Previous Abstract Return to Session D6 Next Abstract

Session D6: Algorithms and Methods

Hardware Validation of an Adaptive Optimization Algorithm for Tap Delay Wavefront
Gregory Reynolds, U.S. Army, Aviation and Missile Research, Development, and Engineering Center (AMRDEC); Laurie Joiner, University of Alabama in Huntsville
Location: Spyglass

Hardware Validation of an Adaptive Optimization Algorithm for Tap Delay Wavefront
PLANS Abstract, Gregory Reynolds, October 2017
Some of the biggest investments surrounding Global Navigation Satellite System (GNSS) receivers are the development and integration of adaptive nulling technology. Researchers, developers, and integrators need to understand the effects that these nulling technologies have on the navigation receivers and systems using the modified RF output signals. Due to the sensitive nature of the GNSS spectrum, outdoor testing of Anti-Jam (AJ) and Controlled Reception Pattern Antenna (CRPA) technologies is difficult, expensive, and often cause schedule delay. It is becoming more important to develop cost effective means to evaluate adaptive nulling advancements and their effects on the receivers with which they are being integrated. Aside from open air testing, this evaluation and integration is best accomplished through modeling and simulation techniques.
Demand for modeling and simulation of adaptive nulling technologies continues to grow because it provides a safe and repeatable environment for researching and developing CRPA and AJ electronics. A primary challenge for wavefront simulation design is the ability to create phase coherent signals using time invariant hardware at an affordable price. This paper describes a wavefront architecture using mechanical tap delay lines that brings an affordable, high fidelity simulation capability to the forefront.
Most methods of using active RF front ends to create the wavefront simulation are time varying due to temperature, clock input, and various other effects. These effects of the varying phase coherency are unknown during the test and lead to erroneous truth data that corrupts the performance evaluation. The wavefront simulation system described in this paper is comprised of mechanical tap delay lines to control the per-element phase differences and inherently applies a time offset and Doppler shift during the phase changes. The mechanical tap delay lines are very precise and not affected by run time or temperature, and do not require reference clocks for phase coherent output or synchronization.
Due to the simulation input for antenna manifold data and the discrete step size of the tap delay output the methodology degrades in its ability to simulate a stable signal when the source is near planar with the antenna elements or near zenith. This results from the lack of resolution in the tap delay architecture as the system components of the antenna and AJ electronics have more visibility into the RF signals characteristics. All wavefront simulations have this challenge, but it has been shown that this gap in capability can be overcome when the system is linear and time-invariant. Adaptive optimization algorithms have been developed for this simulation architecture and are demonstrated in a digital simulation environment and a hardware evaluation of a Commercial Off-The-Shelf (COTS) system.
Previous work is presented in an IEEE paper that fully describes the theory behind the wavefront RF system model and the method of adaptively optimizing this architecture. This theory has also been implemented in a digital simulation environment. Typical systems of this nature are calibrated on a per-channel basis and the corresponding taps are chosen in the same manner. The analysis looks at using an observer model to relate the incoming angle of arrival signals to the antenna frame and relates the signal through a cone-sphere intersection that defines the observation ambiguity of each element-reference pair. A Steiner tree model is then applied to the observation intersections to define the branches that provide the method by which the optimization is applied.
This paper continues the development of the Steiner tree theory by providing hardware simulation results of the adaptive algorithms using tap delay lines and COTS anti-jam electronics. The system is evaluated through various simulation trajectories based on desired input variables. The results are produced through covariance analysis of the perceived signal from the AJ unit as compared to the simulated signal. The results include the all-digital simulations for the same trajectories in addition to the hardware evaluation. This provides a basis for the accuracy and expandability of the analysis in the absence of physical AJ or CRPA technology. This optimization creates a simulation of a wavefront that overcomes the lack of resolution in the zenith and planar regions and creates a predictable incident wave that provides the user with better truth data for performance analysis of the unit under test.



Previous Abstract Return to Session D6 Next Abstract