Intra-Cell Spatial Reuse in LEO Downlinks: Beam-Leakage-Aware Satellite Scheduling

arXiv:2611.06726 · eess.SP · Submitted 2026-11-04 · 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: "Intra-Cell Spatial Reuse in LEO Downlinks".

Dev: Here's the gist of the paper. Dense low-earth-orbit (LEO) constellations offer an opportunity to increase local broadband capacity by serving multiple terminals in the same cell through concurrent satellite beams.

Rosa: This paper proposes beam-leakage-aware satellite scheduling for Ku-band LEO downlink transmissions.

Dev: Problem Statement and Formulation The core problem is to select a prescribed number of satellites at each scheduling interval to maximize the instantaneous cell sum rate, while accounting for link strength and inter-layer beam leakage.

Rosa: The system model focuses on intervals where geometry and large-scale channel states are fixed, aiming to maximize the cell sum rate by jointly accounting for received powers and pairwise leakage coefficients.

Paper summary: Rosa: So, wrapping up this paper called "Intra-Cell Spatial Reuse in LEO Downlinks: Beam-Leakage-Aware Satellite Scheduling", the main point is that just picking satellites based on simple distance or link strength isn't good enough for dense constellations.

Dev: They show that you need to jointly account for satellite dependent received powers and pairwise receive array leakage if you want to maximize the cell sum rate properly, which is what this paper proposes.

Taro: The implication for us is that when we are designing these massive LEO networks, we can't just rely on simple greedy selection; we have to factor in how those beams actually interfere with each other when trying to reuse the same frequency space.

Rosa: It means that for increasing capacity in these dense cells, you get less than you might expect just by adding more beams. The BL BnB method gives a concrete way to find that better balance between using more satellites and managing the interference.

Dev: And because they managed to do this with millisecond scale runtimes, it suggests that implementing this kind of careful scheduling for real-time network control is actually achievable in practice within our current constraints.

Taro: It supports the idea that we should be considering link quality and interference together when we think about spatial reuse in these systems.

Conclusion: Rosa: So we're wrapping up this look at "Intra-Cell Spatial Reuse in LEO Downlinks: Beam-Leakage-Aware Satellite Scheduling." Basically, these authors are showing how to pick the right satellites for a cell so they don't just overlap and cause interference.

Dev: Right. They’re focusing on this specific problem where you have lots of satellites in a dense low earth orbit, and if you just look at distance or link strength, it doesn't work because of the beam leakage between them.

Taro: So they’re building this scheduling method that takes into account both how strong the signal is and how much one satellite's beam leaks into another.

Rosa: Exactly. They use a branch and bound algorithm to find these schedules that maximize the total data capacity within that cell, making sure they’re not just picking satellites randomly.

Dev: The big result they show is that with eight beams running at once, this method improves the mean per-beam capacity by about twenty-seven point two percent compared to just picking your nearest satellite.

Taro: That improvement comes mainly from helping those weaker links get better service, which is a practical thing for real network design.

Rosa: It shows that increasing the beam count helps total capacity, but not in a straight line; you get less of an increase as you add more beams.

Dev: The authors also show how fast this works, and they hit millisecond scale runtimes when they’re testing it against their starlink visibility traces. That means it's computationally feasible for real time decisions.

Taro: So, the core idea is that for spatial reuse in these dense LEO systems, you have to look at both the link quality and the interference at the same time when you decide which satellites to use.

Rosa: That’s it for this paper. It’s a good way to think about how we manage these complex aerial networks, and it sets up a lot of new ways we can approach satellite placement.

Shizhen Jia, Hao Guo, Ruoyu Sun, Hans Geerdes, Mark Poletti, Songwei Dong, Lauryn Smith

New York University WIRELESS Department of Wireless Technologies CableLabs

eess.SP

Submitted: 2026-11-04

Updated: 2026-11-04

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

Importance score: 90/100

The gist: Dense low-earth-orbit (LEO) constellations offer an opportunity to increase local broadband capacity by serving multiple terminals in the same cell through concurrent satellite beams.

Key concepts

Beam Leakage
This refers to the interference that occurs when multiple satellite beams overlap in space. When a satellite transmits, its signal spreads beyond its intended area, causing unwanted interference to other nearby beams. The scheduling method considers this leakage to ensure better spatial reuse and capacity.
Sum-Rate Maximization
This is the goal of the optimization problem: finding the best way to select satellites for a given time slot so that the total data rate (sum rate) received by all users in a cell is as high as possible. This involves balancing link strength and interference constraints.
Branch-and-Bound (BL-BnB)
This is an optimization algorithm used to find the absolute best schedule. It systematically explores possible satellite combinations, using an estimated upper limit (bound) to quickly discard schedules that cannot be better than the best one found so far. This ensures a globally optimal solution is found.
Reference-Point Model
This is a simplified system model used for the optimization process. It fixes the geometry and large-scale channel states for specific time intervals, allowing researchers to focus on how satellite selection affects capacity under these controlled conditions.

Terminology

Summary

Dense low-earth-orbit (LEO) constellations offer an opportunity to increase local broadband capacity by serving multiple terminals in the same cell through concurrent satellite beams. p. 0 This paper proposes beam-leakage-aware satellite scheduling for Ku-band LEO downlink transmissions. p. 0

Problem Statement and Formulation

The core problem is to select a prescribed number of satellites at each scheduling interval to maximize the instantaneous cell sum rate, while accounting for link strength and inter-layer beam leakage. The system model focuses on intervals where geometry and large-scale channel states are fixed, aiming to maximize the cell sum rate by jointly accounting for received powers and pairwise leakage coefficients. p. 1

Optimization Methodology

The authors employ a beam-leakage-aware branch-and-bound (BL-BnB) algorithm to obtain globally optimal schedules for the referencepoint model. This method combines an admissible upper bound with greedy-plus-swap initialization to certify optimality for the referencepoint model. p. 0 The search starts at (∅, S) and prunes branches when the admissible upper bound does not exceed the best feasible sum rate found so far. p. 2

Performance Evaluation

The method is evaluated using a 24-h ephemerisderived Starlink visibility trace with antenna and transmit-power models calibrated to public SpaceX filings. The results show that total capacity increases with beam count, but less than proportionally. p. 0 With eight concurrent beams, BL-BnB improves mean per-beam capacity by 27. p. 02% over nearest-satellite selection, primarily benefiting weaker links. p. 0

Key Findings

The study quantifies the effects of beam count, cell radius, and scheduling on intra-cell performance. Mean per-beam capacities show that increasing the number of paired beams raises total capacity sublinearly. p. 5 For selecting eight of 28 candidates, BL-BnB reduces mean runtime from 529. p. 548 to 1.98 ms. The results support jointly considering link quality and interference for satellite spatial reuse. p. 5 p. 5

Complexity and Practicality

The complexity analysis shows that the overall complexity is O(N 2Nr + (1 + Isw)NK cubed + MBB N squared log N). BL-BnB remains exponential in the worst case as MBB may approach exhaustive enumeration, while admissible pruning reduces practical search cost while preserving global optimality. p. 3 The mean solver time for selecting K=8 of N candidates falls from 529. p. 348 to 1.98 ms when using BL-BnB. p. 0 The results support jointly considering link quality and interference for satellite spatial reuse.

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. Beam-leakage-aware scheduling formulation: The improved system can jointly account for satellitedependent received powers and pairwise receive-array leakage when formulating a fixed-cardinality sumrate maximization problem.

  2. Globally optimal satellite selection: The system can achieve globally optimal schedules for the reference point model by employing a beam-leakage-aware branch-and-bound (BLBnB) algorithm, which combines an admissible upper bound with greedy-plus-swap initialization.

  3. Practical spatial reuse assessment: The system can quantify the effects of beam count and cell radius on capacity, showing that total capacity increases with beam count, but less than proportionally and improving mean per-beam capacity by 27.2% over nearest-satellite selection when using eight concurrent beams.

  4. Optimized scheduling objective: The scheduler can maximize the cell sum rate by solving the nonseparable objective function using a BLBnB search that maintains the best feasible schedule found so far (the incumbent) and prunes branches whose admissible upper bounds do not exceed its sum rate.

  5. Initialization refinement: The system can use an initialization strategy combining greedy–swap set with the top-K and bottom-K received-power sets to obtain a strong starting point for the global search.

Related papers