High utility pattern mining has been widely applied in the field of data mining.Some top-k high utility pattern mining algorithms based on tree-like and list-like structures were proposed.However,tree-like algorithms generated a large number of candidates,and comparing operation was costly during the process of utility pattern growth in list-like algorithms.In addition,the amount of information data increased exponentially in information society.Thus,it required memory usage and computational cost in mining process,especially the dataset size was huge.In order to address above issues,this paper proposed top-k high utility pattern mining algorithm based on MapReduce,called TKHUP_MaR.TKHUP_MaR needed to scan database twice and used three MapReduce phases to parallelize top-k high utility pattern mining.The experiment results show that TKHUP_MaR is effective in the process of mining top-k high utility patterns on parallel environment.

Setting an appropriate threshold was essential to high utility pattern mining,however it was a difficult task for users.If the threshold was too low,a large number of low utility patterns were generated;if the threshold was too high,no high utility patterns might be generated.According,Top-k high utility pattern mining was proposed,in which k was the number of highest utility patterns.Most studies on high utility mining were only designed for static database,yet in real-world applications some new transactions inserted in original database.To address the above issues,this paper proposed an incremental Top-k high utility pattern mining algorithm,named TOPK-HUP-INS.The algorithm adopted four effective strategies to mine the high utility patterns that the number was user-specified in the case of incremental data.Comparing the experimental results on different datasets show that TOPK-HUP-INS performs well in terms of time and space.

High utility pattern mining is a fundamental research in data mining,in which more and more algorithms about top-k high utility pattern mining algorithms are proposed,where k refers to the number of high utility patterns that users need to mine.It can be classified into two types:two-phase algorithm and single-phase algorithm.The former generated a huge number of candidates in mining process,which was the primary factor to decreasing the performance of algorithm;the latter mined top-k high utility patterns without candidate generation.To mine the k of the most valuable patterns more efficiently,this paper proposed a single-phase algorithm TKHUP.The proposed algorithm used four effective strategies to save time and space consumption during mining process.A large number of experiments indicates that the performance of TKHUP is the state-of-the-art topk high utility mining algorithm on time.

An inexpensive ,efficient and intelligent system for checking attendance and managing information based on Bluetooth communication is designed and realized .Bluetooth short-range transmission and poor through-wall ability charac-teristics are used to achieve the distance control in classroom .Combined with Wi-Fi and Android technology ,the information management and real-time display for roll-call information are achieved .Experimental result shows that this system's credi-bility and speed have obvious advantages over manual attendance check ,which is an important component in intelligent cam-pus management .

论文设计并实现了一种基于蓝牙通信的低成本高效智能学生考勤和信息管理系统。充分利用蓝牙短距离传输和穿墙能力差的特性,有效实现了对教室内通讯信号的短距离控制,并且结合现有校园Wi-Fi覆盖条件和智能化手机等优势技术实现了实时、快速、高效的考勤信息采集和考勤信息管理与实时展示。实验结果表明,该系统较人工考勤方式在可信度和速度上具有明显优势,也可以成为目前高校智慧校园管理中的重要组成部分。

Maintaining the energy of sensors in Wireless Sensor Network (WSN) is important in critical applications. It has been a challenge to design wireless sensor networks to enable applications for oceanographic data collection, pollution monitoring, offshore exploration, disaster prevention, assisted navigation and tactical surveillance applications. WSN consists of sensor nodes which sense the physical parameters such as temperature, humidity, pressure and light etc and send them to a fusion center namely Base Station (BS) from where one can get the value of physical parameters at any time. Requirement of monitoring the environment might be anywhere, like middle of the sea or under the earth where man cannot go often to recharge the batteries which supplies the sensing device, transceiver and memory unit in the sensor node. So the usage of the battery power must be judicious in WSN. Earlier attempts have been made to prolong the network lifetime, but still it is a challenging task. In this paper we propose a Novel Energy efficient Surface water Wireless Sensor Network Algorithm (NES-WSN) to optimize the energy consumption by WSN. The present work concentrates on energy saving of sensor nodes when they are deployed in the surface of the sea water. Whenever the sea surface temperature increases there will be a power loss which is reduced by clustering the nodes and by transferring data through multihop routing. Experimental results show that due to increase in temperature there is a definite power loss and it can be minimized by using NES-WSN algorithm definitely.

A wireless sensor network (WSN) consists of hundreds or thousands of energy-limited sensors. Distributing energy dissipation evenly throughout the sensors and saving energy are important goals in designing of WSN protocol. In this paper, we propose an improved energy-efficient algorithm based on L-DCHS (IEAL), it adds a cluster head in each cluster with a novel method, and prolong the steady-state phase time. Additionally, we redefine the candidate set in each round. Simulations show that IEAL can achieve a longer system lifetime and save energy dissipation.

The energy efficiency in the WSN is one of the very important Performance Indicators. This paper proposed a new algorithm—PDCH, on the bases of PEGASIS to make every notes load balance and extent the network lifetime. Protocol PEGASIS is based on the chain structure, every chain have only one cluster head, it is in charge with every note's receiving and sending messages who belong to this chain, the cluster head consumes large energy and the times of every round increasing. In this situation, We chose double cluster heads in one chain and used hierarchical structure in the new algorithm on the basis of avoiding the long chain existing. According to the simulation result, we could know that the new algorithm increased the efficiency of energy-using and the load balance, and extend the lifetime of the whole network.

Most of sensors in a wireless sensor network (WSN) are energy-limited. In order to prolong the network lifetime, some energy-efficient algorithms are proposed. Low Energy Adaptive Clustering Hierarchy with Deterministic Cluster-Head Selection (L-DCHS) is a well-known distributed clustering algorithm. In this paper, we propose a novel energy-efficient algorithm for wireless sensor networks (NEAW) based on L-DCHS. Our simulation shows that NEAW can prolong the network lifetime, and it is energy-efficient.

FAST TCP是一种针对当前高性能网络的高速、长延迟等特点的TCP拥塞算法.论文首先对传统的TCP Reno进行介绍,讨论其算法与FAST TCP的不同点.然后重点分析FAST TCP体系结构和算法.利用NS2仿真工具测试了传统TCP Reno与FAST TCP在吞吐量方面的性能差异.实验验证了FAST TCP较传统的TCP Reno在吞吐量方面具有很大的性能优势.

In this paper, a new method of touring into the picture (TIP) is proposed to present one of the most famous ancient Chinese paintings in order to renew the essence of traditional art works. In the beginning, framework of the whole presentation system is detailed to introduce our system's working principle. Then rendering path which is traversing the exquisite artwork is set for camera capturing key scenes. Then 3d-depth information and 3d scene structure is built with Markov random fields (MRF). In order to make better visual effect, we give an improved method to the projection system. With parameters is set for our camera, it become possible to make sequential dynamically video through dynamical computation.

Traditional collaboration systems is not reliable or scalable. In addition, this architecture can not make good use of computing resources and network bandwidth on the client side. Peer-to-peer architecture is a new computing model. It is highly scalable and more suitable for collaboration. The advantages of using P2P architecture in collaboration systems are analyzed, the details of the design and implementation of a P2P NetWork for collaboration system.

A ploy to ensure the security distribution of the key and the identity authentication of the peers in P2P network is provided in this paper. The proposed security and authentication protocol made use to the combination of public key and secret key to provide such ploy. We also applied to the authentication test method to prove the correctness of the protocol.

This paper introduces application of JXTA in P2P network and analysis JXTA technology advantage and disadvantage in distributed parallel computation. Parallel computation task scheduling in JXTA network environment have not realize load balance, so we design a solution which depend on load prediction to solve parallel computation task scheduling in JXTA. Finally, we develop a parallelization searching prime application program in peer to peer JXTA network environment in order to test this solution's performance.

本文首先分析了SAML的结构、工作原理和特点,并针对它的具体应用,提出了基于S AML的Web Service的安全模型,最后作者结合实际应用对此模型进行了分析.

基于角色的访问控制提供了灵活安全管理授权机制,公钥基础设施(PKI)提供了一个强有力的认证系统,授权管理基础设施(PMI)作为一项新的技术提供了有效授权和访问控制管理.为了满足现在的安全需求,结合X.509公钥证书(PKCs)和属性证书(ACs),本文提出了一个基于角色的访问控制模型.