Spatial Reuse in Multi-Cell Pinching-Antenna Systems

arXiv:2611.07906 · eess.SP · Submitted 2026-11-05 · 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: "Spatial Reuse in Multi-Cell Pinching-Antenna Systems".

Dev: Here's the gist of the paper. Spatial reuse in multi-cell pinching-antenna systems is investigated to determine how flexible antenna placement can reduce required resource blocks by mitigating inter-cell interference The spatial-reuse capability is thus characterized by the minimum number of RBs required to satisfy all users’ signal-to-interference-plus-noise ratio (SINR) targets under per-cell power budgets System Model and Channel Characterization The system considers K adjacent rectangular cells, each equipped with an independent dielectric waveguide extending along the x-direction over its cell length The channel power gain is approximated as h LoS k',k(sk')2 ≈ η q k',k + (sk' − ξ k,k)2, where q k',k, (y U k − y C k')2, and H > 0 > The received SINR at user k ∈ A is given by Γk = p kgk(sk) / P sum pk'gk'(sk') + σ2, where σ2 > 0 is the receiver noise power Two-Cell Case Analysis The two-cell case involves transforming the RB-reuse feasibility problem into the minimization of the maximum required transmit power through PA placement The minimum peak transmit power is given by Q⋆ 1,2 = min x∈

l1,u1: , y∈

l2,u2: ∆(x,y)>0 max(p1(x, y), p2(x, y) > The exact algebraic solution involves characterizing stationary candidates in the interior and boundary regions of the PA-position box Multi-Cell Optimization Algorithm For general multi-cell networks, a low-complexity algorithm for joint RB allocation, PA placement, and power control is developed This involves constructing a weighted compatibility graph based on the optimal pairwise peak power Q⋆ i,j derived in Section III-B The algorithm progressively constructs reuse groups by selecting minimum-weight feasible pairs and admitting additional cells through PA-placement refinement and power-feasibility checks Performance Comparison and Scaling Laws Numerical results validate that PASS significantly reduces the average RB requirement compared to conventional fixed antenna systems in both two-cell and multi-cell networks For large cell dimensions, the average number of required RBs under all three PA-placement strategies approaches one, while the conventional system requires substantially more RBs The asymptotic scaling laws show that E

Ts: = 1 + d2s2/r2 for sufficiently large r, where s ∈ S = N, I, P The analysis demonstrates that nearest-PA placement achieves the same asymptotic decay order as power-optimal placement for the excess RB requirement in the two-cell case The enhanced spatial reuse offered by PASS could enable more neighboring cells to simultaneously utilize the same physical resource blocks (PRBs), thus improving spectrum utilization and reducing the need for orthogonal resource allocation across cells Future work will investigate how PA placement can simplify intercell coordination and resource scheduling in practical NR networks Appendix Details The proof of Lemma 3 shows that the maximum value of J(x, d) over

0, xI(d): is zero, implying limd↓dF J(d) = 0 The proof of Lemma 4 establishes that for sufficiently large r, the average number of required RBs is E

Ts: = 1 + FD(Hds) = 1 + d2s2/r2 The comparison shows that lim inf r→∞ (E

TC: − E

Ts: ) ≥ ε2γ = √γ − 1/2(√γ + 1)2 > 0 Numerical Validation The numerical results show that all three PASS strategies significantly reduce the excess RB requirement compared to the conventional fixed-antenna system in both the two-dimensional and multi-cell cases The proposed method achieves slightly lower RB requirements than nearest-PA placement, although their performances remain close The probability Pr(Ts = 1) increases with Dx for both PASS schemes, whereas it remains close to zero for the conventional fixed-antenna system over the considered configurations Reference List

1: L.

Rosa: Zhu, W.

Dev: Ma, and R.

Rosa: Zhang, “Movable antennas for wireless communication: Opportunities and challenges,” IEEE Commun.

Title and authors: Rosa: So, diving into the specifics of "Spatial Reuse in Multi-Cell Pinching-Antenna Systems," we’re looking at Zhenqiao Cheng, Chongjun Ouyang, and Zhiguo Ding.

Dev: Yeah, they focus on how flexible antenna placement improves desired link gain and lowers transmit power to reduce interference across cells, which is a big deal when you are dealing with many small cells.

Taro: When you talk about spatial reuse here in the context of pinching antennas, what’s the main hurdle they’re trying to get over? Is it just managing the interference between adjacent cells?

Rosa: Exactly. Because simultaneous transmission on one resource block becomes infeasible due to inter-cell interference, and they need to figure out how many blocks are required for all users under those per-cell power budgets.

Dev: They characterize this capability by finding that minimum number of resource blocks needed to satisfy everyone’s signal-to-interference-plus-noise ratio targets while respecting the individual transmit power budget for every cell.

Taro: That sounds like a tough constraint set, especially when you have to move the antenna around dynamically across multiple cells.

Rosa: That’s what they tackle by formulating this joint optimization problem—RB allocation, placement, and power control—to find that minimum number of resource blocks required.

The paper's summary: Dev: So, to break down what the paper actually says in its summary section, they start by focusing on the two-cell case as a starting point for their analysis.

Rosa: In that two-cell scenario, they transform the problem of whether those resource blocks can be reused into a minimization problem focused on finding the maximum required transmit power.

Taro: That sounds like they’re trying to find the tightest constraint on how powerful those antennas need to be to even think about reusing a block between just two cells.

Dev: That’s right, they define this minimum peak transmit power as something that depends on the placement of the antennas, which is what they call Q star one two.

Rosa: And they characterize this globally optimal antenna placement by looking at interior stationary candidates and also checking the boundary regions of that antenna position box.

Taro: So, when you talk about those candidates in the interior versus the boundaries, are you talking about finding where the math gets easiest versus where it might fail?

Dev: Precisely. The algebraic solution involves characterizing these stationary points to find the global minimum peak transmit power Q star one two over that box defined by cell lengths and user positions.

Rosa: And this leads directly into their main contribution: they investigate spatial reuse in multi-cell PASS, where each cell uses a dielectric waveguide with one activated antenna to serve a single user.

The paper's improvements: Taro: I see they’re moving from just two cells to more complex scenarios, and I want to know what the main practical benefit of this pinching antenna approach is compared to fixed antennas.

Rosa: The paper points out that flexible PA placement helps reduce the transmit power needed for a user's link, which in turn helps mitigate inter-cell interference between those adjacent cells.

Dev: By reducing that required transmit power, you alleviate the interference problem, and the antenna placement also reshapes the interference channels between cells.

Taro: So it’s like you can steer your antenna so it doesn't blast its neighbor with signal when it’s trying to serve its own user.

Rosa: That’s a good way to think about it; the paper shows this flexibility enables more cells to share the same resource block, potentially cutting down that total resource block requirement.

Dev: They also developed a low-complexity algorithm for general multi-cell networks that uses a weighted compatibility graph based on those optimal pairwise peak powers they found earlier.

Taro: Is this algorithm practical? How complex is it to run in real time when you’re trying to make these decisions while the network is actually operating?

Rosa: The complexity of that group-wise refinement is O I max Q n to the power four, where I max denotes the maximum number of iterations.

Conclusion: Dev: So, wrapping up on "Spatial Reuse in Multi-Cell Pinching-Antenna Systems," they demonstrate that PASS significantly reduces the average resource block requirement compared to conventional fixed antenna systems in both two-cell and multi-cell settings.

Rosa: They show that for large cell dimensions, the average number of required resource blocks under all three pinching antenna placement strategies approaches one, whereas the conventional system needs substantially more blocks.

Taro: I just want to weigh in on the scaling laws they presented. They found an asymptotic scaling law where E T s equals one plus d squared s squared over r squared for sufficiently large r, where s is N I or P.

Dev: That scaling law shows that nearest-PA placement achieves the same asymptotic decay order as power-optimal placement for the excess resource block requirement in the two-cell case.

Rosa: The implication is that this enhanced spatial reuse offered by PASS could allow more neighboring cells to use the same physical resource blocks, which would improve spectrum utilization and reduce the need for orthogonal allocation across cells.

Taro: If you look at their numerical validation figures, they show that while all three PASS strategies reduce the excess resource block requirement compared to fixed antennas, there’s a slightly lower requirement achieved by nearest-PA placement than by feasibility-optimal placement.

Dev: They also noted that the probability of needing only one resource block increases as the cell dimension d x increases for both PASS schemes, while it stays close to zero for the conventional fixed antenna system.

Rosa: So, in short, they prove that nearest-PA placement achieves a similar asymptotic decay order to power-optimal placement for excess resource blocks in two-cell systems with this paper on Spatial Reuse in Multi-Cell Pinching-Antenna Systems.

Zhenqiao Cheng, Chongjun Ouyang, Zhiguo Ding

eess.SP

Submitted: 2026-11-05

Updated: 2026-11-05

Comments: 13 pages

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 71/100

The gist: Spatial reuse in multi-cell pinching-antenna systems is investigated to determine how flexible antenna placement can reduce required resource blocks by mitigating inter-cell interference The…

Key concepts

Spatial Reuse
This is the capability of allowing multiple cells to use the same physical resource blocks (PRBs) simultaneously without causing excessive interference. The paper aims to maximize this by intelligently placing pinching antennas.
Pinching Antenna Systems (PASS)
PASS uses a dielectric waveguide as an antenna, offering spatial flexibility. By moving the antenna's position, the system can improve link gain and reduce transmit power needed to meet signal quality targets, thereby mitigating inter-cell interference.
Resource Blocks (RBs)
RBs are discrete units of spectrum allocated for communication. The goal of the optimization is to find the minimum number of these RBs necessary to ensure all users achieve their required signal-to-interference-plus-noise ratio (SINR) targets under strict power limits.
Peak Transmit Power ($Q^ ext{⋆}_{1,2}$)
This represents the minimum transmit power needed from two adjacent cells to satisfy their respective SINR requirements when they are allowed to reuse the same resource block. Finding this minimum power is crucial for determining if interference can be managed effectively.

Terminology

Summary

Spatial reuse in multi-cell pinching-antenna systems is investigated to determine how flexible antenna placement can reduce required resource blocks by mitigating inter-cell interference The spatial-reuse capability is thus characterized by the minimum number of RBs required to satisfy all users’ signal-to-interference-plus-noise ratio (SINR) targets under per-cell power budgets

System Model and Channel Characterization

The system considers K adjacent rectangular cells, each equipped with an independent dielectric waveguide extending along the x-direction over its cell length The channel power gain is approximated as h LoS k',k(sk')2 ≈ η q k',k + (sk' − ξ k,k)2, where q k',k, (y U k − y C k')2, and H > 0 > The received SINR at user k ∈ A is given by Γk = p kgk(sk) / P sum pk'gk'(sk') + σ2, where σ2 > 0 is the receiver noise power

Two-Cell Case Analysis

The two-cell case involves transforming the RB-reuse feasibility problem into the minimization of the maximum required transmit power through PA placement The minimum peak transmit power is given by Q⋆ 1,2 = min x∈[l1,u1], y∈[l2,u2] ∆(x,y)>0 max(p1(x, y), p2(x, y) > The exact algebraic solution involves characterizing stationary candidates in the interior and boundary regions of the PA-position box

Multi-Cell Optimization Algorithm

For general multi-cell networks, a low-complexity algorithm for joint RB allocation, PA placement, and power control is developed This involves constructing a weighted compatibility graph based on the optimal pairwise peak power Q⋆ i,j derived in Section III-B The algorithm progressively constructs reuse groups by selecting minimum-weight feasible pairs and admitting additional cells through PA-placement refinement and power-feasibility checks

Performance Comparison and Scaling Laws

Numerical results validate that PASS significantly reduces the average RB requirement compared to conventional fixed antenna systems in both two-cell and multi-cell networks For large cell dimensions, the average number of required RBs under all three PA-placement strategies approaches one, while the conventional system requires substantially more RBs The asymptotic scaling laws show that E[Ts] = 1 + d2s2/r2 for sufficiently large r, where s ∈ S = N, I, P

Conclusion on Spatial Reuse

The analysis demonstrates that nearest-PA placement achieves the same asymptotic decay order as power-optimal placement for the excess RB requirement in the two-cell case The enhanced spatial reuse offered by PASS could enable more neighboring cells to simultaneously utilize the same physical resource blocks (PRBs), thus improving spectrum utilization and reducing the need for orthogonal resource allocation across cells Future work will investigate how PA placement can simplify intercell coordination and resource scheduling in practical NR networks

Appendix Details

The proof of Lemma 3 shows that the maximum value of J(x, d) over [0, xI(d)] is zero, implying limd↓dF J(d) = 0 The proof of Lemma 4 establishes that for sufficiently large r, the average number of required RBs is E[Ts] = 1 + FD(Hds) = 1 + d2s2/r2 The comparison shows that lim inf r→∞ (E[TC] − E[Ts]) ≥ ε2γ = √γ − 1/2(√γ + 1)2 > 0

Numerical Validation

The numerical results show that all three PASS strategies significantly reduce the excess RB requirement compared to the conventional fixed-antenna system in both the two-dimensional and multi-cell cases The proposed method achieves slightly lower RB requirements than nearest-PA placement, although their performances remain close The probability Pr(Ts = 1) increases with Dx for both PASS schemes, whereas it remains close to zero for the conventional fixed-antenna system over the considered configurations

Reference List

[1] L. Zhu, W. Ma, and R. Zhang, “Movable antennas for wireless communication: Opportunities and challenges,” IEEE Commun. p. 0

Where the paper says it

INTRODUCTION Reconfigurable antennas (RAs) have attracted growing attention due to their ability to adapt wireless channels to different quality-of-service (QoS) requirements. Representative examples include fluid/movable antennas [1], [2], reconfigurable intelligent surfaces [3], and dynamic metasurface antennas [4].

Mag., vol. 62, no. 6, pp. 114–120, Jun. p. 9
Where the paper says it

26: end while 27: return T, {At} T t=1, and the associated PA positions and powers. 20 40 60 80 100 120 140 Cell Length Dx [m] 10-3 10-2 10-1 100 E[T s]-1 Conventional PASS (Nearest) PASS (Feasibility) PASS (Optimal) Analytical Asymptotic 40 0.

2024

[5] A. Fukuda, H. Yamamoto, H. Okazaki, Y. Suzuki, and K. Kawai, “Pinching antenna: Using a dielectric waveguide as an antenna,” NTT DOCOMO Technical J. p. 0

Where the paper says it

To address these limitations, NTT DOCOMO introduced the concept of pinching-antenna system (PASS) [5]. In PASS, radio-frequency signals propagate through a dielectric waveguide, while dielectric pinching antennas (PAs) attached to the waveguide couple energy between the guided wave and free space [6].

, vol. 23, no. 3, pp. 5–12, Jan. 2022

[6] Y. Liu, Z. Wang, X. Mu, C. Ouyang, X Xu, and Z Ding, “Pinching antenna systems (PASS): Architecture designs, opportunities, and outlook,” IEEE Commun. p. 0

Where the paper says it

1 Spatial Reuse in Multi-Cell Pinching-Antenna Systems Zhenqiao Cheng, Chongjun Ouyang, and Zhiguo Ding, Fellow, IEEE Abstract—In this paper, spatial reuse in a multi-cell pinchingantenna system (PASS) is investigated, where each cell employs a dielectric waveguide with one activated pinching antenna (PA) to serve a single user. Spatial reuse refers to the simultaneous use of the same time-frequency resource block (RB) by multiple spatially separated cells.

Mag., vol. 64, no. 1, pp. 190–196

[8] Y. Liu, H Jiang, X Xu, Z Wang, J Guo, C Ouyang, X Mu, Z Ding et al. p. 0

Where the paper says it

1 Spatial Reuse in Multi-Cell Pinching-Antenna Systems Zhenqiao Cheng, Chongjun Ouyang, and Zhiguo Ding, Fellow, IEEE Abstract—In this paper, spatial reuse in a multi-cell pinchingantenna system (PASS) is investigated, where each cell employs a dielectric waveguide with one activated pinching antenna (PA) to serve a single user. Spatial reuse refers to the simultaneous use of the same time-frequency resource block (RB) by multiple spatially separated cells.

, “Pinching-antenna systems (PASS): A tutorial,” IEEE Trans. p. 0
Where the paper says it

1 Spatial Reuse in Multi-Cell Pinching-Antenna Systems Zhenqiao Cheng, Chongjun Ouyang, and Zhiguo Ding, Fellow, IEEE Abstract—In this paper, spatial reuse in a multi-cell pinchingantenna system (PASS) is investigated, where each cell employs a dielectric waveguide with one activated pinching antenna (PA) to serve a single user. Spatial reuse refers to the simultaneous use of the same time-frequency resource block (RB) by multiple spatially separated cells.

Commun., vol. 74, pp. 4881–4918

[23] D. A Cox, J. Little, and D. O’shea, Using Algebraic Geometry, 2nd ed. p. 5

Where the paper says it

Following Section III-C2, we assume xI ≤ min{x¯1, x¯2} and focus on γ > 1 and d > dF. The cases with more restrictive PA-position constraints can be addressed using the exact algebraic solution in Section III-B.

, New York, NY, USA: Springer, 2005

Page Breakdown

--- Page 1 ---

Spatial reuse in multi-cell pinching-antenna systems is investigated to determine how flexible antenna placement can reduce required resource blocks by mitigating inter-cell interference The spatial-reuse capability is thus characterized by the minimum number of RBs required to satisfy all users’ signal-to-interference-plus-noise ratio (SINR) targets under per-cell power budgets

--- Page 2 ---

The system model considers K adjacent rectangular cells, each equipped with an independent dielectric waveguide extending along the x-direction over its cell length The channel power gain is approximated as h LoS k',k(sk')2 ≈ η q k',k + (sk' − ξ k,k)2, where q k',k, (y U k − y C k')2, and H > 0 > The received SINR at user k ∈ A is given by Γk = p kgk(sk) / P sum pk'gk'(sk') + σ2, where σ2 > 0 is the receiver noise power

--- Page 3 ---

PASS provides additional spatial flexibility to alleviate this limitation by placing the PA close to its intended user, which can improve its desired-link gain and reduce the transmit power required to satisfy the SINR target The reduced transmit power can alleviate ICI while PA placement can also reshape the interference channels between cells

--- Page 4 ---

The two-cell case involves expressing PA positions relative to the projections of their intended users onto the respective waveguide axes The minimum required transmit power is given by Q⋆ 1,2 = min x∈[l1,u1], y∈[l2,u2] ∆(x,y)>0 max(p1(x, y), p2(x, y) >

--- Page 5 ---

The problem formulation couples PA placement and power control with combinatorial RB allocation The objective is to minimize the number of orthogonal RBs required to serve all K users

--- Page 6 ---

The problem formulation couples PA placement and power control with combinatorial RB allocation To characterize the impact of PA placement on spatial reuse, we first investigate the two-cell case and derive analytical results for the minimum required transmit power

--- Page 7 ---

A finite nonnegative power solution exists if and only if ∆(x, y) > 0 ⇐⇒ γ2a(x)e(y) / b(x)c(y) < 1 Consequently, the two cells can reuse the same RB if and only if Q⋆ 1,2 ≤ Pmax

--- Page 8 ---

The global solution is obtained as Q⋆ 1,2 = min (x,y)∈V max(p1(x, y), p2(x, y) >

--- Page 9 ---

The overall procedure is summarized in Algorithm 1 The complexity of the group-wise refinement is O(ImaxQn4), where Imax denotes the maximum number of iterations

--- Page 10 ---

The overall procedure is summarized in Algorithm 1 The complexity of the group-wise refinement is O(ImaxQn4), where Imax denotes the maximum number of iterations

--- Page 11 ---

Fig. 5 shows that the average RB requirement versus Dx for different numbers of cells PASS significantly outperforms the conventional fixed-antenna system across all considered network sizes

--- Page 12 ---

Fig. 4(a) illustrates the average RB requirement versus the cell length Dx All three PASS strategies significantly outperform the conventional fixed-antenna system, and their average RB requirements decrease toward one as Dx increases

--- Page 13 ---

Fig. 4(b) presents the average RB requirement versus Rth The RB requirements of all four schemes increase with Rth, but PASS consistently outperforms the conventional fixed-antenna system throughout the considered range

--- Page 14 ---

Fig. 3(a) compares the conventional fixed-antenna system with PASS under nearest-PA, feasibility-optimal, and poweroptimal PA placement All three PASS strategies significantly reduce the excess RB requirement compared to the conventional fixed-antenna system

--- Page 15 ---

Fig. 4(b) presents the average RB requirement versus Rth The RB requirements of all four schemes increase with Rth, but PASS consistently outperforms the conventional fixed-antenna system throughout the considered range

--- Page 16 ---

Fig. 5(a) compares the average RB requirements of the three schemes versus Dx for different numbers of cells PASS significantly outperforms the conventional fixed-antenna system across all considered network sizes

--- Page 17 ---

Fig. 5(b) shows that the one-RB reuse probability decreases with Rth Nevertheless, PASS achieves a substantially higher probability than the conventional system, especially at moderate target rates

--- Page 18 ---

Fig.

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. Spatial Reuse Capability Characterization: The AI system can determine the minimum number of RBs required to satisfy all users’ signal-to-interference-plus-noise ratio (SINR) targets under per-cell power budgets by solving the joint optimization problem of RB allocation, PA placement, and power control.

  2. Global Optimization for Two Cells: The system can determine the globally optimal PA placement for two cells by minimizing the peak transmit power required by the two cells, which is achieved through characterizing interior stationary points and boundary candidates of a Karush-Kuhn-Tucker (KKT) formulation.

  3. Multi-Cell Resource Management: For general multi-cell networks, the system can employ a greedy RB allocation algorithm with minimum-weight pair initialization and element-wise PA placement optimization to construct reuse groups based on a weighted compatibility graph where edge weights represent the minimum peak transmit powers required for feasible pairwise RB reuse.

  4. Power-Optimal Placement Selection: The AI can select the optimal PA placement by maximizing the function J(x, d) over the feasible interval [0, xI], which is derived from minimizing the required transmit power p1 = p2 = p0 J(x, d).

  5. Asymptotic RB Requirement Prediction: For large cell dimensions (Dx/H), the system can predict the average number of required RBs using scaling laws such as E[Ts] = 1 + d squared s squared / (2r 2) for nearest-PA placement, feasibility-optimal placement, and power-optimal placement.

  6. Performance Comparison: The AI can compare the performance of PASS strategies against a conventional fixed-antenna system by quantifying the strictly positive asymptotic RB gap between PASS and the conventional fixed-antenna system as cell coverage expands.

Abstract

In this paper, spatial reuse in a multi-cell pinching-antenna system (PASS) is investigated, where each cell employs a dielectric waveguide with one activated pinching antenna (PA) to serve a single user. Spatial reuse refers to the simultaneous use of the same time-frequency resource block (RB) by multiple spatially separated cells. Due to inter-cell interference (ICI), simultaneous transmission by all cells on the same RB may be infeasible, and multiple RBs may be required to serve all users. The spatial-reuse capability is thus characterized by the minimum number of RBs required to satisfy all users' signal-to-interference-plus-noise ratio (SINR) targets under per-cell power budgets. Flexible PA placement can enhance the desired-link gain of each user and reduce the required transmit power of its serving cell, which in turn mitigates ICI and allows more cells to share the same RB. To exploit these benefits, a joint optimization problem of RB allocation, PA placement, and power control is formulated. For the two-cell case, the globally optimal PA placement is determined by minimizing the peak transmit power required by the two cells. A one-dimensional special case is further analyzed to derive analytical expressions for the average RB requirement and its scaling laws with respect to cell length under different PA-placement strategies. For the general multi-cell case, the optimal two-cell solution is exploited to construct a weighted compatibility graph, whose edge weights represent the minimum peak transmit powers required for feasible pairwise RB reuse. Based on this graph, a greedy RB allocation algorithm with minimum-weight pair initialization and element-wise PA-placement optimization is developed. Numerical results validate the analytical findings and demonstrate that PASS significantly reduces the required RBs compared with conventional fixed-antenna systems.

Sources

Related papers