Abstract:在当前绿色云计算的背景下,本专刊将进一步凸显软件在节能减排中的独特而日益重大作用和角色,从效能的角度重新审视软件的发展道路,不仅是自身执行的高效能,而且必须考虑其运行中所依托计算系统的高效能.两者的有效融合,已经对传统的基于开发效率和用户友好的软件开发与设计理念提出了重大的挑战——我们必须在软件的开发和执行两者间寻求新的效能平衡,因为软件是硬件的绿色行为和规格的规范者,在硬件的设计和发展中发挥着日益重大的角色.为在硬件之外需求大规模分布式计算环境下系统效率的提升、能量有效性的提升和系统价格的下降,探索绿色云计算环境下高效能软件及其相关的编程模型与语言等的设计、分析与实现的技术、理论与实例.

Abstract:Internet data centers are developing towards a diversified, intelligent, automated, large-scaled, and standardized direction. With the increasing of scale and complexity, it brings great challenges in how to effectively manage resources. Currently, resource management has become a major issue in Internet data centers, and its importance and urgency cannot be ignored. This paper analyzes two major challenges of resource management with which the Internet data center is facing: (1) meeting the compatibility of concurrent and multiple application SLAs (service level agreements); (2) improving the energy efficiency of system service. Based on the challenges, this paper thoroughly analyzes and summarizes the related work of resource management guaranteeing the SLA, reducing power, and incorporating the objectives of guaranteeing the SLA and reducing power simultaneously during the last ten years. Finally, the paper summaraizes the research and points out future research directions.

Abstract:This paper presents an EE (energy efficiency) model and measuring approach for cloud computing. A mathematical expression of EE is first defined, as well as the measuring and calculation approaches, and the extreme conditions of EE are deduced. Then, to facilitate calculating EE, the mathematical expression between the computer power and the CPU working state is improved, thus EE can be calculated through CPU usage and CPU frequency, which simplifies the measurement of EE. In addition, the study designs and implements a series of experiments to verify the correctness of the proposed model. CPU-intensive, I/O intensive and interactive jobs are performed in both stand-alone environment and cloud environment to evaluate their EEs and to summarize their features as well as the optimization approaches. Both the theory and experiments have proved that the proposed EE model and measuring approach can accurately evaluate the EE of cloud systems and lay the foundation for EE optimization.

Abstract:This paper suggests the green proxy software system to keep wireless devices accessible while saving the energy. It is a system that uses a proxy server to build virtual NICs and virtual images for wireless devices in the same WLAN, so that the devices can sleep during idle time while their functionality are pretended to be carried on by the virtual image in the proxy. If a network request is sent to a device, the proxy server will wake up the device. Using this software system does not require any changes to hardware system or network topology; therefore, it has strong flexibility and universal applicability. The green proxy software system is deployed in a test bench with 11 wireless or wired computers, and real-time measurements of energy consumption are provided to sense and record the energy consumption of the system. The results show significant energy reduce that reaches 60% of the total PCs energy consumption. The economic benefit is obvious if the system is widely deployed.

Abstract:This paper starts at a software architecture level, considers the functional relation between software characteristic quantities and embedded software energy as nonlinear (linear functional relation can be considered as a special nonlinear functional relation). Next, the paper presents an energy model at architecture level by using BP neural network. The energy model measures 5 software characteristic quantities at architecture level and uses BP neural network to fit the functional relation between software characteristic quantities and embedded software energy. Experimental results show that this model is effective.

Abstract:With the aim to address the increasing difficulty of efficiently using large number of cores in many-core processors, a core-partitioned adaptive scheduling algorithm, named CASM (core-partitioned adaptive scheduling for many-core systems), is proposed. CASM dynamically aggregates cores into different partitions by splitting or merging task-clusters, which ensures the efficiency of isolated accessing in these core partitions. To improve the scheduling efficiency of CASM, equi-partitioning scheduling algorithm is adopted to reallocate the cores among task-clusters, and the feedback-driven adaptive scheduling algorithm is implemented within the task-clusters. Online competitive analysis shows that CASM achieves 2-competitiveness ratio with respect to the execution time of parallel jobs, which indicates that CASM has better performance and scalability. The experimental results demonstrate that compared with WS (work-stealing), AGDEQ (adaptive greedy dynamic equi-partitioning) and EQUI?EQUI, CASM reduces the execution time of the same workload by nearly 46%, 32% and 15% respectively. Under the same power consumption, CASM greatly enhances the system throughput.

Abstract:This study extends the current productivity models for a typical Web search engine system, which consists of a Web crawling system and an indexing system. Five different design schemata are characterized according to this model and are compared through power consumption, networking cost, system scale, and query efficiency. The half-WAN scheme, which consists of a WAN-based crawling system and a multi-cluster indexing system, is proved to be the best choice for a large-scale highly-efficient Web search engine.

Abstract:In the running process of cloud computing system, the idle compute nodes will generate a large amount of idle energy consumption. Furthermore, the unmatching task scheduling strategy will also cause a great waste of energy consumption. This paper presents a policy of energy optimal management for cloud computing system based on task scheduling strategy. First, use queueing system to model the cloud computing system for analyzing the mean response time, mean power consumption of cloud computing system, and constructing the energy consumption model of cloud computing system. In order to reduce waste of energy, a high service utilization task scheduling and a low execution energy task scheduling strategy are propsed, which are used to reduce idle energy and “luxury” energy respectively. Based on the idea of the strategies, an algorithm is designed which is called minimum expectation execution energy with performance constraints (ME³PC). Repeated experiments show that this energy management strategy can reduce the energy consumption considerably while meeting performance constraints.

Abstract:This paper analyzes the problem of sensitivity to noise in the mean square criterion of Gaussian- Laplacian regularized (GLR) algorithm. A robust semi-supervised learning algorithm based on maximum correntropy criterion (MCC), called GLR-MCC, is proposed to improve the robustness of GLR along with its convergence analysis. The half quadratic optimization technique is used to simplify the correntropy optimization problem to a standard semi-supervised problem in each iteration. Experimental results on typical machine learning data sets show that the proposed GLR-MCC can effectively improve the robustness of mislabeling noise and occlusion as compared with related semi-supervised learning algorithms.

Abstract:This paper introduces two models to describe dynamic evolution of network information: identify and analysis the document collection on the same topic in different stages. In order to construct dynamic of evolution content differences, two dynamic multi-document summarization models are presented, which are matrix subspace analysis model, text similarity cumulative model. Based on these models, some efficient dynamic sentence weighting algorithms are implemented. Experiments on the test data of Update Summarization in TAC 2008 and comparative results between new models and TAC 2008 evaluation, shows the effectiveness of the models.

Abstract:This paper proposes a method of pedestrian detection that takes both motion information and appearance features into account. This could be done by integrating motion information into the segmentation algorithm in the framework, which performs the validation of segmentation on candidate detection windows obtained by the appearance detector. The paper considers that better segmentation results can raise the detection accuracy. Shape features are obtained by integrating color information indirectly into pedestrian detection by using motion information to model foreground/background distribution of moving object. Better detection performance benefits from the complementary advantages between shape features and pedestrian appearance detector. The claim is supported by these experiments based on CAVIAR and the test video with pedestrians.

Abstract:This paper studies the problem of surveillance area of WVSN (wireless video sensor network) with obstacles, which is finding the optimized view-orientations for all video sensors to maximize multimedia coverage. In this paper, the characteristics of direction adjustable sensing model are discussed and the relation between sensors and obstacles is analyzed. Then, a coverage enhancement algorithm PFOFSA (potential field based occlusion-free surveillance algorithm) is proposed, which could effectively minimize the negative effect of occlusion and overlapping in the sensing field. In PFOFSA, three kinds of virtual centroid point are defined for simplifying the structure of effective region, occlusion region and overlapping region. Two kinds of relationship between virtual force and orientation rotation are constructed and made comparison in experiments. Finally, a set of simulation results are performed to demonstrate the effectiveness of PFOFSA. According to the experiments, PFOFSA could maximize the coverage of WVSN with obstacles and make higher coverage rate than other algorithms.

Abstract:The performance analytical model for (X,S)-bottleneck cell is proposed and perform some probabilistic analysis on the performances of (X,S)-bottleneck cell, such as the probability of balance state, the transmission probability of a flow, and the throughput of (X,S)-bottleneck cell. To capture the essential aspects of (X,S)- bottleneck cell, two-hierarchy Quasi Birth-Death models (QBDs) are used to model the (X,S)-bottleneck cell. The general characters of (X,S)-bottleneck cell are govern by the first hierarchy QBDs, while the characters of each flow are captured by the second hierarchy QBDs. Based on the analytical model, a methodology to derive the balance-state probability of (X,S)-bottleneck cell is proposed, and some theoretic ranges for the transmission probability of a flow and the throughput. The simulations show that the proposed model can analyze effectively the performance of (X,S)-bottleneck cell. The results are helpful for improving and managing wireless networks.

Abstract:Based on the idea of SLoPS, a probing technique, self-loading decreasing rate train (SLDRT), is proposed. SLDRT measures the available bandwidth by using a single decreasing rate packet train. The special train can efficiently avoid bias caused by multiple sampling and makes the tool converge fast. Analysis and simulations verify the effectiveness of the method under both the single-hop and multi-hop path with the bursty cross traffic environment, which also show that SLDRT estimates available bandwidth more, accurately and with less measurement overhead and intrusiveness than other existing techniques such as Pathload, pathChirp.

Abstract:This paper proposes a trajectory privacy measure for Silent Cascade, which is a prevalent trajectory privacy preserving method in LBS (location-based services). In this measure, the user’s trajectory is modeled as a weighted undirected graph, and the user’s trajectory privacy level is calculated through the use of information entropy. It is pointed out in literatures that any privacy preserving methods will be subject to privacy threats once the attacker has new background knowledge. Therefore, adversarial background knowledge is hierarchically integrated into this measure. The privacy metric result composes of the assumptive background knowledge and the corresponding trajectory privacy level. (K_UL(K_i+,K_i-),K_L(K_i+,K_i-)) association rules is also proposed to describe the assumptive background knowledge. Simulation results show that this metric is an effective and valuable tool for mobile users and the designers of trajectory privacy preserving methods to measure the user’s trajectory privacy level correctly, even the attacker has variable background knowledge.

Abstract:Network anomaly detection is very important in order to guarantee the reliable operation of network. Existing methods only utilize temporal correlation or spatial correlation of network traffic individually. Aiming at this deficiency, this paper considers the spatio-temporal correlation of traffic matrix together and puts forward a network-wide anomaly detection method based on MSPCA. The method utilizes the multiscale modeling ability of wavelet transform and dimensionality reduction ability comprehensively to model normal network traffic, and then analyzes residual traffic using Shewart and EWMA control charts. In addition, the MSPCA anomaly detection method is extended to online MSPCA anomaly detection method through applying gliding window mechanism. Real Internet measurement data analyses and simulation experiment analyses show that the detection performance of MSPCA algorithm is superior to PCA algorithm and KLE algorithm proposed recently. Analyses also show that the detection performance of online MSPCA algorithm is close to MSPCA algorithm, and the single step execution time of online MSPCA algorithm is very short, which can fully meet the need of real-time detection.

Abstract:This paper proposes a semantic-based approach to malware behavioral signature extraction and detection. This approach extracts critical malware behaviors as well as dependencies among these behaviors, integrating instruction-level taint analysis and behavior-level semantics analysis. Then, it acquires anti-interference malware behavior signatures using anti-obfuscation engine to identify semantic irrelevance and semantically equivalence. Further, a prototype system based on this signature extraction and detection approach is developed and evaluated by multiple malware samples. Experimental results have demonstrated that the malware signatures extracted show good ability to anti obfuscation and the detection based on theses signatures could recognize malware variants effectively.

Abstract:This paper studies generalized signcryptions in the certificateless public key settings. The paper proposes the formal definition and security model of certificateless generalized signcryption. The Type II adversary in the security model is malicious, but a passive Type II attacker. Also an efficient construction of certificateless generalized signcryption scheme using bilinear maps is also implemented. The security of proposed scheme is based on the infeasibility of the Computational Diffie-Hellman problem and Decision Bilinear Diffie-Hellman problem. The scheme is formally proven without random oracles under the security model of certificateless generalized signcryption schemes. Due to its security, high efficiency and freedom from certificate management, it may have practical applications in electronic cash payment, firewall and key distribution, etc.

Abstract:This paper proposes a fault tolerance approach for OpenMP programs, named PR-OMP, which makes use of a novel fault recovery scheme, parallel recomputing. By redistributing the workload of the failed thread to all the surviving threads, PR-OMP remarkably reduces the overhead for fault recovery. The paper discusses the key issues including program division, computational state saving, workload redistribution, and fault detection of PR-OMP and details concerning implementation. Furthermore, the paper also presents an extended data flow analysis for OpenMP, which is used to decrease the data amount of computational state saving. Through the experimental evaluation, it has been proven that this approach achieves a minor overhead in fault recovery.

Abstract:Considering the characteristics of the grid computing environment, dynamic, heterogeneous and distributional, and the problem of the low utilization ratio of resources and benefit imbalance in the grid resource distribution, this paper proposes a grid resource auction model which is multi-winners and based on the microeconomics theory. The contributions of this paper are listed as follows: first the study predicts the status of consumer’s bidding price using the hidden Markov model; second, the paper presents the multi-winners auction model using Nash equilibrium of the incomplete information game, where it could enhance the utilizable rate of resources; thirdly, the condition of dominant strategy incentive compatibility is analyzed; finally, the paper proves the profits both of buyers and the seller all are maximal. Moreover, the utilizable ratio of the resource is proved to be increased through the contradistinctive experiment with other algorithms.