Teaching Ad-hoc Networks Using a Simple Agent Framework
Maja Pantić, Reinier Zwitserloot, R.J. Grootjans · 2005
This paper describes a method of teaching agent technology and ad-hoc networks using a novel, simple agent framework developed specifically for the purposes of teaching introductory artificial intelligence (AI) to undergraduate students. The agent framework in question is Java-implemented and it embodies the concepts of concurrency, multi-agency, persistency, and mobility. The introductory AI course work in question is a set of assignments that requires the students to use intelligent agents to route SMS messages through an ad-hoc network. While many AI courses teach basic AI concepts such as heuristic search, rule-based reasoning, and neural networks, by means of programming assignments, we were not able to find any AI course introducing students to relatively new AI concepts such as distribution, mobility and ad-hoc networks by similar means. The coursework presented in this paper represents a synthesis of the traditional objectivist approach and a real-world oriented, constructivist approach to introducing students to these "hot topics".