Cloud Computing forAgent-Based UrbanTransportation Systems
Abstract
Agent-based traffic management systems can use the autonomy, mobility, and adaptabilityof mobile agents to deal with dynamic traffic environments. Cloud computing can help such systemscope with the large amounts of storage and computingresources required to use traffic strategyagents and mass transport data effectively. Thisarticle reviews the history of the development oftraffic control and management systems within theevolving computing paradigm and shows the stateof traffic control and management systems basedon mobile multiagent technology.Intelligent transportation clouds could provideservices such as decision support, a standard development environment for traffic managementstrategies, and so on. With mobile agent technology,an urban-traffic management system basedon Agent-Based Distributed and Adaptive Platformsfor Transportation Systems (Adapts) is bothfeasible and effective. However, the large-scaleuse of mobile agents will lead to the emergence ofa complex, powerful organization layer that requiresenormous computing and power resources.To deal with this problem, we propose a prototypeurban-traffic management system using intelligenttraffic clouds.
System Diagram
Proposed System
Agent-based computing and mobile agents were proposed to handle thisvexing problem. Only requiring aruntime environment, mobile agentscan run computations near data to improve performance by reducingcommunication time and costs.This computing paradigm soon drewmuch attention in the transportation field. From multiagent systems andagent structure to ways of negotiatingbetween agents to control agent strategies,all these fields have had varyingdegrees of success.
Cloud computing provides on demandcomputing capacity to individualsand businesses in the form ofheterogeneous and autonomous services.With cloud computing, usersdo not need to understand the detailsof the infrastructure in the “clouds;”they need only know what resourcesthey need and how to obtain appropriateservices, which shields thecomputational complexity of providingthe required services.
Modules
1. Agent-Based Traffic Management Systems
The organization layer consists of amanagement agent (MA), three databases(control strategy, typical trafficscenes, and traffic strategy agent), andan artificial transportation system. Asone traffic strategy has been proposed,the strategy code is saved in the trafficstrategy database. Then, according tothe agent’s prototype, the traffic strategywill be encapsulated into a trafficstrategy agent that is saved in the trafficstrategy agent database. Also, thetraffic strategy agent will be tested bythe typical traffic scenes to review itsperformance. Typical traffic scenes,which are stored in a typical intersectionsdatabase, can determine the performanceof various agents. With thesupport of the three databases, theMA embodies the organization layer’sintelligence.
2. Intelligenttraffic Module
With the development of intelligenttraffic clouds, numerous trafficmanagement systems could connectand share the clouds’ infinite capability,thus saving resources. Moreover,new traffic strategies can be transformedinto mobile agents so suchsystems can continuously improvewith the development of transportationscience.
3. Traffic-strategy agent Module
The more typical traffic scenes used totest a traffic-strategy agent, the moredetailed the learning about the advantagesand disadvantages of differenttraffic strategy agents will be. In thiscase, the initial agent-distributionmap will be more accurate. To achievethis superior performance, however,testing a large amount of typical trafficscenes requires enormous computingresources.Researchers have developed manytraffic strategies based on AI. Someof them such as neural networks consumea lot of computing resources fortraining in order to achieve satisfactoryperformance. However, if a trafficstrategy trains on actuator, the actuator’slimited computing power andinconstant traffic scene will damage the performance of the traffic AIagent. As a result, the whole system’sperformance will deteriorate. If thetraffic AI agent is trained before movingit to the actuator, however, it canbetter serve the traffic managementsystem.
4. Intelligent Traffic Clouds Storage
We propose urban-traffic managementsystems using intelligent trafficclouds to overcome the issues we’vedescribed so far. With the supportof cloud computing technologies,it will go far beyond other multiagenttraffic management systems, addressingissues such as infinite systemscalability, an appropriate agentmanagement scheme, reducing theupfront investment and risk for users,and minimizing the total cost ofownership.
System Requirements:
Hardware Requirement:
Minimum 1.1 GHz PROCESSOR should be on the computer.
128 MB RAM.
20 GB HDD.
1.44 MB FDD.
52x CD-ROM Drive.
MONITORS at 800x600 minimum resolution at 256 colors minimum.
I/O, One or two button mouse and standard 101-key keyboard.
Software Requirement:
Operating System : Windows 95/98/2000/NT4.0.
Technology : JAVA, JFC(Swing),J2EE
Development IDE : Eclipse 3.x