| dc.description.abstract | On our everyday operations there is need to engage agents to perform some duties on our behalf,
hence they are gaining acceptance as a technology and are being used. Most of the networked
offices, networked homes, cyber cafe's, learning institutions and other arenas where computers
are interconnected on a Wi-Fi network, have peer-to-peer networks. In Wi-Fi peer-to-peer
networks, it is difficult to identify the network details of all the network devices connected such
as the IP adresses, mac addresses and computer names of all computers connected on the Wi-Fi
peer to peer network at one go; which we hereby reffer to as fundamental network details. This is
mainly possible in a client-server based architecture where the server monitors all the computers
on the network. From the above gap, we developed a mobile agent that could be run in any
computer on the Wi-Fi peer-to-peer network and it lists these fundamental details of all the
computers connected to the Wi-Fi peer-to-peer network. In developing this mobile agent, we used
the MaSE agent methodology. The mobile agent was coded, implemented and tested. The agent
was then subjected to various controls which it overcame and managed to return the desired
fundermental details with over 80% accuracy. The agent also had the capacity to classify every
computer on the network as either intruder or non intruder based on the list of authorised
computers supplied by the user. The agent suffered major limitation which included: - the agent
took longer time to learn and return the results, as well as it could not communicate to the
intruders or shut them down. In future, the agent could therefore be improoved to reduce its
processing time, communicate with the intruding computers, and shut down the intruding
computers or deny them network access. | en |
