Mobile Agent Based bio-Inspired Mechanisms for Servicing Networked Robots
No Thumbnail Available
The development of middleware for networked robotics offers challenges that are often quite different from those encountered in singular robots. It calls for the design and development of a software framework that can facilitate interactions amongst the distributed, interoperable, heterogeneous robotic entities that comprise the network and the simplification of complex robot control software systems which in turn can ease the associated application development process. Mobile agents form one of the virtual machine based paradigms that are ideally suited to the development of such middleware. A framework that exploits all features of mobile agents to form the basic middleware is still missing. The work described in this thesis aims at the realization of an Artificial Being (AB) comprising a network of mobile robots which are serviced by mobile agents. While the robots form the physical effectors of the being the mobile agents and the network facilitate the movement of information amongst them. The being is portrayed to use bio-inspired paradigms. Application scenarios where mobile agents carry services in the form of programs for networked robots have been depicted. Bio-inspired mechanisms - PherCon and PherCon-C have been proposed to provide for faster servicing by the mobile agents. Both simulation and real-time experiments were carried out and their results have been portrayed and discussed. An increase in the number of agents through cloning can enhance the performance of such networked systems. However, uncontrolled cloning can lead to excessive consumption of system resources including network bandwidth. The work reported in the thesis addresses this problem and proposes a novel stigmergy based method for control of populations of heterogeneous mobile agents. As an offshoot of this research, a mobile agent framework for use in conjunction with an Internet of Things (IoT) has been proposed and its implementation described..
Supervisor: S B Nair
COMPUTER SCIENCE AND ENGINEERING