging stations as a case study. Simulation results reveal that our proposals could achieve delightful performance on global optimization in the software-defined green V2G network. [ABSTRACT FROM PUBLISHER]

EVs provide a promising green solution to reduce dependence on fossil fuel and the emission of greenhouse gas. This article concerns hybrid EV systems that contain both wired charging and wireless charging vehicles. To the best of our knowledge, we are the first to discuss how to jointly use these two types of charging methods whose advantages are complementary to each other; for example, wireless charging has better user friendliness and flexibility, while wired charging can provide higher charging efficiency and bidirectional energy transmission. We believe such a hybrid system can benefit us much more than either a pure wireless or pure wired EV charging system. On the other hand, we also pay attention to a practically important problem with DSM. The existing schemes can hardly ensure efficient and stable charging services in the foreseeable future due to the rapidly increasing number of electric vehicles and the limited capacity of local grids. To address this issue, we propose a software-defined green framework for hybrid EV charging networks, which consists of three planes: the application plane provides customized services for users; the control plane aims to guide both data flow and energy flow to implement an efficient and economic strategy for EV charging; and the physical plane collects the information from a large number of EVs and other infrastructures. Furthermore, we present a DSM approach for EV charging as a case study. Simulation results reveal that our proposals could achieve delightful DSM performance in the proposed software-defined EV charging networks. Although the proposed hybrid framework offers promising opportunities, we still face some technical challenges that solicit further research efforts from both academia and industry, which are discussed in detail at the end of this article. [ABSTRACT FROM PUBLISHER]

With the significant improvements in mobile digital devices and wireless networking technologies, we have witnessed the explosion of multimedia data. Because it is dynamic, vast in volume, and heterogeneous, this data not only evokes various novel data-driven services and applications, but also brings considerable security threats. In this article, the authors focus on privacy leakage issues in multimedia systems and study how to maximize the total privacy weights and upgrade the security level given predefined time and resource constraints. To this end, they propose a selective privacy-preserving method that adaptively allocates encryption resources according to the privacy weight and execution time of each data package. That is, it selects the encryption method with the appropriate complexity and security level for each multimedia data package. It first divides the data randomly into two parts, then performs XOR operations and generates cipher keys in different cloud storages to prevent users’ original information from being attacked by untrusted cloud operators. Extensive simulation results have demonstrated the advantages and superiority of the proposed method over previous schemes. This article is part of a special issue on cybersecurity. [ABSTRACT FROM AUTHOR]

This paper deals with the secure communication issues of wireless cooperative networks in the presence of multiple friendly but selfish intermediate nodes. On account of the broadcast nature of wireless communications, it is assailable to malicious eavesdropping. To tackle this challenge in the paper, we present a relay and jammer selection strategy that selects the jammer and relay nodes from intermediate nodes to improve the security of the eavesdropping attacks. The jammer is used to broadcast artificial interference noise on the eavesdropper. The relay acts as a traditional relay that retransmits source signal from the source to the intended destination. For achieving the maximum secrecy capacity by the selected nodes, we introduce a power allocation approach of intermediate nodes that is formulated as a Bertrand game based on price competition. We prove that an optimal pricing scheme can maximize the secrecy capacity and achieve the optimal profits for the selfish friendly nodes. Then, a new particle swarm with simulated anneal optimization algorithm is employed to obtain the solution of pricing and node selection. Finally, the simulation results verify our theoretical analysis. [ABSTRACT FROM PUBLISHER]

Advanced metering infrastructure (AMI) is an important component for a smart grid system to measure, collect, store, analyze, and operate users consumption data. The need of communication and data transmission between consumers (smart meters) and utilities make AMI vulnerable to various attacks. In this paper, we focus on distributed denial of service attack in the AMI network. We introduce honeypots into the AMI network as a decoy system to detect and gather attack information. We analyze the interactions between the attackers and the defenders, and derive optimal strategies for both sides. We further prove the existence of several Bayesian-Nash equilibriums in the honeypot game. Finally, we evaluate our proposals on an AMI testbed in the smart grid, and the results show that our proposed strategy is effective in improving the efficiency of defense with the deployment of honeypots. [ABSTRACT FROM PUBLISHER]

Driven by the rapidly growing demand for multimedia service, CCN is conceived as an important component for the next-generation network to facilitate information distribution and meet individual needs. This article investigates how to integrate the crowdsourcing technique and the CCN into a coherent whole, for the purpose of significantly improving the performance of CCN. Specifically, we first propose a CCCN that mainly contains two parts, i.e., mobile participants and a server cloud. Second, we design a complementary social-enhanced communication strategy that exploits the social characteristics for this framework, in order to develop a systematic understanding of the interplay between social characteristics and crowdsourcing. Extensive simulations have been conducted, and the results demonstrate that our social strategy can achieve significant performance improvement. Finally, we discuss open issues about security and management that may arise in the future CCCN framework. [ABSTRACT FROM AUTHOR]

Wireless communications are vulnerable to eavesdropping attacks due to their broadcast nature. To deal with their emerging challenge of physical layer security, in this paper, we study the antieavesdropping problem in the presence of selfish jammers, who desire to achieve maximum profit for themselves. We consider both the single-channel multijammer (SCMJ) model and the multichannel single-jammer (MCSJ) model. We investigate the interaction between the source that transmits secret information and friendly jammer nodes who assist the source by interfering with the eavesdropper. This problem is formulated as an oligopoly market consisting of a few firms and a buyer. By modeling the problem as a Bertrand game based on price competition, we obtain the optimal pricing scheme for the friendly, while for selfish jammers, the utility of those jammers is maximized. For the SCMJ model, we prove the existence of Bertrand Equilibrium by deriving a closed-form expression for the optimal price strategy. For the MCSJ model, a closed-form expression for power allocation is derived, based on which a new algorithm is designed to obtain the optimal strategy of the jammer. Finally, via simulations, we verify our theoretical analysis. [ABSTRACT FROM PUBLISHER]

The development of smart grid brings great improvement in the efficiency, reliability, and economics to power grid. However, at the same time, the volume and complexity of data in the grid explode. To address this challenge, big data technology is a strong candidate for the analysis and processing of smart grid data. In this article, we propose a big data computing architecture for smart grid analytics, which involves data resources, transmission, storage, and analysis. In order to enable big data computing in smart grid, a communication architecture is then described consisting of four main domains. Key technologies to enable big-data-aware wireless communication for smart grid are investigated. As a case study of the proposed architecture, we introduce a big-data- enabled storage planning scheme based on wireless big data computing. A hybrid approach is adopted for the optimization including GA for storage planning and a game theoretic inner optimization for daily energy scheduling. Simulation results indicate that the proposed storage planning scheme greatly reduce [ABSTRACT FROM PUBLISHER]

Making full use of V2G services, EVs with batteries may assist the smart grid in alleviating peaks of energy consumption. Aiming to develop a systematic understanding of the interplay between smart grid and EVs, an architecture for the V2G networks with the EV aggregator is designed to maintain the balance between energy suppliers (the grid side) and consumers (the EV side). We propose a combined control and communication approach considering distributed features and vehicle preferences in order to ensure efficient energy transfer. In our model, the integrated communication and control unit can achieve realtime and intelligent management with the logic controller and collected data. On the consumers' side, we theoretically analyze how to satisfy the charging constraints that we incorporate in the form of willingness to pay, and propose a distributed framework to coordinate the energy delivery behaviors for satisfying service demands. Moreover, illustrative results indicate that the proposed approach can yield higher revenue than the conventional pricing mechanism in V2G networks. [ABSTRACT FROM PUBLISHER]

The article focuses on the aspects of the proposed three layer Internet of Things (IoT) framework which include layers such as sense, control, and gateway for energy efficient system for industrial application. Topics discussed include the creation of energy efficient Industrial IoT (IIoT), the several domains comprises the architecture such as service hosted network, user application, and cloud server, and the aspects of the system architecture which focus on the system's sense entities.

The advanced idea of machine-to-machine technology has attracted a new period of network revolution, evolving into a method to monitor and control global industrial user assets, machines, and the production process. M2M networks are considered to be the intelligent connection and communication between machines. However, the security issues have been further amplified with the development of M2M networks. Consequently, it is essential to pour attention into attack detection and forensics problems in M2M networks. This article puts forward the hybrid attack detection and forensics model in M2M networks. It contains two modules: the attack detection module and the forensics analysis module. In addition, we present a distributed anti-honeypot- based forensics strategy to cope with DDoS attacks in the forensics analysis module. Finally, we also discuss some challenges in M2M network security and forensics. [ABSTRACT FROM AUTHOR]