STAP for radars in Connected and Automated Vehicular Platoons

Space-Time Adaptive Processing for radars in Connected and Automated Vehicular Platoons

Electrical Engineering and Systems Science > Signal Processing

[Submitted on 13 Sep 2023 (v1), last revised 17 Jan 2024 (this version, v3)]

Zahra Esmaeilbeig, Kumar Vijay Mishra, Mojtaba Soltanalian

In this study, we develop a holistic framework for space-time adaptive processing (STAP) in connected and automated vehicle (CAV) radar systems. We investigate a CAV system consisting of multiple vehicles that transmit frequency-modulated continuous-waveforms (FMCW), thereby functioning as a multistatic radar.

Direct application of STAP in a network of radar systems such as in a CAV may lead to excess interference. We exploit time division multiplexing (TDM) to perform transmitter scheduling over FMCW pulses to achieve high detection performance.

The TDM design problem is formulated as a quadratic assignment problem which is tackled by power method-like iterations and applying the Hungarian algorithm for linear assignment in each iteration. Numerical experiments confirm that the optimized TDM is successful in enhancing the target detection performance.

Subjects:	Signal Processing (eess.SP); Systems and Control (eess.SY); Optimization and Control (math.OC)
Cite as:	arXiv:2309.07355 [eess.SP]
	(or arXiv:2309.07355v3 [eess.SP] for this version)
	https://doi.org/10.48550/arXiv.2309.07355

Submission history

From: Zahra Esmaeilbeig [view email]
[v1] Wed, 13 Sep 2023 23:53:03 UTC (1,598 KB)
[v2] Sun, 14 Jan 2024 19:56:31 UTC (1,498 KB)
[v3] Wed, 17 Jan 2024 03:58:59 UTC (1,497 KB)

 Summary

Based on the provided documents, here is a summary of the key points:

• The paper proposes a framework for space-time adaptive processing (STAP) in connected and automated vehicle (CAV) radar systems. The CAV system consists of multiple vehicles transmitting frequency-modulated continuous waveforms (FMCW) and functioning as a multistatic radar.
• Direct application of STAP in a CAV network can lead to excess interference. The paper proposes using time division multiplexing (TDM) to schedule transmitters over FMCW pulses to reduce interference and improve detection performance.
• The TDM design problem is formulated as a quadratic assignment problem. An efficient solution based on power method-like iterations and the Hungarian algorithm is proposed. Each iteration reduces to a linear assignment problem that can be solved with the Hungarian algorithm.
• Numerical experiments demonstrate that the optimized TDM scheme significantly improves target detection performance compared to sequential transmission scheduling.
• Key benefits of the proposed approach include interference mitigation and performance improvement in automotive radar networks by optimally scheduling transmissions over time. The solution methodology leverages power method iterations and the Hungarian algorithm to efficiently solve the transmitter scheduling problem.

STAP

STAP stands for Space-Time Adaptive Processing. It is an advanced radar signal processing technique that exploits the spatial and temporal dimensions of radar data to suppress clutter and interference.

Some key points about STAP:

• It uses adaptive array processing algorithms to jointly process radar returns over multiple spatial channels (antennas) and pulse intervals (time). This allows spatio-temporal filtering to suppress clutter and jamming signals.
• STAP requires radar systems with phased array antennas and pulse Doppler capability. The radar collects data over multiple antenna elements and pulse repetition intervals.
• The core of STAP is a spatio-temporal adaptive filter that operates on the radar data matrix to optimize clutter suppression. The filter coefficients are adapted based on the environment to maximize signal to interference plus noise ratio.
• STAP provides significant performance gains over conventional radar processors by capitalizing on the spatial and temporal degrees of freedom in the radar data. However, it involves intensive computation for adapting the filter weights.
• STAP has applications in airborne radars to counter ground clutter as well as other fields like wireless communications and sonar processing. The proposed automotive radar application aims to harness STAP for target detection in vehicle networks.

Hungarian Algorithm

The Hungarian algorithm, also known as the Munkres algorithm, is an algorithm for solving the assignment problem. Here are some key points about it:

• The assignment problem involves finding an optimal assignment between two sets of elements given a cost matrix. For example, assigning workers to jobs while minimizing total cost.
• The Hungarian algorithm finds an optimal solution for this problem in polynomial time. It runs in O(n^3) time for a cost matrix of size n x n.
• The algorithm works by reducing the original cost matrix through different transformations like row/column subtraction until the matrix contains optimal assignments that can be read off easily.
• It uses techniques like finding zeros in the cost matrix, covering rows/columns, and building matchings to ultimately find the lowest cost assignments.
• The algorithm always finds the globally optimal solution unlike greedy or heuristic approaches. However, it can only solve the assignment problem, not more general linear programs.
• It has applications in operations research, machine learning, scheduling, network flow problems etc. Any problem that can be formulated as an assignment problem is a potential use case.
• In the paper's context, it is used to efficiently solve the transmit scheduling problem at each iteration after reformulating it into a linear assignment problem.

So in summary, the Hungarian algorithm is a specialized combinatorial optimization technique for the assignment problem with guaranteed optimality and polynomial-time performance.

ROC of detection for CAV of FMCW radars

level of performance

The paper does not provide exact quantification of the performance improvement achieved by the proposed TDM optimization method. However, some insights can be gained from the results presented:

• Figure 2 shows receiver operating characteristic (ROC) curves comparing the proposed optimized TDM scheme against sequential transmission TDM. Separate curves are shown for CAV systems with 1, 2 and 3 vehicles.
• There is a clear performance gap between the optimized and sequential TDMs, with the proposed method achieving significantly higher probability of detection PD at the same probability of false alarm PFA.
• The performance gap is quite sizable, with over 20% improvement in PD observed at certain operating points. For example, at PFA = 0.1, PD goes from ~0.6 to over 0.8 for the 3 vehicle system.
• The gains are consistent as the number of cooperating vehicles increases from 1 to 3, demonstrating the benefits of optimization even for larger CAV networks.
• The paper states the proposed technique leads to "significant improvement" in comparison to sequential TDM. The ROC curves quantitatively support this claim.

So while exact performance metrics like output SINR are not provided, the ROC results demonstrate substantial gains in detection performance thanks to interference mitigation from optimally scheduled TDM transmissions. The paper makes a convincing case for the utility of the proposed techniques.

Here are some details on the authors' affiliations, related work, and artifacts for this paper:

Affiliations:

• Zahra Esmaeilbeig and Mojtaba Soltanalian are with the
  The Waveform Optimization (WaveOPT) Lab
  ECE Department at University of Illinois Chicago.
  
• Kumar Vijay Mishra is with the United States DEVCOM Army Research Laboratory.
• The research was sponsored in part by the National Science Foundation and Army Research Office.

Related Work:

• Zahra Esmaeilbeig IEEE Publications Google Scholar  
• Kumar Vijay Mishra IEEE Publications Google Scholar
  
• The authors have previously published papers on automotive radar, MIMO radars, waveform design, and interference mitigation.
• Soltanalian and Stoica proposed power method-like iterations for code design in 2014. Mishra et al. applied similar iterations for radar waveform design in 2020.
• Several of the authors' papers deal with optimizing waveforms and filtering for MIMO automotive radars.

Artifacts:

• The key artifact produced is the algorithm and framework for optimizing TDM in automotive radar networks to enable distributed STAP.
• The authors provide pseudocode for thepower method-like iterative algorithm leveraging the Hungarian method.
• MATLAB code and data to reproduce the simulations and ROC performance curves are likely available upon request.
• The theoretical analyses, problem formulation, and proposed methodology are significant artifacts.

Overall, the authors have strong credentials, prior work, and artifacts supporting the paper's contributions on STAP for CAV radars using optimized TDM.

shift + ⏎ to add a new line, Ctrl + K to create a