Shyam Venkatasubramanian

Shyam Venkatasubramanian

ECE PhD Student @ Duke University

Talks and presentations

Data-Driven Radar Target Position and Velocity Estimation

August 10, 2023

Talk, U.S. Air Force Research Laboratory (AFRL), Dayton, OH

Target position and velocity estimation is a crucial aspect in the design of modern radar systems, being foundational to various radar applications such as surveillance, navigation, and military operations. This task involves determining the position and velocity of a target in the presence of clutter and noise using radar STAP techniques. Much of the framework for estimating these target attributes is defined post radar STAP detection, where received radar signals are spatially and temporally processed to detect and subsequently estimate target positions and velocities. Despite their advancements, however, radar STAP methods have several limitations, including (1) sensitivity to noise and interference, and (2) high computational costs. Expanding upon (1), the interference due to clutter, modeled via the clutter covariance matrix, must be estimated from limited available samples when it is unknown a priori. This weakens any optimality claims made regarding the adaptive radar detection and subsequent target position and velocity estimation procedures (i.e., finding the peak location from a detection test statistic). Resultantly, while optimal detection can be achieved in the non-adaptive case (where the true clutter covariance matrix is known), we cannot make claims of invariance in the adaptive case (where the clutter covariance matrix is unknown). As a notional proof in the context of suboptimal detection for the adaptive case, we analyze whether the accuracy of target position and velocity estimation post radar STAP detection can be improved using a data-driven methodology.

Subspace Perturbation Analysis for Data-Driven Radar Target Localization

May 04, 2023

Conference proceedings talk, 2023 IEEE Radar Conference, San Antonio, TX

Recent works exploring data-driven approaches to classical problems in adaptive radar have demonstrated promising results pertaining to the task of radar target localization. Via the use of space-time adaptive processing (STAP) techniques and convolutional neural networks, these data-driven approaches to target localization have helped benchmark the performance of neural networks for matched scenarios. However, the thorough bridging of these topics across mismatched scenarios still remains an open problem. As such, in this work, we augment our data-driven approach to radar target localization by performing a sub-space perturbation analysis, which allows us to benchmark the localization accuracy of our proposed deep learning framework across mismatched scenarios. To evaluate this framework, we generate comprehensive datasets by randomly placing targets of variable strengths in mismatched constrained areas via RFView®, a high-fidelity, site-specific modeling and simulation tool. For the radar returns from these constrained areas, we generate heatmap tensors in range, azimuth, and elevation using the normalized adaptive matched filter (NAMF) test statistic. We estimate target locations from these heatmap tensors using a convolutional neural network, and demonstrate that the predictive performance of our framework in the presence of mismatches can be predetermined.

Data-Driven Target Localization Using Adaptive Radar Processing and Convolutional Neural Networks

July 29, 2022

Talk, U.S. Air Force Research Laboratory (AFRL), Dayton, OH

Facilitated by the recent emergence of radio frequency (RF) modeling and simulation tools purposed for adaptive radar processing applications, data-driven approaches to classical problems in radar have rapidly grown in popularity over the past decade. Despite this surge, limited focus has been directed toward the theoretical foundations of these data-driven approaches. In this regard, using adaptive radar processing techniques, we propose a data-driven approach in this work to address the classical problem of radar target localization post adaptive radar detection. To give context to the performance of this data-driven approach, we analyze the asymptotic breakdown signal-to-clutter-plus-noise ratio (SCNR) threshold of the normalized adaptive matched filter (NAMF) test statistic within the context of radar target localization, and augment this analysis by benchmarking our proposed deep learning framework for target location estimation across matched and mismatched settings.

Toward Data-Driven STAP Radar

March 23, 2022

Conference proceedings talk, 2022 IEEE Radar Conference, New York, NY

Using an amalgamation of techniques from classical radar, deep learning, and computer vision, we characterize our ongoing data-driven approach to space-time adaptive processing (STAP) radar. We generate a rich dataset of received radar signals for random target locations and strengths via the RFView simulation software, and generate heatmap image tensors in range, azimuth, and elevation of this received signal data. We present a regression convolutional neural network (CNN) for estimating target locations from these image tensors, and conclude by demonstrating the feasibility of and significant improvements provided by our data-driven approach.