Article de revue avec comité de lecture (5)
QI Chengming , ZHOU Zhangbing , SUN Yunchuan , SONG Houbing , HU Lishuan , WANG Qun
Feature selection and multiple kernel boosting framework based on PSO with mutation mechanism for hyperspectral classification. Neurocomputing, january 2017, vol. 220, pp. 181-190
Hyperspectral remote sensing sensors can capture hundreds of contiguous spectral images and provide plenty of valuable information. Feature selection and classification play a key role in the field of HyperSpectral Image (HSI) analysis. This paper addresses the problem of HSI classification from the following three aspects. First, we present a novel criterion by standard deviation, Kullback-Leibler distance, and correlation coefficient for feature selection. Second, we optimize the SVM classifier design by searching for the most appropriate value of the parameters using particle swarm optimization (PSO) with mutation mechanism. Finally, we propose an ensemble learning framework, which applies the boosting technique to learn multiple kernel classifiers for classification problems. Experiments are conducted on benchmark HSI classification data sets. The evaluation results show that the proposed approach can achieve better accuracy and efficiency than state-of-the-art methods.
XIANG Jianming , ZHOU Zhangbing , SHU Lei , RAHMAN Taj , WANG Qun
A mechanism filling sensing holes for detecting the boundary of continuous objects in hybrid sparse wireless sensor networks. IEEE access, 2017, pp. 1-14 (document in press - published online 19 Jan 2017)
Nowadays, the rapidly developed Internet of Things require the ability handling information efficiently to deal with the intelligent applications. Wireless sensor networks (WSNs), which act as an important interface between physical environment and Internet of Things, have been applied in numerous applications. As a kind of important application of WSNs, the continuous objects boundary detection is popular in industry. However, the long-term maintenance for the traditional WSNs, which are used to monitor the leakage of continuous objects, is expensive. Thus, we use sparse WSNs to address this issue. But, the inaccuracy of the sparse network is a big problem while the information of continuous objects is used to arrange retreat path for people. To access this problem, we propose our mechanism which used hybrid network to compromise the accuracy and cost of maintenance. The sensing holes will be detected by using Voronoi diagram, before the network starts to work. After the static sensor nodes get the value of the toxic air, the mechanism can calculate the high variation location, which give weights to the sensing holes, in the static sensor networks. Thus, the sensing holes which selected by both spatial and data variation factors will be list in a target nodes list for the mobile sensor node. Finally, the optimal path considering both distance and priority for the mobile sensor will be plan out. Experimental evaluation shows that there is an optimal amount of the static nodes decided by the sensing radius and the size of area. And it reduces the energy consumption by the static networks.
ZHOU Zhangbing , CHENG Zehui, ZHANG Liang-Jie, GAALOUL Walid, NING Ke
Scientific workflow clustering and recommendation leveraging layer hierarchical analysis. IEEE transactions on services computing, 2017, pp. 1-14 (document in press - published online 16 Mar 2016)
This article proposes an approach for identifying and recommending scientific workflows for reuse and repurposing. Specifically, a scientific work- flow is represented as a layer hierarchy, which specifies hierarchical relations between this workflow, its subworkflows, and activities. Semantic similarity is calculated between layer hierarchies of workflows. A graphskeleton based clustering technique is adopted for grouping layer hierarchies into clusters. Barycenters in each cluster are identified, which refer to core workflows in this cluster, for facilitating cluster identification and workflow ranking and recommendation. Experimental evaluation shows that our technique is efficient and accurate on ranking and recommending appropriate clusters and scientific workflows with respect to specific requirements of scientific experiments.
ZHOU Zhangbing , ZHAO Deng, HANCKE Gerhard , SHU Lei , SUN Yunchuan
Cache-aware query optimization in multi-application sharing wireless sensor networks. IEEE transactions on systems, man, and cybernetics. Systems, 2017, pp. 1-17 (document in press - published online 24 Aug 2016)
Hosting multiple applications in a shared infrastructure of wireless sensor networks is a trend nowadays, and sharing sensory data for answering concurrent applications is a promising and energy-efficient strategy. To address this challenge, this paper proposes an energy-efficient query optimization mechanism for supporting multiple concurrent applications leveraging our two-tier cooperative caching mechanism. Specifically, query requests for concurrent applications are represented as binary strings, which are reduced to a single one for avoiding the reprocessing of shared subquery requests. This reduced query request is answered through our cooperative caching mechanism, where sensory data, which are highly possible to be reused for answering forthcoming query requests, are cached at the sink node (SN). Besides, the gray model GM(1, 1) is adopted for forecasting sensory data units which may be interested mostly by forthcoming query requests. These units of sensory data may be prefetched from the network and cached at the SN. Experimental evaluation shows that this approach can reduce the energy consumption significantly, and improve the network capacity to an extent, especially when the number of concurrent query requests is relatively large
ZUO Liyun , SHU Lei , DONG Shoubin , ZHU Chunsheng , ZHOU Zhangbing
Dynamically weighted load evaluation method based on self-adaptive threshold in cloud computing. Mobile networks and applications, 2017, pp. 1-15 (document in press - published online 21 Jan 2016)
Cloud resources and their loads possess dynamic characteristics. Current research methods have utilized certain physical indicators and fixed thresholds to evaluate cloud resources, which cannot meet the dynamic needs of cloud resources or accurately reflect their resource states. To address this challenge, this paper proposes a Self-adaptive threshold based Dynamically Weighted load evaluation Method (termed SDWM). It evaluates the load state of the resource through a dynamically weighted evaluation method. First, the work proposes some dynamic evaluation indicators in order to evaluate the resource state more accurately. Second, SDWM divided the resource load into three states, including Overload, Normal and Idle using the self-adaptive threshold. It then migrated those overload resources to a balance load, and releases the idle resources whose idle times exceeded a threshold to save energy, which could effectively improve system utilization. Finally, SDWM leveraged an energy evaluation model to describe energy quantitatively using the migration amount of the resource request. The parameters of the energy model were obtained from a linear regression model according to the actual experimental environment. Experimental results showed that SDWM is superior to other methods in energy conservation, task response time, and resource utilization, and the improvements are 31.5 %, 50 %, 50.8 %, respectively. These results demonstrate the positive effect of the dynamic self-adaptive threshold. More specially, SDWM shows great adaptability when resources dynamically join or exit
Communication dans une conférence à comité de lecture (1)
EL RACHKIDI Elie, AGOULMINE Nazim, BELAID Djamel, CHENDEB Nada
Towards an efficient service provisioning in Cloud of Things (CoT). GLOBECOM 2016 : IEEE Global Communications Conference, Los Alamitos : IEEE Computer Society, 04-08 december 2016, Washington, Dc, United States, 2017, pp. 1-6, ISBN 978-1-5090-1328-9
The Cloud offers virtually unlimited resources and the ability to scale up or down applications as needed on the fly. Hence, the Cloud emerged as a suitable solution for large-scale IoT applications to cope with the rapidly increasing devices and data volume. Furthermore, IoT broadened the scope of the Cloud to the real world and enabled new service models such as the Sensing as a Service model. The convergence of both technologies stimulated innovations in both fields, we refer to this convergence as the Cloud of Things. The Cloud of Things enables users to request a complex IoT service (IoT application composed of several interconnected micro services) and deploy it seamlessly. However, deploying a complex IoT service in the Cloud of Things infrastructure is a difficult process due to the different types of physical nodes (Cloud data centers, IoT devices, gateways, etc.) and multiple architectures to collect and process data. Furthermore, network usage largely depends on the placement of different services across the network. In this paper, we present an efficient provisioning model of IoT services formulated as a Mixed Integer Problem. The objective is to minimize the cost of the deployment of IoT services in Cloud of Things infrastructure, through optimizing resources usage across physical nodes and bandwidth consumption over the network