We propose a cloud service composition framework that selects the optimal based on an end user's long-term Quality of Service (QoS) requirements. In typical environment, existing solutions are not suitable when providers fail to provide QoS provision advertisements. The proposed uses new multivariate analysis predict provisions from providers' historical data and short-term advertisements represented using Time Series. quality prediction is improved by incorporating attributes' intra...

We propose a theoretical model based on the concept of multiferroic tunnel junction. The is capable producing eight different logic states by combining spin-filter effect and screening polarization charges between two electrodes through general spintronic tunneling. dependence conductance ratio with very large magnitude electric polarization, exchange splitting, barrier width, bias voltage investigated. result may provide some insights into realization octal data storage (namely, are used as...

In autonomous driving community, numerous benchmarks have been established to assist the tasks of 3D/2D object detection, stereo vision, semantic/instance segmentation. However, more meaningful dynamic evolution surrounding objects ego-vehicle is rarely exploited, and lacks a large-scale dataset platform. To address this, we introduce BLVD, 5D semantics benchmark which does not concentrate on static detection or segmentation tackled adequately before. Instead, BLVD aims provide platform for...

We propose a new framework for composing Sensor-Cloud services based on dynamic features such as spatio-temporal aspects. To evaluate services, two quality attributes are introduced. present heuristic algorithm A* to compose in terms of In addition, technique 3D R-tree access is proposed. Analytical and simulation results presented show the performance proposed approach.

Traditional hierarchical intrusion detection systems have a central manager which attracts hackers to attack and might overload when there are too many client requests. To overcome these drawbacks, some researchers suggested apply Peer-to-Peer approaches in detection. Most current only allow hosts collect related information from "neighbours" (one hop hosts). The limitation of sources may lead system make inaccurate decisions. In this paper, we propose Mobile Agent Based Distributed...

This article considers cloud service composition from a decision analysis perspective. Traditional QoS-aware techniques usually consider the qualities available at time of because compositions are immediately consumed. is fundamentally different in environment where typically lasts for relatively long period time. The two most important drivers when composing long-term nature and economic motivation outsourcing tasks to cloud. We propose an model, which we represent as Bayesian network,...

To improve data availability and reduce user access latency, geo-cloud based replication is widely used in large global Web sites, such as Facebook. However, the popularity of different will change time goes by, simple static replica creation strategies that assign same number replicas to all data, nev er changing thereafter, are not suitable. this issue, we propose a two-layer dynamic strategy called TGstag. TGstag addresses issue with twofold: policy constraint heuristic inter-datacenter...

Cross datacenter data replication has been widely used in geo-cloud environment due to its ability increase application's availability and improve the performance. However, with large scale of cloud, it is difficult determine location replicas among datacenters order minimize overall user access latency. The correlation between each other makes replica placement problem more complex. To address these issues, we propose a two-step approach called GCplace. Before applying GCplace, network...

many distributed storage systems have been proposed to provide high scalability and availability for modern web applications. However, most of those applications only aware data skew while actually request is also widely exist needed be considered as well. In this paper, we present a heterogeneous system based on Cassandra-a famous NoSQL database aiming manage very large scale without single point failure. We improve Cassandra through two ways: 1) minimize forward load by shifting the node...

QoS-aware service composition intends to integrate services from different providers and maximize the global QoS in order increase user's satisfaction degree while subjecting dynamic context constraints. Current approaches only focus on optimizing a single process for one party. When multiple processes are performed concurrently by their selfish users resource-constrained environment, new issues will arise, i.e., undesirable competition resources, extra waiting frequent change of contexts....

Computer systems are becoming extremely complex, while system anomalies dramatically influence the availability and usability of systems. Online anomaly prediction is an important approach to manage imminent anomalies, high accuracy relies on precise monitoring data. However, data not easily achievable because widespread noise. In this paper, we present a method which integrates improved Evidential Markov model ensemble classification predict for with Traditional models use explicit state...

Accurate carbon (C) stock estimation is crucial for C sequestration research, environmental protection, and policy formulation related to management. Although, research on in forests, oceans, soil, desert has received increasing attention, relatively few studies have focused urban stock. Moreover, the current mainstream methods assessment, including field surveys satellite mapping, are characterized by notable limitations, being labor-intensive having limited real-time data acquisition...

Extraction of the breast skin-line is crucial in computeraided analysis mammograms. This paper presents an effect adaptive-neighborhood contrast enhancement (ANCE) [1] on extraction. ANCE used to enhance parenchyma and suppress background noise. Suppression noise can improve Our extraction method based work by Ojala et al. [2]. We use Hausdorff distance [3, 4] for quantitative comparison skin-lines. shows that improves due its ability suppressing while improving contrast. have defined...

To defend against distributed denial of service (DDoS) attacks, one critical issue is to effectively isolate the attack traffic from normal ones. A novel DDoS defense scheme based on TCP_IP Header Analysis and Proactive Tests (THAPT) hereby proposed. Unlike most previous schemes that are passive in nature, proposal uses proactive tests identify malicious traffic. Simulation results validate effectiveness our proposed scheme.

The key management is the most important way to protect communication in Wireless Sensor Networks (WSNs). In this paper, we present a Distance-based Key Management (DKM) which can efficiently enhance security and survivability hierarchical WSNs. Different from previous works, generate system as well cluster formation. And DKM distributes keys based on distance (hop counts), not only localizes things but also has no overhead. network compromised nodes by reducing high probability of common...

Using deep learning technology to detect individual bamboo trees in UAV images is a valuable research field. Similar other object detection tasks, scenario, data augmentation can greatly increase the number and appearance diversity of samples, making trained model has good generalization ability performs better when predicting trees. Since different operations have effects on improving performance model, just treating them equally randomly selecting combining some training will not achieve...