Recommendation technology is an important part of the Internet Things (IoT) services, which can provide better service for users and help get information anytime, anywhere. However, traditional recommendation algorithms cannot meet user's fast accurate recommended requirements in IoT environment. In face a large-volume data, method finding neighborhood by comparing whole user will result low efficiency. addition, system ignores inherent connection between preference time. reality, interest...

MapReduce is becoming the state-of-the-art computing paradigm for processing large-scale datasets on a large cluster with tens or thousands of nodes. It has been widely used in various fields such as e-commerce, Web search, social networks, and scientific computation. Understanding characteristics workloads key to achieving better configuration decisions improving system throughput. However, workload characterization MapReduce, especially production environment, not well studied yet. To gain...

Understanding the characteristics of MapReduce workloads in a Hadoop cluster is key to making optimal configuration decisions and improving system efficiency throughput. However, workload analysis on cluster, particularly large-scale e-commerce production environment, has not been well studied yet. In this paper, we performed comprehensive using trace collected from 2000-node at Taobao, which biggest online enterprise Asia, ranked 10th world as reported by Alexa. The results are...

In cloud computing communications, information security entails the protection of elements (e.g., multimedia data), only authorized users are allowed to access available contents. Fingerprint recognition is one popular and effective approaches for priori authorizing protecting during communications. However, traditional fingerprint have demerits easy losing rich poor performances due complex inputs, such as image rotation, incomplete input image, quality enrollment, so on. order overcome...

Security is paramount in public places such as subways, airports, and train stations, where security screeners use X-ray imaging technology to check passengers' luggage for potential threats. To streamline this process make it more efficient, researchers have turned object detection techniques with the help of deep learning. While some progress has been made, there are few comprehensive literature reviews. This paper provides a overview standard algorithms principles dangerous goods...

Resource management is one of the main issues in cloud computing. order to improve resource utilization large data centers while delivering services with higher QoS clients, a allocation strategy based on market (RAS-M) proposed. Firstly, architecture and model RAS-M are constructed, which QoS-refection utility function designed according different requirements client, equilibrium state defined proof its optimality given. Secondly, GA-based price adjusted algorithm introduced deal problem...

Previous packet length optimizations for sensor networks often employ a fixed optimal scheme, while in this study we present DPLC, Dynamic Packet Length Control scheme. To make DPLC more efficient terms of channel utilization, incorporate lightweight and accurate link estimation method. We further provide two easy-to-use services, i.e., small message aggregation large fragmentation, to facilitate upper-layer application programming. The implementation based on TinyOS 2.1 is lightweight, with...

Sensing coverage is a fundamental problem in wireless sensor networks (WSNs), which has attracted considerable attention. Conventional research on this topic focuses the 0/1 model, only coarse approximation to practical sensing model. In paper, we study target problem, where objective find least number of nodes randomly-deployed WSNs based probabilistic We analyze joint detection probability with multiple sensors. Based theoretical analysis probability, formulate minimum ϵ-detection problem....

Tensor data (i.e., the having multiple dimensions) are quickly growing in scale many practical applications, which poses new challenges for modeling and analysis approaches, such as high-order relations of large complexity, gross noise, varying scale. Existing low-rank methods, effective at analyzing matrix data, may fail regime tensor due to these challenges. A robust scalable method is heavily desired. In this paper, we develop an online (ORLTM) address The ORLTM leverages correlations...

This paper investigates the use of wireless sensor networks for multiple event source localization using binary information from nodes. The events could continually emit signals whose strength is attenuated inversely proportional to distance source. In this context, faults occur due various reasons and are manifested when a node reports wrong decision. order reduce impact on accuracy localization, we introduce trust index model evaluate fidelity which nodes report in detection process,...

In this paper we present results from test and quantitative analysis of leading hypervisors Xen, Kvm Vmware. Our comparison was carried out using benchmarks which can performance on the overall performance, isolation, scalability virtual machines running these three hypervisors. To compare, need to measure efficiency virtualization technologies. We define equal scalability. These make follow SPEC TPC standard easily repeatable .Our goals are understand how different architectural decisions...

Wireless charging is an important issue in wireless sensor networks, since it can provide emerging and effective solution the absence of other power supplies. The state-of-the-art methods employ a mobile car predefined moving path to charge nodes network. Previous studies only consider factor network (i.e., residual energy node) as constraint design strategy. However, factors, such travelled distance car, also affect effectiveness In this work, we study strategy based on analysis combination...

In wireless sensor networks (WSN), faults occurring to nodes are common due the device itself and harsh deployment environment. order avoid degradation of service faults, it is necessary for WSN be able detect early. this paper we propose a localized fault diagnosis algorithm which executes in tree-like effectively. It based on local comparisons sensed data dissemination test results remaining sensors. Furthermore, Times redundancy used diagnose intermittent sensing communication. Simulation...

Abstract Network passive monitoring is an important method in debugging and analyzing wireless sensor networks. The for networks (WSN) has the following problems: traces are incomplete contain only limited information because some packets may be misheard; how to infer implicit about network protocol behavior detect events from challenging issues system. To resolve above issues, we propose use of time tuning trace merging a finite state machine inferring missing packets, event detection. (1)...

Resource management is one of the main issues in Cloud Computing. In order to improve resource utilization large Data Centers while delivering services with higher QoS Clients, an automatic allocation strategy based on market Mechanism ( A RAS-M) proposed. Firstly, architecture and model RAS-M are constructed, which a QoS-refectitive utility function designed according different requirements Client . T he equilibrium state defined proof its optimality given. Secondly, Genetic Algorithm (GA)...