An Intelligent Wireless Communication Scheme for Remote Telemetry in Forest Terrains using a Multi-Band Network
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
UNIVERSITI MALAYSIA SARAWAK
Abstract
Description
The thesis presents a communication scheme called Mi-WiRe-Net that makes use of a multiband scheme employing Artificial Intelligence for optimal channel allocation to achieve wireless remote telemetry in difficult terrains. The domain chosen for this study is remote forest terrains with thick foliage. In such, popular wireless signals have difficulty in penetration due to their sharp attenuation in dense foliage. The domain also has other constraints such as the available amount of computing resources, power supply, infrastructure available and skilled personnel as well. The lower frequency signals overcome obstacles present in difficult terrains such as heavy undergrowth and addresses the issue of attenuation. Multi-band with Artificial Intelligence will optimise the frequency allocation for maximum throughput as well as reliability. The pure implementation of the Ant Colony Artificial Intelligence algorithm makes the system simple, yet satisfying complex requirements such as self-configuring and self-adapting, optimized, and consistent. The new system can have significance in wider applications than specified in the scope and can be extended to provide services for a wide domain of urban and rural computing services. Mi-WiRe-Net is a new communication scheme that makes use of low-frequency signals to get over the connectivity obstacle in dense foliage. The scheme is also adaptive to the environment to select the signal channel with the best throughput from a set of open channels forming a multi-band scheme. The scheme proposed is unique in that it keeps multiple channels open to achieve parallel transmissions for an overall improvement in throughput. The minimal intelligent algorithm based on a pure implementation of ant-colony optimisation will ensure that the best available channel is always used first before the slower channels are employed. The intelligence will focus on end to end throughput, which encompasses all parameters involved in the communication, which also ensures that the scheme is intelligent as well as computationally resource efficient. The performance benchmarking was done using simulation resulted in the system improve transmission throughput (isolated) from about 150% and more for the observed worse-case scenario. The Monte Carlo based simulation shows that the AI optimised scheme can show up to about 95% gain in transmission time when compared to the multi band scheme in the nomlal case. Hence proved that 1) the new scheme is able to achieve connectivity, which is otherwise absent, in thick foliage using low-frequency carrier signals, 2) make the new scheme more resilient and efficient by using a parallel multi-band approach and 3) optimise the scheme proposed using a least-cost artificial intelligence algorithm adaptation of the Ant Colony Optimisation.