Abstract
Abstract : Cellular communication networks are large and distributed systems that provide billions of people around the world with means of communication. Antennas as used currently in cellular communication networks do not provide efficient resource management given the growth in the current communication network scenario. Most of the problems are related to the number of devices that can connect to an antenna, the coverage map of an antenna, and frequency management. A smart antenna grid can cover the same area as traditional cellular system towers with some enhancements. Smart antenna grids can include a device in an area that requires connectivity rather than covering of the entire area. Frequencies are handled per antenna base, with more focus on providing stable communication. The objective of the dissertation is to improve resource management of smart antenna grids by making use of a multi-agent system. The dissertation uses a simulation environment that illustrates a smart antenna grid that operates with a multi-agent system that is responsible for resource management. The simulation environment is used to execute ten scenarios that intends to place large amounts of strain on the resources of the smart antenna grid to determine the effectiveness of using a multi-agent system. The ten scenarios show that when resources deplete, the multi-agent system intervenes, and that when there are too many devices connected to one smart antenna, the devices are managed. At the same time, when there are antennas that have frequency problems, the frequencies are reassigned. One of the scenarios simulated the shutdown of antennas forcing devices to disconnect from the antenna and connect to a different antenna. The multi-agent system shows that the different agents can manage the resources in a smart grid that is related to frequencies, antennas and devices.
M.Sc. (Computer Science)