Author: Jeffrey Twigg
Date: Thursday, April 7th, 2022 at 3:30pm
Location: Glenn L. Martin Hall, Room EGR-2164
Committee Members:
Professor Nikhil Chopra, Chair/Advisor
Professor Miao Yu
Professor Yancy Diaz-Mercado
Professor Shapour Azarm
Professor Dinesh Manocha, Dean’s Representative
Dr. Brian Sadler
Title: On the Control of Robotic Parasitic Antenna Arrays
Abstract:
Wireless networking is challenging in safety, security, and rescue contexts where network infrastructure may be damaged or compromised. Radio communication between ground robots at the lower end of the Very High Frequency (low-VHF) band is generally more reliable in complex indoor and urban environments when compared to communication systems such as Wi-Fi and cellular which operate at Ultra High Frequencies (UHF) and higher frequencies. Exciting antenna design research in the last 5 to 10 years has approached what is theoretically possible to create compact, moderately high bandwidth antennas at low-VHF. At the beginning of this dissertation research, we discovered that we could distribute these low-VHF antennas across closely positioned ground robots to create a robotic parasitic antenna array. When these robots are optimally positioned, they create a directional signal through the mutual coupling of their antennas. Consequently, these low-VHF arrays have the potential to extend the communication range of a reliable signal in urban and indoor environments with a proportionally small amount of robotic motion.
In this dissertation, we research the control of robotic platforms constituting these arrays from two perspectives. First, we research how robots control their positions to optimize or maintain the gain of a single robotic parasitic array to improve the quality of a communication link. Then, we investigate where these robots should collect to form an array in a network of robotic parasitic arrays to increase a metric of overall network connectivity.
To improve individual network communication links, we consider a two-element parasitic array formed by a static antenna and a ground robot and propose a technique by which this array can optimize its gain in a direction of interest. First, we propose and test an optimization approach for actuating spacing between the two antennas and passive antenna length to increase gain. Next, we propose and test an approach for using robotic motion to rotate the antenna array. In these experiments, we show that the robotic parasitic antenna array can provide a gain of 2 dB which is close to twice the effective transmission power in line-of-sight and non-line-of-sight conditions.
From a network perspective, we research where robots should form arrays to maintain a metric of overall connectivity. However, existing control formulations for maintaining connectivity are not general enough to support this new capability. We first propose a generalized model that we integrate into a Fiedler value maximization approach for maintaining communication. Afterward, we develop approaches for allocating a finite number of robots for forming these robotic parasitic arrays while ensuring that our metric for overall network connectivity between robotic parasitic array forming robots remains lower bounded.