Date of Award

January 2022

Document Type


Degree Name

Doctor of Philosophy (PhD)


Aerospace Sciences

First Advisor

Joseph Vacek


This dissertation provides a numerical analysis of a Receiver Only Synchronization (ROS) protocol which is proposed for use by Unmanned Aircraft Systems (UAS) in Beyond Visual Line of Sight (BVLOS) operations. The use of ROS protocols could reinforce current technologies that enable transmission over 5G cell networks, decreasing latency issues and enabling the incorporation of an increased number of UAS to the network, without loss of accuracy. A minimum squared error (MSE)-based accuracy of clock offset and clock skew estimations was obtained using the number of iterations and number of observations as independent parameters. Although the model converged after only four iterations, the number of observations needed was considerably large, of no less than about 250. The noise, introduced in the system through the first residual, the correlation parameter and the disturbance terms, was assumed to be autocorrelated. Previous studies suggested that correlated noise might be typical in multipath scenarios, or in case of damaged antennas. Four noise distributions: gaussian, exponential, gamma and Weibull were considered. Each of them is adapted to different noise sources in the OSI model. Dispersion of results in the first case, the only case with zero mean, was checked against the Cramér-Rao Bound (CRB) limit. Results confirmed that the scheme proposed was fully efficient. Moreover, results with the other three cases were less promising, thus demonstrating that only zero mean distributions could deliver good results. This fact would limit the proposed scheme application in multipath scenarios, where echoes of previous signals may reach the receiver at delayed times. In the second part, a wake/sleep scheme was imposed on the model, concluding that for wake/sleep ratios below 92/08 results were not accurate at p=.05 level. The study also evaluated the impact of noise levels in the time domain and showed that above -2dB in time a substantial contribution of error terms disturbed the initial estimations significantly. The tests were performed in Matlab®. Based on the results, three venues confirming the assumptions made were proposed for future work. Some final reflections on the use of 5G in aviation brought the present dissertation to a close.