Research work · 2014
Long Range Channel Predication in Aeronautical Telemetry
Marilynn P. Wylie-Green
UA Campus Repository (The University of Arizona)
Abstract
Long range prediction exploits the principal that the memory span of a predictive filter increases when it samples the input signal at a lower rate. Within the context of a communication system that employs adaptive modulation and coding communication, this implies that by measuring the channel at a rate that is minimally twice the highest Doppler frequency rather than twice the data rate (the latter typically being orders of magnitude greater than the former), one can predict the channel fades much farther ahead into the future. In this paper, we consider an application of a recursive weighted least squares estimation (WLSE) algorithm to the problem of long range channel prediction. The algorithm is applied to ten sets of channel measurements that have been made during channel sounding exercises on the Edwards Air Force Base test range and include the following scenarios: taxiway, takeoff, in-flight and final approach & landing. In this subset of the measurement campaigns, the test article speed varies from ≈ 5 m/s along the taxiway to ≈ 100 m/s while in flight. As our results show, the algorithm can be used in real time to predict several milliseconds in the future while using a modest number of filter coefficients (d = 2 and d = 4 are considered), thus demonstrating that channel prediction is viable in the aeronautical telemetry testing environment.