Keyuan Jiang

An energy-optimization-based method of task scheduling for a cloud video surveillance center

The number of cloud video surveillance (CVS) systems has been increasing rapidly over the last decade. Since CVS systems are big energy consumers, it is urgent to take the problem of optimizing the energy consumption of CVS systems into consideration. In this study, we build a task scheduling model, and present a method of scheduling that minimizes energy consumption by reducing the number of virtual machines. The optimization problem is first formulated as a multi-dimensional bin-packing problem due to the constrains on the resources (sizes of the bandwidth, the memory, the hard disk, the CPU utilization, etc.). We convert the problem into a one-dimensional bin-packing problem by making use of the relationships between the resources, and solve it using the greedy best-fit search algorithm. This method greatly reduces the computational expense and can be used in a real-time fashion. An experimental system is designed to evaluate the method, and four experiments are carried out to demonstrate the validity of the method. Experimental results show that the method not only largely improved the resource utilization and reduces energy consumption but also the scheduling time was significantly decreased when handling the same number of video tasks. And it is obviously superior to the common approach and First Fit Decreasing (FFD) algorithm. HighlightsWe build a model of power consumption of the cloud video surveillance center.We build a model of video task scheduling of the cloud video surveillance center.We formulate the model of task scheduling into multi-dimensional bin-packing problem.The multi-dimensional problem is converted into a one-dimensional bin-packing problem.A greedy best-fit algorithm is presented with better effective and lower expense.

Construction of a Personal Experience Tweet Corpus for Health Surveillance

Studies have shown that Twitter can be used for health surveillance, and personal experience tweets (PETs) are an important source of information for health surveillance. To mine Twitter data requires a relatively balanced corpus and it is challenging to construct such a corpus due to the labor-intensive annotation tasks of large data sets. We developed a bootstrap method of finding PETs with the use of the machine learning-based filter. Through a few iterations, our approach can efficiently improve the balance of two class dataset with a reduced amount of annotation work. To demonstrate the usefulness of our method, a PET corpus related to effects caused by 4 dietary supplements was constructed. In 3 iterations, a corpus of 8,770 tweets was obtained from 108,528 tweets collected, and the imbalance of two classes was significantly reduced from 1:31 to 1:3. In addition, two out of three classifiers used showed improved performance over iterations. It is conceivable that our approach can be applied to various other health surveillance studies that use machine learning-based classifications of imbalanced Twitter data.

A Low Energy Consumption Storage Method for Cloud Video Surveillance Data Based on SLA Classification

With the continuous expansion of the amount of data with time in mobile video applications such as cloud video surveillance (CVS), the increasing energy consumption in video data centers has drawn widespread attention for the past several years. Addressing the issue of reducing energy consumption, we propose a low energy consumption storage method specially designed for CVS systems based onthe service level agreement (SLA) classification. A novel SLA with an extra parameter of access time period is proposed and then utilized as a criterion for dividing virtual machines (VMs) and data storage nodes into different classifications. Tasks can be scheduled in real time for running on the homologous VMs and data storage nodes according to their access time periods. Any nodes whose access time periods do not encompass the current time will be placed into the energy saving state while others are in normal state with the capability of undertaking tasks. As a result, overall electric energy consumption in data centers is reduced while the SLA is fulfilled. To evaluate the performance, we compare the method with two related approaches using the Hadoop Distributed File System (HDFS). The results show the superiority and effectiveness of our method.

A Task Scheduling Method for Energy-Efficient Cloud Video Surveillance System Using a Time-Clustering-Based Genetic Algorithm

Demands for cloud video surveillance systems are growing rapidly. Addressing to the issue of low energy-efficiency in cloud video datacenters, a task scheduling method using a time-clustering-based genetic algorithm is proposed. Firstly, an off-line scheduling model with SLA (service level agreement) time constraint is proposed after the analysis of the constrain relationship between the SLA and surveillance tasks. Then, a time-clustering-based genetic algorithm (TCGA) is proposed to solve the model for an optimal energy-efficient solution. According to the solution, the service quality is guaranteed and the total operating time of virtual machines is minimized. Meanwhile, idle virtual machines are shut down to reduce energy consumption. Simulations of large scale tasks scheduling are conducted. Several comparison experiments verify that the proposed method can improve the resource utilization greatly and achieve energy saving extremely.

An Energy-Aware Workload Balancing Method for Cloud Video Data Storage Management

The data centers of cloud video surveillance (CVS) systems based on Hadoop have a couple of common problems such as large energy consumption, low power utilization, etc. Addressing to the issue of reducing energy consumption while guaranteeing quality of service, we propose an energy-aware workload balancing method for efficient data storage management in cloud video data centers. A dynamic adjusting algorithm is designed to control the running status of nodes in the data centers according to the access frequency of video data blocks for the purpose of reducing the energy consumption, in order to eliminate the potential influence on the service quality posed by the changes of running status of nodes, the workload balancing between nodes in ring network topology is executed, a nonlinear programming model is established to obtain the minimum number of data blocks transferred during the workload balancing. Experimental results in the GridSim simulation environment show that the proposed method can achieve more energy saving and better performance than the original Hadoop.

Prototyping A Workstation-based Icu Charting System

ARSTRACT We describe a prototype ICU charting system based on workstation technology. The system is designed to provide a natural, paper-style user interface. Significant graphical capabilities are included. The prototypc is implemented using Microsoft Excel running under MS Windows. We discuss the issues and methods involved in the creation of the prototype. INTRODUCTION AND RACKGROUND The explosion of data acquired through intcnsive care monitoring of critically ill patients has posed a major problem for physicians and nurses for the past 10-15 years. Several serious attcmpts have bcen madc, both commercially and by individual medical centers to develop systems capable of handling these data in a manner not intimidating to non- computer competent mcdical personnel. These attempts have gcnerally been less than fully successful because sufficient skills wcrc not ablc to be maintained across a wide range of personncl. Further, most commcrcially available systems are extremely cumbersome to use (command driven or multiple level menu systems), are relatively slow due to hardware constraints, and support only relatively low resolution graphics output with small viewing areas. However, new computing technology, namcly, high performance workstations with high resolution bit-mapped displays promise to provide a development and user environmcnt capablc of addressing the problems described above. Only rcccntly havc conimcrcial products been announced that begin to address the ICU data problem utilizing workstation technology. Our goal is to prototype an ICU workstation meeting local design requirements. Thcse dcsign requirements include automation of the data collection process, improving tk quality of charted data, increasing the effcctiveness of interpretation of charted data through new visualization tcchniques, crcating a ncw research tool through the creation of research data bases, decreasing the workload of nursing staff, creating a more uniform charting cnvironment, and creating a charting environment that is natural to use. The system is developed around the concept of rcplicating the functions of the ICU chart, a large, two dimensional spreadsheet, whcrc rows reprcscnt itcms and columns represent timc. Typically, the flow sheet represents one full day of patient data. On this flowshcct the nurse records all vital information related to the patient including viral signs, medication, fluid balance, lab results, interventions, etc.

An energy-aware task consolidation algorithm for cloud computing data centre

Energy efficiency has become an increasingly prominent issue in cloud computing. Task consolidation is an effective way to maximise utilisation of cloud computing resources and reduce energy consumption. Presented in this paper is an improved energy-aware task consolidation algorithm to optimise the scheduling of tasks in cloud computing data centres. Our algorithm was developed based on the linear relationship between energy consumption and the CPU utilisation. Instead of consolidating tasks to a service node until its CPU utilisation reaches 100%, our algorithm uses an optimal point, where the CPU utilisation of a service node reaches 70%, and ensures that every task is preferentially assigned to service nodes that satisfy the 70% CPU utilisation. Our experiments demonstrate that our algorithm can significantly reduce the energy consumption of cloud computing data centres without performance degradation while maintaining good performance compared to the energy-conscious task consolidation (ECTC) approach.

A Johnson's-Rule-Based Genetic Algorithm for Two-Stage-Task Scheduling Problem in Data-Centers of Cloud Computing

One of the keys to making cloud data-centers (CDCs) proliferate impressively is the implementation of efficient task scheduling. Since all the resources of CDCs, even including operating systems (OSes) and application programs, can be stored and managed on remote data-centers, this study first analyzed the task scheduling problem for CDCs and established a mathematical model of the scheduling of two-stage tasks. The Johnsons rule was combined with the genetic algorithm to create a Johnsons-rule-based genetic algorithm (JRGA), which takes into account the characteristics of multiprocessor scheduling in CDCs. New crossover and mutation operations were devised to make the algorithm converge more quickly. In the decoding process, the Johnsons rule is used to optimize the makespan for each machine. Simulations were used to compare the performance of the JRGA with that of the list scheduling algorithm and an improved list scheduling algorithm. The results demonstrate the validity of the JRGA.

A Streaming Execution Method for Multi-services in Mobile Cloud Computing

Commercially available mobile devices with various kinds of hardware and software platforms have resulted in a huge amount of mobile services. This has a new challenge in designing multiple services that are compatible with heterogeneous devices and operating systems (OSes). This paper presents a method of multi-services mobile cloud computing (MSMCC) for mobile device, which regards OS as a service to improve user experience. A multi-services pre-boot firmware (MSPF) is designed to stream multi-OS images and multi-application images to the mobile devices over wireless network. After that, we use the MSMCC method to run on the S3C6410 board equipped with MSPF, results illustrated that MSPF is able to support remote boot and streaming execution of multi-services with stable performance.