Junde Song

Beijing University of Posts and Telecommunications, Peping, Beijing, China

Are you Junde Song?

Claim your profile

Publications (153)2.69 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Nowadays in many real-world scenarios, high speed data streams are usually with non-uniform misclassification costs and thus call for cost-sensitive classification algorithms of data streams. However, only little literature focuses on this issue. On the other hand, the existing algorithms for cost-sensitive classification can achieve excellent performance in the metric of total misclassification costs, but always lead to obvious reduction of accuracy, which restrains the practical application greatly. In this paper, we present an improved folk theorem. Based on the new theorem, the existing accuracy-based classification algorithm can be converted into soft cost-sensitive one immediately, which allows us to take both accuracy and cost into account. Following the idea of this theorem, the soft-CsGDT algorithm is proposed to process the data streams with non-uniform misclassification costs, which is an expansion of GDT. With both synthetic and real-world datasets, the experimental results show that compared with the cost-sensitive algorithm, the accuracy in our soft-CsGDT is significantly improved, while the total misclassification costs are approximately the same.
    08/2013;
  • Guan Le, Ke Xu, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Cloud computing is the promising key technology to build future architecture of massive IT systems and one of key benefits of cloud computing is to provide its customers with elastic resources according to the fluctuation of request workloads. In this paper, we propose adaptive resource management policy to handle requests of deadline-bound application with elastic cloud. Adaptive resource management architecture has been proposed, and we divide resource management into two parts, resource provision and job scheduling. We design analytical provision model for adaptive provision based on queuing theory, by introducing a key metric named average interval time. Three job scheduling policies are raised to dequeue appropriate jobs to execute, First-Come-First-Service (FCFS), Shortest Job First (SJF) and Nearest Deadline First (NDF), for different preference toward execution order. Simulation evaluation has been set up with realistic grid workload, and results show that our provisioning model gives elastic resource provisioning for dynamic workload and FCFS achieves better performance compared with other scheduling policies.
    Service Sciences (ICSS), 2013 International Conference on; 01/2013
  • Ke Xu, Hui Zhang, Meina Song, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: The enormous increase in Internet traffic usage has been leading to problems such as increased complexity of routing topology, explosion in routing table entries, provider-dependent addressing, which reduce the speed of network service. The emerging new techniques such as CDN, P2P, VPN, etc. speed up the network from different perspectives. A new speed up system called CANR, content aware and name based routing, is proposed in this paper, which integrates benefits of several existing mehanisms. CANR consists of a cluster of proxy peers deployed in different network domains, which can work as collaborative routers, forwarding requests to each other to speed up the cross-domain visits. CANR can automatically aware the changes of network and re-construct name-based routing table based on a new multi objectie k shortest algoritm by itself, finding a set of cheapest and most fast k routing paths, which is different from current static preconfigured systems.
    Proceedings of the 2012 international conference on Pervasive Computing and the Networked World; 11/2012
  • Jie Chang, Wen'an Zhou, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: The service continuity mechanism is to provide multimedia application with seamless multimedia experience over multiple networks (Wifi, 3G, Wimax, LTE) as well as multiple devices (mobile, PCs, PDA, Settop box) in ubiquitous network. However, most of the proposed approaches of service continuity are either not adaptive resources provisioning or based on a specific definition of context in ubiquitous network environment. This paper presents a multi-device service continuity with adaptive resources provisioning mechanism under ubiquitous network relying on IMS which helps perform service continuity according to user's contexts (such as preferences and habits), network context, service context and device context. The contribution of this paper is to propose: 1) service continuity scenario, architecture and requirements; 2) signaling flow of resource switching for the proposed scenario is proposed; 3) process of adaptive resources provisioning is put forward. Evaluation results show that the service continuity approach based on context-aware, multi-device coordination and adaptive resources provisioning process outperforms the traditional service continuity.
    01/2012;
  • Guan Le, Ke Xu, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Federated clouds establish a model for independent cloud providers to cooperate and share resource to cope with unforeseen demand loads. One of the challenges in federated clouds is how to find suitable virtual machines to locate resource requests, with constrains of multiple resource attributes. This paper proposes a Gossip-based Hybrid Multi-attribute Overlay (GHMO) for effective resource discovery in federated clouds. GHMO enhances structured overlay with gossip protocol, expanding possible routing ranges to improve the efficiency multi-attribute search. A weight overlay is introduced in the case of routing costs among federated clouds, and an improved neighbor selection strategy is raised to reduce routing costs in process of multi-attribute search. Experimental evaluations show that the performance of the proposed approach is acceptable and stable, and the routing hop and cost is reduced compared with other multi-attribute search methods.
    e-Business Engineering (ICEBE), 2012 IEEE Ninth International Conference on; 01/2012
  • Xianqi Lu, Si Chen, Wen-an Zhou, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Coordinated multipoint (CoMP) transmission/reception is a new transmission scheme in LTE-Advanced system to satisfy the system requirements and improve the cell-edge throughput. It is essential to design a series of feasible and effective signaling processes to implement and deploy CoMP in the real network and achieve the expected performance. In this contribution, we discuss some affecting factors that may lead to variability of the CoMP signaling process design, including the network structure, the scheduling strategy, and the activation and deactivation model. We also propose some common requirements that keep the generality of the CoMP signaling process design, during the sub process of activation, data transmission, and deactivation. In the end, we provide a brief summary of the contribution and an outlook of the future studies.
    Computer Science and Automation Engineering (CSAE), 2012 IEEE International Conference on; 01/2012
  • Wen'an Zhou, Jie Chang, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: The service adaptation mechanism is one of the key issues in ubiquitous network. However, most of the proposed approaches of service adaptation are either not context-aware or based on a specific definition of context in specific environment. This paper presents a context-aware service adaptation mechanism under ubiquitous network relying on user-to-object, space-time interaction patterns which helps perform service adaptation according to user's contexts (such as preferences and habits), network context, service context and device context. The contribution of this paper is to propose: 1) importing user similarity into the service adaptation process and also considering users' trust value; 2) similar users-based service adaptation algorithm (SUSA) is proposed, by combining entropy theory and fuzzy analytic hierarchy process algorithm (FAHP). Evaluation results show that the service adaptation approach based on context-aware, user similarity and SUSA algorithm outperforms the traditional service adaptation algorithm in the accuracy aspect.
    Vehicular Technology Conference (VTC Spring), 2012 IEEE 75th; 01/2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Using virtualization to consolidate servers is a routine method for reducing power consumption in data centers. Current practice, however, assumes homogeneous servers that operate in a homogeneous physical environment. Experimental evidence collected in our mid-size, fully instrumented data center challenges those assumptions, by finding that chassis construction can significantly influence cooling power usage. In particular, the multiple power domains in a single chassis can have different levels of power efficiency, and further, power consumption is affected by the differences in electrical current levels across these two domains. This paper describes experiments designed to validate these facts, followed by a proposed current-aware capacity management system (CACM) that controls resource allocation across power domains by periodically migrating virtual machines among servers. The method not only fully accounts for the influence of current difference between the two domains, but also enforces power caps and safety levels for node temperature levels. Comparisons with industry-standard techniques that are not aware of physical constraints show that current-awareness can improve performance as well as power consumption, with about 16% in energy savings. Such savings indicate the utility of adding physical awareness to the ways in which IT systems are managed.
    Green Computing Conference and Workshops (IGCC), 2011 International; 08/2011
  • Jing Han, Meina Song, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: The traditional relational database to online storage are becoming increasingly problematic: the low performance do not gracefully meet the needs of mass data, the storage approaches of massive data are still not perfect presents. NoSQL and distributed memory database technologies have the potential to simplify or eliminate many of these challenges. NoSQL database technologies can provide Key-value style of data storage and largely ensure high performance. Distributed memory database technologies provide a means for easily store mass data in cloud in a dynamic and scalable manner. This paper argues for a new architecture called CDSA, which is a distributed memory NoSQL database architecture for Cloud computing to improve the performance of querying data and ensure mass data storage in cloud by using rational strategy. Furthermore, add or deleted any node from the distributed database cluster, the others node can work without stop service. We believe that CDSA can provide durable storage with high throughput and lower access latency.
    Computer and Information Science (ICIS), 2011 IEEE/ACIS 10th International Conference on; 06/2011
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we consider a cross-layer design aimed to enhance performance for uplink transmission in an orthogonal frequency division multiple-access (OFDMA)-based cellular network with fixed relay stations. Because mobile stations (MSs) spend most of the power on the uplink transmission, power efficiency resource allocation becomes very important to MSs. We develop a cross-layer optimisation framework for two types of uplink flows (inelastic and elastic flows) that have different quality-of-service requirements. For inelastic flows with fixed-rate requirement, we formulate the cross-layer optimisation problem as the minimisation of the sum transmission power of MSs under the constraints of flow conservation law, subcarrier assignment, relaying path selection and power allocation. For elastic flows with flexible-service-rate requirement, we consider the cross-layer trade-off between uplink service rate and power consumption of MSs and pose the optimisation problem as the maximisation of a linear combination of utility (of service rates) and power consumption (of MSs). Different trade-offs can be achieved by varying the weighting parameters. Dual decomposition and subgradient methods are used to solve the problems optimally with reduced computational complexity. The simulation results show that, through the proposed cross-layer resource optimisation framework and algorithms, significant benefits of deployment of multiple fixed relays in an OFDMA cellular network can be fully obtained such as reduction in power consumption, increase in service rate and energy savings in the uplink transmission of MSs. Copyright © 2011 John Wiley & Sons, Ltd.
    European Transactions on Telecommunications 01/2011; 22:296-314. · 1.05 Impact Factor
  • Yang Meng, Jing Han, Meina Song, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Cloud computing is a style of computing in which dynamically scalable and often virtualized resources are provided as a service over the Internet. So, we can also use the cloud computing to solve the problem of large-scale file storage. Cloud-based file storage system is a high reliability, high consistency and high availability large-scale file storage system. In this paper, we design a service-oriented file storage architecture for cloud computing, for dealing with the problem of the file storage in a carrier-grade information system. Furthermore, we provide solutions to prevent system crash.
    01/2011;
  • Jie Chang, Wenan Zhou, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: With the rapid development of wireless broadband services and the steady advancement of ubiquitous network, there is a rising demand for multi-homed network sites to provide services. Mobile Stream Control Transmission Protocol (mSCTP) is ideal to provide vertical handover (VHO) feature by enabling a mobile client to freely switch between IP addresses acquired in different networks. Concurrent Multipath Transfer with mSCTP's multi-homing support (mSCTP-CMT), which is the extension of mSCTP protocol, simultaneously transmits new data from a source host to a destination host via two or more end-to-end paths. This paper proposes bandwidth estimation extension and congestion control modifications for mSCTP-CMT to improve the transmission throughput in ubiquitous network scenarios. In the proposed scheme, the main contribution is that improved Non-Renegable Selective Acknowledgments (NR-SACKs) based on available bandwidth estimation and its congestion control mechanism which can efficiently utilize of send buffer, especially when frequency of loss events. From the simulation test, we can see that the performance of the proposed scheme could improve the application throughput under packet loss rate and time varied.
    Proceedings of the 74th IEEE Vehicular Technology Conference, VTC Fall 2011, 5-8 September 2011, San Francisco, CA, USA; 01/2011
  • Le Guan, Xu Ke, Meina Song, Junde Song
    [Show abstract] [Hide abstract]
    ABSTRACT: The rapid development of mobile computing and cloud computing trigger novel computing paradigm-----Mobile Cloud Computing. This paper review current research effort towards Mobile Computing. First, we present several challenges for the design of Mobile Cloud Computing service. Second, a concept model has been proposed to analyze related research work. Third, we survey recent Mobile Cloud Computing architecture, application partition & offloading, and context-aware service.
    Computer and Information Science (ICIS), 2011 IEEE/ACIS 10th International Conference on; 01/2011
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we investigate optimal cross-layer design of congestion control, routing, one-hop two-way inter-commodity (OTIC) network coding and scheduling in wireless multi-hop networks utilizing the broadcast advantage of wireless medium. We first present an achievable rate region with OTIC network coding by introducing virtual flow rates in a node. Then we formulate the network utility maximization problem subject to constraints on this achievable rate region, and analyze the complexities of both node- and path-based formulation with no network coding, OTIC network coding, and overhearing network coding. After that, we solve it using dual decomposition and subgradient method. Based on the solution, we present a new queue model, which is able to facilitate the coding operation between two-way commodities, and then propose a backpressure-based cross-layer optimization algorithm with low coding complexity and overhead. Afterwards, we analyze the stability and asymptotical optimality of the proposed cross-layer algorithm by Lyapunov drift technique, and evaluate its performance through extensive simulation. By comparing with the pure routing scheme under both primary and two-hop interference models in an illustrative topology, it is shown that with the proposed joint optimization algorithms, the OTIC network coding can interact adaptively and optimally with other components in different layers and achieve high throughput gain. Simulation of the proposed algorithm in a mesh network with base station and a random ad hoc network justifies that OTIC network coding can obtain considerable throughput gain with low complexity in various kinds of networks.
    Computer Networks. 01/2011; 55:1747-1769.
  • Junjie Tong, Meina Song, Junde Song, Ke Xu
    ICEIS 2011 - Proceedings of the 13th International Conference on Enterprise Information Systems, Volume 1, Beijing, China, 8-11 June, 2011; 01/2011
  • Xiaoxiang Luo, Junde Song
    ICEIS 2011 - Proceedings of the 13th International Conference on Enterprise Information Systems, Volume 1, Beijing, China, 8-11 June, 2011; 01/2011
  • [Show abstract] [Hide abstract]
    ABSTRACT: With the increasing popularity of Internet-based cloud services, energy efficiency in large-scale Internet data centers has become important not only to curtail energy costs and alleviate environmental concern, but also because such systems can quickly reach the limits of power available to them. This paper investigates to what extent and how energy usage improvements through consolidation can benefit from taking into account the environmental influences and effects seen in data center systems. Toward that end, we present experimental results obtained in a fully instrumented, small scale data center and then use these results to propose a hierarchical energy-aware resource scheduler (HEaRS) for cluster workload placement and server provisioning, also considers the physical environment in which data center systems operate. Specifically, at the rack level, HEaRS tries to maintain a 'thermal balance' across the rack to avoid hot spots and reduce cooling costs. At the chassis level, HEaRS utilizes the proportional plus integral controller to achieve a balance in the levels of usage of electrical current between the two power domains in the chassis, which helps the chassis reach its most energy efficient state. Finally, at server level, HEaRS can employ known methods like dynamic voltage and frequency scaling or core idling to reduce power consumption. This results in a hierarchical set of controllers that jointly, implement holistic solutions to energy-aware resource scheduling for an entire rack, and this hierarchical solution can then be further extended to entire data centers. Our initial experiment result show opportunities for gains, with up to 16% in energy usage compared to methods that are not aware of the physical environment and up to 15% improvements in application performance.
    2011 IEEE International Conference on Cluster Computing (CLUSTER), Austin, TX, USA, September 26-30, 2011; 01/2011
  • Xiaoxiang Luo, Junde Song, Meina Song
    [Show abstract] [Hide abstract]
    ABSTRACT: Service Composition provides a flexible framework for new service construction by integrating atomic services developed independently. Algorithms are needed to select atomic service and service instances with various QoS levels according to some application-dependent performance requirements. Our objective of service selection is to maximize a QoS function under the end-to-end QoS constraints, as well as reducing the complexity and computing cost. In this paper the problem is modeled as a multi-dimension multi-choice 0-1 knapsack problem (MMKP). A novel efficient heuristic algorithm for service selection is presented. This algorithm takes the priority of service components and the priority of QoS attributes. These priorities are useful in promoting the efficiency and reducing the complexity of service selection algorithm. The simulation results show that the improvement on efficiency is obvious and increasing with the number of service components.
    Pervasive Computing and Applications (ICPCA), 2011 6th International Conference on; 01/2011
  • Yulu Hao, Meina Song, Jing Han, Junde Song
    ICEIS 2011 - Proceedings of the 13th International Conference on Enterprise Information Systems, Volume 1, Beijing, China, 8-11 June, 2011; 01/2011
  • [Show abstract] [Hide abstract]
    ABSTRACT: Traffic prediction is of significant importance for telecommunication network planning and network optimization. Since modeling and forecasting using traditional Box-Jenkins' ARIMA is rather a complex process and time consuming, a novel approach called APM is studied and applied in this paper. APM is especially appropriate for time series exhibiting stable seasonal pattern and can be employed much simpler than ARIMA. Traffic series from a certain mobile network of Heilongjiang province in China is studied. Average daily traffic per month for the province as well as its every sub-region from July to December in 2009 is forecasted by using APM. The mean absolute percentage error (MAPE) for one-step ahead prediction is 2.11%, and MAPE for the 6 steps is smaller than 7%. The prediction result is of high precision and can be comparable with ARIMA.
    01/2011;