Umar Siddiqui, Ghalib Ahmed Tahir Attiq-ur-Rehman students of NUST SEECS stood second in COMPPEC 2012 held in Department of Computer Engineering, College of Electrical Engineering & Mechanical Engineering (EME) on 27-28th June 2012
Umar Siddiqui, Ghalib Ahmed Tahir Attiq-ur-Rehman students of NUST SEECS stood second in COMPPEC 2012 held in Department of Computer Engineering, College of Electrical Engineering & Mechanical Engineering (EME) on 27-28th June. It was the 11th Annual All Pakistan Inter Colleges/Universities Computer Projects Exhibition and Competition, one of the most prestigious national competitions in Pakistan since 2002.
The students had to compete in four different categories namely Electro Mechanical, Information Systems, Information Systems and Digital systems. The projects were adjudged by different universities and industry representatives. NUST SEECS students developed a Dynamically Scalable Multi-agent Environments using Cloud Resources under the supervision of Dr. Peter Bloodsworth and Dr. Raihan Ur Rasool, which got awarded second position at COMPPEC.Project Abstract
In recent years multi-agent systems have been applied in important problem areas such as anti-terror applications and logistics. Agents work on the behalf of users to provide decision support in challenging environments. As multi-agent technology continues to develop, the numbers of agents that comprise a system and their usage of computing resources are both likely to grow. Cloud infrastructures including Amazon EC2 can potentially be harnessed in order to dynamically scale a multi-agent platform such as JADE (Java Agent Development Environment.) This project involves the development of automated mechanisms for managing periods of heavy load in multi-agent platforms using cloud resources. In order to achieve this, a load monitoring tool has been developed for the JADE platform. In response to high levels of demand, the system provisions further JADE platforms using cloud instances. The new platforms are automatically linked with existing ones using the JADE protocols. When demand falls back to the original level, then the cloud platforms are closed and the system returns to its original state. In order to evaluate the proposed architecture a prototype anti-terror application has been implemented which may be deployed at an airport to provide extra security. This system uses the cloud to scale the decision support features of a real-time terrorist identification platform.