CC BY
14
Liu Jue-fu, Liu Gao-yuan, Li Bo, Shi Yu-zhen УДК 519.87
RESEARCH ON GRID COMPUTING MANAGEMENT MODEL BASED ON MOBILE AGENT
1. Introduction
Gird is the integration of computing resources, including CPU, memory, database etc distributed in different geographical location through high-speed Internet, which provides high-performance computing, resources management and services. Numerous resources are widely distributed in the Grid system nowadays. Those dynamic resources and services may come into or leave from the different Virtual Organization (VO) dynamically at any time. In order to utilize the resources both conveniently and effectively for Grid users, problems related to resources distribution, resources monitor and update and Grid resources management, which includes resources discovery, should be solved.
This paper analyzes and discusses the Mobile Agent technology for grid resource management model.
2. Mobile Agent Technology
Mobile Agent is an executable program which is able to migrate from one computer to another in the network, searching for suitable computing resource and information. Thanks to the mobility and intelligent, Mobile Agent is particularly suitable for Grid with dynamic and heterogeneous features. The introduction of the Mobile Agent technology to grid resource management and scheduling system contributes to the resources release, discovery, distribution and scheduling [1]. There are several following advantages:
(1) Reducing the traffic of the resources control center. A Mobile Agent moves computation to data, directly deals with computation on data side and only returns the final result, which avoids numeric of intermediate data communications and saves the network bandwidth successfully. Mobile Agent also carries many service requests to the server side for local calls, thereby avoids lots of remote calls, and reduces the network delay by each remote call.
(2) Improving the ability to solve tasks in parallel. Mobile Agent is able to asynchronous execute on multiple heterogeneous network hosts and return the
result to user after the process. To complete a specific task, user can creates multiple Agents that execute on one or more computers.
(3) Better adapting to the dynamic of network resources. Mobile Agent supports offline computation, supports distributed applications of the mobile computing. And the dynamic adaptability of the Mobile Agent is able to interact with the environment, perception of environmental change, response fast and independently [2].
3. Grid Computing Model Features
"Grid computing" is a combination of a variety of independent resources and systems connected in network, which enables resources sharing, teamwork and joint computing, and provide the various types of grid-based integrated services for users as a whole. The key feature of the Grid is to share resource and solve problems in the dynamic, multi-organization Virtual Organization environment [3].
Processing information is the core technology of Grid computing. According to the physical features and architecture features of grid computing environment, its application-layer model should have the following features:
(1) Job classification: To effectively cope with the tasks of grid computing environment, reflect the priority of the tasks, the high-grade job should be given priority in Grid Computing model.
(2) High reliability: model has to guarantee the high-reliable communication between various modus and ensure achieving required resources dynamically and effectively.
(3) Interactivity: the intelligent interaction, exchange of information should be performed among different modules in order to know their own resources information and the job handling capacity, and based on those to make the dynamic decisions, modify their own management decisions and coordinate conflicts.
4. The research of Grid Computing Management based on Mobile Agent
ИРКУТСКИЙ ГОСУДАРСТВЕННЫЙ УНИВЕРСИТЕТ ПУТЕЙ СООБЩЕНИЯ
4.1. Construction of Grid Computing Management model
AS a Grid is to realize the sharing of resources, the organization, scheduling of resources is the core of the grid. In s Grid study case, the toolkit Globus Too-kis 2.2 developed by Globus, the organizing, scheduling of resources are realized by two functional modules: MDS, GRAM. MDS makes an efficient aggregation of the scattered grid resources and provides optimal scheduling functions to certain degree. The synchronization between MDS information and resource state is ensured by GRAM; in addition, GRAM can also monitor the resource load state. By utilizing the organization, scheduling information of the resources provided by Grid will greatly facilitate the selection of target node in the process of migration. One important issue in the Network Resource Management model is resource load balancing, which is the inevitable requirement to achieve effective sharing of resources and improve utilization of system resources. Mobile agent is able to achieve load balancing and high fault tolerance, comparing to the load balancing which has the following advantages. Load balancing system allows tasks or processes redirect or relocate in the network, however, this migration is decided by the OS or the related application of load balancing, the re-migrated tasks or process cannot be known at all and are completely passive, that indicates the migration transparency is required by the load balancing system. On the contrary, the move of the Mobile Agent is active, and is the results of the agent show requests, which is decided by the autonomy of Agent. The mobile initiative of Mobile Agent asks for the language realizing Agent with mobile semantics, which requires the codes of Mobile Agent contain functions or statements with function calls. Consequently, the load balancing of the distributed system can be achieved through using of the Mobile Agent technology.
According to the advantages of Mobile Agent, the Grid Computing Management model is build based on the Mobile Agent technology, the model figure is shown in Figure1
The System is divided into 5 layers: Resource layer, Resource Agent layer, Negotiation Layer, Job Agent layer and User layer.
Resource layer contains a variety of heterogeneous resources in the grid system. Resource Agent layer includes two types of Agent, one is Resource Agent that can manage one or more of the same resources, which is responsible for scheduling those resources; the other one is Resource Providing Negotiation layer: the Agent is created by Resource Agent, used to represent the resources provider perform price negotiation [4]. Negotiation layer mainly enables a
negotiation platform and relevant services for the requester and the provider of resources, and contact with the request memory, if there was the same resource request before, then the resource can be requested through directly contact with the resource Agent. Job Agent layer is due to receive is responsible for receiving the task submitted by end-users, and creates the Job Agent, while the Job Agent create the resources requested Negotiation Agent for resources requirement. User layer contains the humans and software applications.
Fig. 1. Grid Computing Management model based on Mobile Agent
4.2. Analysis of system flow
Once the Job Agent received a job request, a job Agent is created. The Job Agent will be responsible for querying Resource Request Negotiation Agents in order to find the available resources, and contacts with the resources, and then form a cooperative alliance through consultation to jointly complete the job. After the job being completed and the results are delivered to end users, the job Agent will be dissolved. The Job Agent has to express its resource requirement, and achieve the system resources through consultation.
5. Simulation and Analysis of Results
We utilized the Swarm Platform of Santa Fe Institute in U.S. to simulate grid service management [5]. In the Mobile Agent simulation model, we utilized the Java programming language call the class library in Swarm to simulation.
The simulation is designed for a grid service. The internal structure and model of the Aggregate
Agent from the micro-economic simulation modelASPEN model developed by U.S. Sandia Labs, is composed by the Resource Agent, the Resource Providing Negotiation Agent, the Job Agent and the Resource Request Negotiation Agent etc. The system has been simulating for more than 10 years, the refresh rate is 1. (Refreshed after each simulation cycle time) When booting the simulation experiment system, each Agent will be in accordance with the pre-set behavior schedule which is the action-interaction observing window and the change of transaction aggregate through computer. The results of experiment are shown in Figure2.
Agent service request to the target host, the Agent can access to the resource of host directly, and reduce the interaction with resource host. Because of the above reasons, which avoid of lots of data transmissions in the Network, reduces the strict requirement to the Network bandwidth, cuts down on the delay time, and improves the service response speed. In addition, each Agent is able to Asynchronous and autonomy execute on multiple heterogeneous network hosts. We can complete the tasks through implanting to the mobile agent.
REFERENCES
vs. time
-¡equirernent
outpu!
|—i-J-¡ ¡ i -
a ion 2BO son
time
Fig. 2. Simulation results of the system
6. Conclusion
Applying the Mobile Agent technology into Grid Computing Resource Management can build a dynamic self-adaptive resource environment. The advantages of Mobile Agent include: the flow of data on the network can be greatly reduced. Through move
1. Frank Griffel, M. Tuan Tu, Malte Munke. Electronic Contract Negotiation as an Application Niche for Mobile Agents // IEEE, 2000.
2. Zhu Lili, Xiong Qianxing. Technology of Mobile Agent in the Application of Electronic Business // Computer & Digital Engineering. 2008. № (4). P. 165-166.
3. Liu Gao-yuan, Liu Jue-fu. Research of Gird Service Based on Web Service // Jounal of East China Jiaotong University. 2008. № (4). P. 71-73.
4. Ma Yin-qiu, Wu Di. Realization of Resource Searching Based on Mobile Agent and AJAX // Microcomputer Information. 2008. № (4). P. 7173.
5. Jain P. Kircher M. Leasing Pattern. Sandholm TW // PLOP 2000 conference. USA. Illionis : Aller-tonPark, 2000. P. 326-328.
Jiang Xiangang, Xu Miaocun, Jiang Xiaojun
Y^K 004.413
SOFTWARE DESIGNING OF GEOMETRY DISTORTION ADJUSTMENT IN MACHINE VISION
1.Introduction
The images are needed for accurate calibration in CNC process based on machine vision dealing with recorded images by one or more quality demotions and identifying the position of certain surface characteristic points, so that the handled images can max-imumly match the original feature and achieve accurate resoration. The digital image processing methods
can transform distorted image into the original image. The transformation process is divided into two steps: (1) geometric transformation. Transformation by using the coordinates relationship between the original image and the distorted image, which means the rule of the distortion makes geometric transformations on the coordinates space where image distortions lie. (2) gray-scale calibration. If the transformed coordinates
