CRLB Minimization for Multi-Target Localization in Segmented Waveguide-Enabled Pinching-Antenna ISAC Systems

arXiv:2604.00572 · eess.SP · Submitted 2026-04-01 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Electrical Engineering. The day in electrical engineering: signals, systems and the hardware that carries them.

Rosa: I'm Rosa.

Dev: And I'm Dev, here with Taro.

Rosa: We're looking at the paper: "CRLB Minimization for Multi-Target Localization in Segmented Waveguide-Enabled Pinching-Antenna ISAC Systems".

Dev: Here's the gist of the paper. Pinching-antenna (PA) has recently attracted considerable research attention in wireless systems <ref:2604.00572#pg2> Building upon which, the segmented waveguide-enabled pinching-antenna system (SWAN) has been proposed to mitigate the inter-antenna radiation problem in uplink transmissions of conventional PA systems <ref:2604.00572#pg2>.

Rosa: In this work, SWAN-assisted integrated sensing and communication (ISAC) is investigated, where a base station (BS) equipped with SWAN provides downlink communications for multiple communication users (CUs) and performs sensing for multiple targets <ref:2604.00572#pg2>.

Dev: The main contributions are summarized as follows <ref:2604.00572#pg2>.

Rosa: System Architecture and Modeling The proposed system involves a base station (BS) transmitting and receiving signals via a SWAN system consisting of multiple transmit segmented waveguides (TSWs) and receive segmented waveguides (RSWs) <ref:2604.00572#pg2>.

Title and authors: Rosa: What the paper summarizes is that they’ve built this framework to find the best way to set up both the beamforming and the physical positions of these pinching antennas simultaneously, which they call a Riemannian product manifold optimization method >

Dev: They are using this optimization method because it treats those variables—the beamforming settings and where the antennas are placed—as being coupled together on a single surface, which is what makes it challenging to solve with standard math tools >

Taro: The summary mentions they're equipping the transmit power constraint with equality as a Riemannian complex sphere manifold, which is how they handle that physical limit in their optimization setup >

Rosa: It’s about showing that this joint optimization of beamforming and antenna positions leads to better results than just optimizing them separately >

Dev: They are using an iterative method called the Riemannian Broyden–Fletcher–Goldfarb–Shanno algorithm, which they apply on this manifold to find a solution >

Taro: And that iterative process involves calculating the Riemannian gradient by projecting the standard Euclidean gradient onto the tangent space of each manifold, which is a necessary step because you can't just use standard calculus there >

Rosa: It’s about making sure that whatever update you take in one step moves you toward a better solution on that curved space >

The paper's summary: Dev: What they point out as an improvement is that this joint optimization method gives them superior performance compared to existing schemes, specifically mentioning multi-waveguide-enabled pinching antenna assisted ISAC systems >

Taro: They claim this approach outperforms conventional MIMO and mult-waveguide enabled PASS in the context of integrated sensing and communication because of how they handle the signal transmission and reception >

Rosa: And they also point out that their optimization method itself is better than existing separately designed approaches, which highlights why joint optimization is important for this kind of system >

Dev: Plus, they look at a trade-off by showing that increasing the communication rate threshold actually degrades the Cramer Rao lower bound performance for all schemes, and their proposed method finds the best trade-off between rate and CRLB >

Taro: That means you can get better sensing accuracy without hurting your data throughput too much, but it depends on that communication rate setting >

Rosa: So they show that this SWAN assisted ISAC system is consistent across different values of the target position error factor ν, meaning it’s robust in terms of how accurate the localization needs to be >

The paper's improvements: Dev: To wrap up, the paper on "CRLB Minimization for Multi-Target Localization in Segmented Waveguide-Enabled Pinching-Antenna ISAC Systems" confirms that this SWAN assisted system is effective for multi-user multi-target sensing and communication >

Rosa: They focused on using the CRLB of target position estimation as the main metric to evaluate how good the localization is, and they successfully proposed an optimization method to solve that rate-constrained problem by jointly optimizing both beamforming and PA positions >

Taro: From my side, I think what's important is that this joint optimization actually works better in simulation than what you’d expect from just patching two separate solutions together >

Dev: And it does. It shows that the mathematical formulation using the Riemannian product manifold allows them to find a solution that is genuinely better than what you get from running those things separately >

Rosa: So, this work gives us a solid theoretical foundation for designing these complex systems where you need to keep communication quality high while also getting precise location data from several targets >

Conclusion: Rosa: So, we just looked at this paper called "CRLB Minimization for Multi-Target Localization in Segmented Waveguide-Enabled Pinching-Antenna ISAC Systems." Basically, they’re trying to figure out the best way to use these pinching antennas and segmented waveguides when you want to do both communication and sensing at the same time.

Dev: Yeah, they tackle that problem by using this Riemannian product manifold optimization method. It’s not just picking settings randomly; they are jointly optimizing the beamforming and where those antennas are physically placed, which is a big deal for how coupled the variables are.

Taro: When you think about it from a real-world testing standpoint, that joint optimization is what lets them achieve better accuracy in target localization compared to just tuning everything separately.

Rosa: Exactly. They show that this approach beats existing schemes, like those using multi-waveguide enabled pinching antenna assisted ISAC systems, especially when you look at the error factor of the target position.

Dev: And they also point out a trade-off with communication rate. They find that if you want a higher communication rate, your sensing accuracy gets worse for every system, but their method finds the best balance between those two things.

Taro: I just wonder about the practical side though, how stable is this optimization loop when you’re running it in real-time? If the world misbehaves and the measurements change quickly, does this complex optimization method stay reliable?

Rosa: That’s a fair question. The paper shows they used an iterative method called Riemannian Broyden–Fletcher–Goldfarb–Shanno applied on that manifold, and they used a line-search strategy to guarantee that the algorithm keeps moving toward a better solution, which helps with stability.

Dev: Right, the line-search strategy ensures monotonic descent. That’s important because we need to know if our control loop is actually converging and not just oscillating around some local minimum. And they also provided a complexity analysis, showing it runs in something like O of n squared times the number of elements times that stuff.

Taro: I see the complexity, but how does that translate when you move from simulation to an actual mobile robot or a remote sensor setup? Does that big O notation mean it’s fast enough for those tight loops?

Rosa: They did mention the computational complexity, which they estimated as O of n squared times the number of elements times that stuff. It’s a lot, but they also used a limited-memory strategy within their algorithm to make the inverse Hessian approximation efficient, which helps ease that high computational cost.

Dev: That limited memory trick is smart. It shows they thought about making the real-time execution feasible, even though the math behind it is quite heavy. So, to sum up this paper on "CRLB Minimization for Multi-Target Localization in Segmented Waveguide-Enabled Pinching-Antenna ISAC Systems," they proved that joint beamforming and antenna position optimization using Riemannian geometry leads to superior localization accuracy compared to previous methods.

Taro: It’s solid work, definitely proving the theoretical advantage of coupling those variables. But it leaves the question open about how robust this whole framework is when you push it into a truly chaotic environment where things don't behave according to their spherical wave models perfectly.

Rosa: Yeah, that’s exactly what we want to test next. We need to see if this performance holds up outside the lab and when the environment gets really messy and unpredictable. Next week, we’re looking at a different angle on MIMO decoding architectures for sort-free box decoding.

Yue Geng, Tee Hiang Cheng, Kah Chan Teh, Zhiguo Ding

Nanyang Technological University

eess.SP

Submitted: 2026-04-01

Updated: 2026-11-09

Comments: 14 pages, 6 figures

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 85/100

The gist: Pinching-antenna (PA) has recently attracted considerable research attention in wireless systems <ref:2604.00572#pg2> Building upon which…

Key concepts

Pinching-Antenna (PA)
A specialized antenna design that focuses energy in a specific direction. This paper builds upon PA research, using a segmented waveguide structure to improve performance and mitigate radiation issues when multiple antennas are used together.
Segmented Waveguide-Enabled Pinching-Antenna (SWAN)
An advanced antenna system where the transmitting and receiving elements are divided into segments connected by waveguides. This structure helps manage inter-antenna radiation problems, improving the overall performance of the communication and sensing system.
Cramer-Rao Lower Bound (CRLB)
A mathematical tool used to determine the theoretical minimum possible error in estimating a parameter, such as a target's location. Minimizing this bound helps establish the best possible performance achievable by the system for target localization.

Terminology

Summary

Pinching-antenna (PA) has recently attracted considerable research attention in wireless systems <ref:2604. p. 000572#pg2> Building upon which, the segmented waveguide-enabled pinching-antenna system (SWAN) has been proposed to mitigate the inter-antenna radiation problem in uplink transmissions of conventional PA systems <ref:2604. p. 000572#pg2>. In this work, SWAN-assisted integrated sensing and communication (ISAC) is investigated, where a base station (BS) equipped with SWAN provides downlink communications for multiple communication users (CUs) and performs sensing for multiple targets <ref:2604. p. 000572#pg2>. The main contributions are summarized as follows <ref:2604. p. 100572#pg2>.

System Architecture and Modeling

The proposed system involves a base station (BS) transmitting and receiving signals via a SWAN system consisting of multiple transmit segmented waveguides (TSWs) and receive segmented waveguides (RSWs) <ref:2604.00572#pg2>. The BS transmits dual-functional signals for communication users (CUs) and sensing targets, which are radiated by the TPA on each TSW, while echo signals from targets are captured by the RPA on each RSW <ref:2604. p. 000572#pg2>. The channel models for both communication and sensing channels are derived using spherical wave models for near-field propagation <ref:2604. p. 200572#pg2>. Specifically, the overall channel between the BS and a CU is given by equation (6) <ref:2604. p. 200572#pg2>. The target-PA channel for the k-th target is obtained as equation (13) <ref:2604. p. 200572#pg2>.

Optimization Framework

The performance evaluation is formulated as a Cramer-Rao lower bound (CRLB) minimization problem to evaluate the system performance, where the CRLB of location estimation is minimized under communication rate constraints <ref:2604.00572#pg2>. To obtain a feasible solution to this non-convex problem, a Riemannian manifold optimization (RMO) method is developed, which unifies beamforming and PA position variables on a Riemannian product manifold (RPM) <ref:2604. p. 100572#pg2>. The RPM construction involves equipping the transmit power constraint with equality as a Riemannian complex sphere manifold (CSM) <ref:2604. p. 500572#pg2>.

Optimization Algorithm Details

The RMO method solves the reformulated problem by integrating a penalty method with a smoothing technique to handle inequality constraints <ref:2604.00572#pg2>. The optimization is performed iteratively using the Riemannian Broyden–Fletcher–Goldfarb–Shanno (RBFGS) algorithm applied on the RPM <ref:2604. p. 000572#pg2>. This iterative process involves obtaining the Riemannian gradient by projecting the Euclidean gradient onto the tangent space of each manifold <ref:2604. p. 600572#pg2>. A limited-memory strategy is employed within Algorithm 1 to construct an efficient approximation of the inverse Hessian, alleviating high computational cost <ref:2604. p. 600572#pg2>.

Performance Analysis and Results

Simulation results demonstrate that the proposed SWAN-assisted ISAC system yields superior CRLB performance for target localization compared to existing schemes including the multi-waveguideenabled pinching-antenna-assisted ISAC systems <ref:2604.00572#pg2>. The proposed scheme achieves better CRLB performance than conventional MIMO and mult-waveguide-enabled PASS in the considered ISAC system <ref:2604. p. 100572#pg2>. Furthermore, the proposed RMO method outperforms those obtained by existing separately designed approaches, highlighting the benefit of joint optimization <ref:2604. p. 100572#pg2>. The study also shows that increasing the communication rate threshold degrades CRLB performance for all schemes, and the proposed method achieves the best rate–CRLB trade-off <ref:2604. p. 900572#pg2>.

Conclusion

In this paper, we have investigated a novel SWAN-assisted multi-user multi-target ISAC system <ref:2604.00572#pg2>. We adopted the CRLB of target position estimation as the sensing metric, and proposed an RMO method for solving the rate-constrained CRLB minimization problem by jointly optimizing the beamforming and PA positions <ref:2604. p. 400572#pg2>. Simulation results have verified the effectiveness of the proposed RMO method and the advantages of the SWAN scheme for multi-user multi-target ISAC systems compared with baseline schemes including conventional MIMO and multi-waveguide-enabled PASS schemes <ref:2604. p. 100572#pg2>. The work mainly focus on the theoretical performance of the SWAN-assisted ISAC systems <ref:2604. p. 100572#pg2>.

APPENDIX A

EXPRESSIONS OF THE SUB-MATRICES IN THE FIM

The elements of the sub FIMs can be obtained as equation (62) <ref:2604.00572#pg2>. The partial derivatives with respect to the auxiliary TPA and RPA position variables are derived through aggregation of derivatives through the conjugate spatial channels <ref:2604. p. 600572#pg2>.

APPENDIX B

CLOSED-FORM EUCLIDEAN GRADIENTS DERIVATIONS

The Euclidean gradients of the communication rates are obtained by applying chain rule to obtain vec(∇Wf) = P F in the set F=xx,xy,yy <ref:2604.00572#pg2>. The Euclidean gradients of the CRLB with respect to the auxiliary TPA and RPA position variables are derived through aggregating derivatives through the conjugate spatial channels <ref:2604. p. 600572#pg2>.

APPENDIX C

COMPUTATIONAL COMPLEXITY ANALYSIS

The computational complexity of the proposed RMO method summarized in Algorithm 2 is estimated as O IA2Imax(M3NK2T + K3TM2 + K2CM + S(M(KC + KT + N))) <ref:2604.00572#pg2>.

APPENDIX D

SUMMARY OF THE RMO METHOD

The overall RMO method for solving (39) is summarized in Algorithm 2 <ref:2604.00572#pg2>. The line-search strategy in (51) guarantees the monotonic descent of Algorithm 2 <ref:2604. p. 700572#pg2>.

APPENDIX E

SIMULATION RESULTS

Figure 3 shows the achieved CRLB versus the transmit power Pt of either the CUs or targets <ref:2604.00572#pg2>. Figure 8 shows the achieved CRLB of each scheme versus the target position error factor ν <ref:2604. p. 9 p. 900572#pg2>. The results demonstrate that schemes with optimized TPA and RPA positions are able to alleviate this performance loss <ref:2604. p. 900572#pg2>. Figure 8 shows the achieved CRLB of each scheme versus the target position error factor ν <ref:2604.00572#pg2>. The proposed SWAN scheme consistently outperforms the multiwaveguide-enabled PASS scheme across different values of the target position error factor <ref:2604. p. 900572#pg2>.

APPENDIX F

REFERENCES

[1] Z. Wang et al., “A tutorial on extremely large-scale MIMO for 6G: Fundamentals, signal processing, and applications,” IEEE Commun. p. 0 Surv. Tutor., vol. 26, no. 3, pp. 1560–1605, 2024 <ref:2604. p. 000572#pg2>.

[8] Z. Ding, R. Schober, and H. Vincent Poor, “Flexible-antenna systems: A pinching-antenna perspective,” IEEE Trans. p. 0 Commun., vol. 73, no. 10, pp. 9236–9253, 2025 <ref:2604. p. 000572#pg2>.

[18] C. Ouyang, H. Jiang, Z. Wang, Y. Liu, and Z. Ding, “Uplink and downlink communications in segmented waveguide-enabled pinchingantenna systems (SWANs),” IEEE Trans. p. 0 Commun., vol. 74, pp. 3688–3703, 2026 <ref:2604. p. 000572#pg2>.

[25] N. Boumal, An Introduction to Optimization on Smooth Manifolds <ref:2604. p. 400572#pg2>.

[31] W. Huang, K. A. Gallivan, and P.-A. Absil, “A Broyden class of quasi-Newton methods for Riemannian optimization,” SIAM Journal on Optimization, vol. p. 5 25, no. 3, pp. 1660–1685 <ref:2604. p. 000572#pg2>.

[32] N. Boumal, P.-A. Absil, and C. Cartis, “Global rates of convergence for nonconvex optimization on manifolds,” IMA J. p. 4 Numer. Anal., vol. 39, no. 1, pp. 1–33 <ref:2604. p. 000572#pg2>.

[27] C Liu and N Boumal, “Simple algorithms for optimization on Riemannian manifolds with constraints,” Appl Math Optim. p. 4, vol. 82, no. 3, pp. 949–981 <ref:2604. p. 000572#pg2>.

--- Page 1 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng, Tee Hiang Cheng, Kah Chan Teh, Senior Member, IEEE, and Zhiguo Ding, Fellow, IEEE <ref:2604.00572#pg2>.

--- Page 2 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 3 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 4 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 5 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 6 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 7 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 8 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 9 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 10 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

--- Page 11 ---

CRLB Minimization for ISAC Systems with Segmented Waveguide–Enabled Pinching Antenna

Yue Geng et al. <ref:2604.00572#pg2>.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Optimal joint optimization of beamforming matrix and pinching antenna positions can be achieved by solving a Riemannian product manifold optimization problem, which allows for joint beamforming and PA position optimization to exploit the coupling among variables. This enables the system to achieve superior performance to both conventional MIMO and mult-waveguide-enabled PASS in the considered ISAC system.

  2. The system can perform integrated sensing and communication by jointly optimizing dual-functional signals, where the dual-functional signal transmitted by the BS are radiated by the SWAN, and the echo signals reflected by the targets are captured by the SWAN and relayed to the BS for estimating the locations of the targets.

  3. The AI system can achieve superior target localization accuracy in multi-user multi-target scenarios because the proposed SWAN scheme achieves better CRLB performance than existing multiwaveguide-enabled PASS schemes and outperforms them across various system parameter settings.

  4. The system can maintain high communication rates while ensuring sensing requirements are met by minimizing the Cramer Rao lower bound, as the objective is formulated as min W,ψ,ϕ CRLBξ s.t. Rk ≥ Γk.

Sources

Related papers