⚡ electrical engineering

Full Duplex ISAC with Cluster Ray Targets: Parameter Estimation and Beamforming

This paper proposes a full-duplex ISAC framework for spatially distributed systems that employs a two-stage FFT and Gauss-Newton estimator to overcome the limitations of conventional methods in localizing distributed targets, alongside an adaptive beamformer that simultaneously satisfies sensing resolution and multi-user communication requirements.

Original authors: Muhammad Talha, Besma Smida, David González G

Published 2026-06-23

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Muhammad Talha, Besma Smida, David González G

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a future where your car's radio doesn't just play music, but also acts like a super-powered flashlight and a sonar system simultaneously. This is the world of ISAC (Integrated Sensing and Communication). The paper you provided tackles a specific, tricky problem within this technology: how to "see" objects that aren't just single dots, but are actually fuzzy, spread-out clouds (like a whole group of cars or a large building) while still talking to many people at once.

Here is a breakdown of the paper's ideas using everyday analogies.

1. The Problem: The "Fuzzy Cloud" vs. The "Pinpoint"

Most current radar and communication systems are designed to see targets as single points, like a pinprick of light on a map. This works great for a single car far away.

However, in the real world (especially for self-driving cars), targets are often distributed clusters. Think of a target not as a single pin, but as a foggy cloud or a swarm of bees.

The Issue: Traditional high-tech math tools (like the "MUSIC" algorithm mentioned in the paper) are like trying to find a single needle in a haystack. When the target is a whole "cloud" of rays, these tools get confused because the signal looks "full" and messy rather than sharp. They fail to tell you where the center of the cloud is or how wide the cloud spreads.
The Goal: The authors wanted a system that could accurately measure both the center of this "cloud" and its width (spread), even when the signal is weak (low battery/signal strength).

2. The Solution: A Two-Stage Detective

To solve this, the authors built a two-step estimation process. Think of it like a detective trying to identify a suspect in a crowded room.

Stage 1: The Wide-Angle Search (The "Coarse" Estimate)
- The Tool: They use a Fast Fourier Transform (FFT).
- The Analogy: Imagine scanning a dark room with a wide-angle flashlight. You can't see the details yet, but you can quickly spot roughly where the movement is happening. This stage uses a clever math trick (called "Manifold Separation") to turn a complex, curved problem into a straight line, allowing the computer to use a very fast "searchlight" algorithm to find the general direction and spread of the targets.
- Result: It gives a quick, rough guess of where the "clouds" are.
Stage 2: The Magnifying Glass (The "Fine" Estimate)
- The Tool: They use the Gauss-Newton method.
- The Analogy: Once the wide-angle flashlight spots the general area, the detective pulls out a magnifying glass. This step takes the rough guess from Stage 1 and "fine-tunes" it. It adjusts the numbers slightly to get a precise measurement of the center and the exact width of the cloud.
- Result: This step dramatically improves accuracy, especially when the signal is weak (like trying to see in the fog). The paper claims this method is three times better at measuring the "width" of the target than older methods when the signal is weak.

3. The Beamformer: The "Smart Spotlight"

The system isn't just a passive observer; it has to send out signals to do the sensing. This is where Beamforming comes in.

The Challenge: The system has to do two things at once:
1. Talk to several users (like sending data to your phone).
2. Sense the spread-out targets (like illuminating a whole parking lot, not just one car).
3. Avoid Self-Interference: Since the system is "Full Duplex" (talking and listening at the exact same time), the loud voice of the transmitter can deafen its own ears.
The Solution: The authors designed an adaptive beamformer.
- The Analogy: Imagine a stage spotlight operator who has to do three things simultaneously:
  1. Shine a tight, bright beam on a specific actor (the communication user).
  2. Shine a wide, diffuse floodlight over the entire stage to see the scenery (the distributed target).
  3. Wear noise-canceling headphones so the loud spotlight doesn't deafen them.
- The paper's math creates a "smart spotlight" that shapes the beam perfectly to cover the whole "cloud" of the target while still delivering fast data to users and keeping the system from getting overwhelmed by its own noise.

4. The Results: Why It Matters

The authors ran computer simulations to test their new "Two-Stage Detective" and "Smart Spotlight."

Speed and Efficiency: Their method is much faster than older methods. While older methods had to do heavy, slow math for every single possible angle, their method uses a "lookup table" (like a pre-calculated cheat sheet) to find the answer almost instantly.
Accuracy:
- In low-signal conditions (like a stormy night), their method was three times better at measuring how "spread out" a target is compared to the next best method.
- Even for targets that are very tight (narrow clouds), they were twice as accurate.
Balance: The system successfully managed to send data to users and sense the targets at the same time without the system "blinding" itself.

Summary

In simple terms, this paper presents a new way for future wireless systems to see "fuzzy" objects (like groups of cars or buildings) rather than just single points. It does this by using a two-step process (a quick scan followed by a precise adjustment) and a smart beam-shaping technique that allows the system to talk to people and scan the environment simultaneously without getting confused or overwhelmed. The result is a system that is faster, more accurate, and better at handling weak signals than current technology.

Technical Summary: Full Duplex ISAC with Cluster Ray Targets

Problem Statement
This work addresses critical limitations in Integrated Sensing and Communication (ISAC) systems, specifically within the context of next-generation 6G wireless networks. Conventional high-resolution estimation methods, such as MUSIC, fail to accurately localize distributed targets because the signal subspace becomes full-rank even in single-target scenarios. Furthermore, existing literature often models radar targets as point objects in the far field, an assumption valid for communication users but invalid for vehicle-to-infrastructure (V2I) and vehicle-to-everything (V2X) sensing scenarios where targets possess spatial extent. Additionally, existing extended target models (e.g., truncated Fourier series) often incur prohibitive computational costs and rely on deterministic assumptions that ignore the stochastic nature of radar channels. Finally, prior beamforming designs often neglect the constraints of self-interference (SI) in full-duplex (FD) systems and the simultaneous requirement to illuminate distributed targets while satisfying user data rates.

Methodology
The authors propose a framework for a Full-Duplex (FD) MIMO ISAC system operating at millimeter-wave frequencies using OFDM waveforms. The system serves multiple downlink communication users while sensing $K$ distributed Cluster Ray (CR) targets, where each target is modeled as a cluster of rays with specific angular and range spreads, adhering to 3GPP standards.

The core of the proposed solution consists of two main components:

Two-Stage Parameter Estimator:
- Stage 1 (Coarse Estimation): The authors employ a Manifold Separation Technique (MST) to transform the nonlinear directional phase components of the array response vectors into linear phase components. This allows the construction of the target spectrum using a Fast Fourier Transform (FFT). By decoupling the spatial manifold from the array dimensions, the method constructs a 2D lookup table via 1D and 2D FFT operations. This enables the rapid evaluation of the spatial spectrum across a grid of angles and angular spreads using a moving-average filter, circumventing the need for expensive matrix multiplications at every grid point.
- Stage 2 (Fine Estimation): To refine the coarse estimates, a Gauss-Newton optimization method is applied. This stage minimizes a weighted least-squares objective function based on the difference between the theoretical and estimated covariance matrices. Analytical derivatives are used to update the parameter vector (mean angle, angular spread, and signal power) iteratively.
Adaptive Beamforming Optimization:
The paper proposes an optimization framework to design the transmit beamformer ( $V$ ). The objective is to minimize the weighted distance between the designed beamformer and an "ideal" beamformer that fully illuminates the spatial extent of all targets. This optimization is subject to three constraints:
- Data Rate: Meeting the minimum data rate requirements for all communication users.
- Self-Interference: Limiting the residual SI power at the receiver RF chains (after analog cancellation) to prevent saturation.
- Power: Adhering to total transmit power limits.
  The non-convex data rate constraint is reformulated into a Second-Order Cone (SOC) constraint, rendering the overall problem convex and solvable via standard tools like CVX.

Key Contributions

Distributed Target Modeling: The work is the first to study spatially spread or clustered targets within the ISAC context using a computationally efficient distributed CR model (proposed by 3GPP), moving beyond point-object assumptions.
Low-Complexity Estimation: The proposed two-stage estimator successfully resolves multiple distributed targets. It achieves this without the $O(N_A^3)$ computational burden per grid point associated with conventional subspace-based methods (like Zoubir's, DISPARE, or Generalized Capon), instead utilizing FFT-based operations to reduce complexity significantly.
Joint Sensing and Communication Beamforming: The authors introduce an optimization framework that synthesizes adaptive beams capable of covering the full extent of distributed targets while strictly maintaining user Quality of Service (QoS) and managing self-interference in FD systems.

Results
Simulation results demonstrate the efficacy of the proposed framework:

Spread Estimation: The proposed method shows a threefold improvement in estimating angular spread under low Signal-to-Noise Ratio (SNR) conditions compared to baseline subspace-based schemes. For low-spread targets, a twofold improvement is observed.
Angle Estimation: While the proposed estimator tracks the performance of baselines in low SNR regimes, it outperforms them significantly once SNR exceeds 10 dB.
Robustness: The fine-tuning stage provides greater accuracy improvements for angular spread ( $\sigma_\theta$ ) than for the mean angle ( $\theta$ ). The method remains robust across varying target spreads, maintaining superior performance even as dispersion widens, whereas some baseline methods (like Generalized Capon) show degradation or bias at extreme spreads.
Beamforming: The designed beampattern closely matches the ideal sensing pattern required to illuminate targets, even while satisfying user data rate constraints and keeping SI power below the RF saturation threshold.

Significance
The paper claims that its primary significance lies in resolving the computational and modeling bottlenecks of ISAC systems dealing with distributed targets. By utilizing the MST-based approach, the authors provide a scalable solution that avoids the prohibitive complexity of traditional grid-search estimators for massive MIMO arrays. Furthermore, the work bridges the gap between theoretical target modeling and practical system design by integrating target illumination requirements directly into the beamforming optimization, ensuring that sensing performance is not compromised by communication demands or hardware limitations like self-interference. The results suggest that this framework offers a viable path toward high-resolution, full-duplex ISAC systems capable of handling the complex, stochastic nature of real-world extended targets.

1. The Problem: The "Fuzzy Cloud" vs. The "Pinpoint"

2. The Solution: A Two-Stage Detective

3. The Beamformer: The "Smart Spotlight"

4. The Results: Why It Matters

Summary

More like this