Joint Air Corridor and BS Coverage Planning for Large-Scale High-Density Low-Altitude Aerial Networks via Channel Knowledge Maps

arXiv:2611.13574 · eess.SP · Submitted 2026-11-10 · 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: "Joint Air Corridor and BS Coverage Planning for Large-Scale High-Density Low-Altitude Aerial Networks via Channel Knowledge Maps".

Dev: Here's the gist of the paper. Joint air corridor and base station (BS) coverage planning for large-scale high-density low-altitude aerial networks via channel knowledge maps studies the joint air corridor and base station (BS) coverage planning problem for high-density lowaltitude aerial networks supported by terrestrial cellular infrastructures The joint design is formulated as a mixed-integer linear program (MILP), which can be optimally solved using off-the-shelf solvers.

Rosa: System Model Description The considered low-altitude aerial network involves unmanned aerial vehicles (UAVs) operating within a structured air corridor planned over a grid-based airspace, and each BS serves selected corridor grid cells by sharing its communication resources among all UAVs therein.

Dev: The airspace is discretized into an N ×N arrangement of cuboidal grid cells with dimensions ∆x, ∆y, and ∆z along the x, y, and z directions.

Rosa: An air corridor connecting two points is represented by a sequence of face-adjacent grid cells according to the four-neighbor definition.

Title and authors: Dev: So we've established the setup—the MILP formulation and the channel knowledge map—and now let's look at what the paper actually summarizes about this problem.

Rosa: They start by describing the system model again, explaining that it’s a structured airspace, an N times N arrangement of cuboidal grid cells defined by dimensions Delta X, Delta Y, and Delta Z.

Dev: And they explain how the air corridor is represented as a sequence of face-adjacent grid cells following a four-neighbor definition between a departure and destination point.

Rosa: The key takeaway from their summary is that each base station serves selected corridor grid cells by sharing its communication resources among all the UAVs within those specific cells.

Dev: And they define the channel knowledge map, or CKM, as the tool used to characterize the channel gains between base stations and candidate airspace cells by exploiting prior knowledge of the local propagation environment.

Rosa: They then define spectral efficiency from a base station k to a grid cell i, j as being log2 one plus gamma k i j.

Dev: And they frame the whole joint optimization problem as minimizing a weighted objective that balances corridor length and the supportable average UAV density along the entire corridor.

Rosa: They emphasize that this minimization happens subject to these minimum communication rate requirements for all UAVs, which is what drives the need for a feasible solution.

Dev: So, in simple terms, they are taking this complex physical layout problem and turning it into a mathematical optimization task where the goal is to find the best spatial arrangement of corridors and base station coverage that satisfies all communication needs.

Taro: That sounds like they are trying to solve a spatial configuration puzzle under strict communication constraints.

Rosa: Precisely. The paper summarizes that this specific problem—joint planning of air corridors and BS coverage for supporting dense UAV traffic—is less explored compared to the existing work focusing on optimizing trajectories or just optimizing communication performance along a given corridor or specific mission.

Dev: So they are highlighting that the gap in research is exactly where the need for this comprehensive joint planning solution lies.

Taro: I wonder if they mention any specific limitations regarding the assumptions made about that channel knowledge map being perfect?

Rosa: They do mention it, but implicitly by saying they exploit prior knowledge to acquire such information, which implies that the quality of their planning depends heavily on how accurate and complete that CKM is.

Dev: So if the input data—the CKM—is flawed, the resulting optimal corridor and coverage plan might not be good in reality.

Taro: That makes sense. If you rely on a map to do your whole design, you need to be confident in that map’s fidelity for it to work outside of the simulation.

The paper's summary: Rosa: Now let’s talk about how this research improves upon what came before by looking at the specific improvements they suggest in their methodology.

Dev: They move from just optimizing trajectories or communication performance along a corridor to a more holistic approach that optimizes the whole system structure at once.

Rosa: Yes, the main improvement is shifting from single-task optimization to joint planning of air corridors and base station coverage, which addresses that coupling problem directly.

Dev: By formulating it as this MILP, they are making sure you don't just find a path that happens to work; you get a solution where the physical layout and the coverage are intrinsically linked.

Taro: So, they’re not just optimizing one variable anymore; they’s also optimizing the spatial relationship between infrastructure and traffic flow.

Rosa: That's it. They are balancing corridor length against supportable average UAV density while ensuring all minimum communication rates are met within the constraints of resource sharing.

Dev: I think another point is how they introduce that scalable suboptimal algorithm for large maps, which addresses the computational cost issue.

Taro: So, they’re using Lagrangian relaxation and A star search to reduce computation time instead of just brute-forcing the entire problem space.

Rosa: They generate candidate corridors first and then solve a recovery MILP for each corridor, which is a way to manage the complexity without getting bogged down in one giant calculation.

Dev: That’s a practical approach for real-world application because it lets you get good solutions on maps that are too big for the exact solver to handle efficiently.

Taro: I'm interested in how they handle the resource sharing constraint explicitly, since that’s where most of the complexity lies.

Rosa: They incorporate load-based rate estimates, like rk i j equals B u k eta k i j into their optimization constraints and objective function.

Dev: So they aren't ignoring that base station capacity is a major factor; it’s being explicitly used to manage how much traffic can fit in.

Taro: That seems like a necessary step for making the model realistic because without accounting for the actual load on the BS, you wouldn't get an accurate picture of what’s possible.

The paper's improvements: Rosa: So, to wrap up the discussion on this paper "Joint Air Corridor and BS Coverage Planning for Large-Scale High-Density Low-Altitude Aerial Networks via Channel Knowledge Maps," the main implication is that joint planning is essential for maximizing drone traffic in these dense networks.

Dev: The main thing to remember is that you can achieve better results by not treating corridor layout and base station coverage as separate design choices.

Taro: And the fact that they showed their method works across different minimum rate requirements R0 shows it’s a more robust way to approach this problem than just picking one single requirement.

Rosa: They conclude that corridors with the same geometric length can support markedly different UAV densities, which really highlights how crucial it is to jointly plan geometry and coverage under resource sharing.

Dev: So, the big picture here is that when designing these low-altitude networks, you need a coordinated strategy for both path planning and infrastructure placement.

Taro: I think I'm just glad they showed how to make this mathematically tractable for large maps without sacrificing the quality of the solution too much.

Rosa: They did it, and it’s a solid piece of work for anyone designing these kinds of systems.

Conclusion: Rosa: So we've seen how they’re tackling the joint planning of air corridors and base station coverage for those high-density low-altitude networks using channel knowledge maps in this paper, "Joint Air Corridor and BS Coverage Planning for Large-Scale High-Density Low-Altitude Aerial Networks via Channel Knowledge Maps."

Dev: Exactly. The core idea is optimizing the spatial relationship between the infrastructure layout and the traffic density, all while making sure every little drone gets enough communication.

Taro: It’s interesting how they use that channel knowledge map to get those initial gains before they even start planning where everything goes.

Rosa: Right, because it shows that just optimizing a path or just optimizing the coverage separately doesn't give you the best outcome when you have these shared resources.

Dev: They introduce this mixed integer linear program, or MILP, to make sure they’re finding the globally optimal solution for minimizing corridor length against traffic density.

Taro: And for those massive maps, they didn't just try to solve that huge MILP directly because it would take forever.

Rosa: They developed this scalable algorithm based on Lagrangian relaxation and A star search, which lets them generate candidate corridors first before doing the detailed work.

Dev: That means they reduce the computational load significantly by breaking the big problem into smaller, manageable pieces for each corridor they consider.

Taro: It makes sense to relax those rate constraints into resource constraints in that way so you can use A star search to guide your pathfinding across all those grid cells.

Rosa: And when they run the numbers on small maps, they show that this joint approach gives them better results and higher traffic density than just looking at a corridor layout alone.

Dev: They even found that the benefit of jointly planning corridors and coverage keeps working even when you change how much minimum communication rate R0 you require from those UAVs.

Taro: That suggests the physical shape of the corridor itself really matters, not just how long it is in straight line distance.

Rosa: So to wrap up, this paper confirms that coordinating the corridor layout and which base stations serve which cells is key to balancing length and density while respecting resource sharing across all those base stations.

Dev: It’s a strong framework for designing these complex aerial systems because it shows how to handle the interaction between physical geometry and communication capacity.

Taro: I think it sets a good foundation for future work, especially if you want to see how this plays out when the world gets really messy and the channel knowledge map isn't perfect.

Rosa: Exactly, that’s where we need to look next—what happens when the environment misbehaves or you need real-time adjustments.

Dev: Next time, we’ll be looking at how they handle those failure modes in their own work and see what the safety case looks like in practice.

Jiaxuan Li, Yilong Chen, Yuan Guo, Fan Liu, Jie Xu

School of Science and Engineering at The Chinese University of Hong Kong · National Mobile Communications Research Laboratory at Southeast University

eess.SP

Submitted: 2026-11-10

Updated: 2026-11-10

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

Importance score: 94/100

The gist: Joint air corridor and base station (BS) coverage planning for large-scale high-density low-altitude aerial networks via channel knowledge maps studies the joint air corridor and base station (BS)…

Key concepts

Channel Knowledge Map (CKM)
A map that uses prior knowledge about the local radio propagation environment to estimate how well a base station can communicate with specific grid cells. This map helps determine the effective signal strength for service at any given location, which is crucial for planning where BSs should be placed and how corridors should be routed.
Mixed-Integer Linear Program (MILP)
A mathematical optimization framework used to solve the joint planning problem. It involves finding the best combination of discrete decisions (like which cells to cover or which paths to use) and continuous variables (like traffic density) while minimizing a specific objective function, subject to various constraints like minimum data rates.
Lagrangian Relaxation and A* Search
A scalable computational technique used for large maps. It simplifies the complex MILP by relaxing strict rate constraints into simpler resource constraints. The A* search then efficiently explores potential air corridors, guided by costs derived from these relaxed constraints, allowing for fast planning even on very large areas.

Terminology

Summary

Joint air corridor and base station (BS) coverage planning for large-scale high-density low-altitude aerial networks via channel knowledge maps studies the joint air corridor and base station (BS) coverage planning problem for high-density lowaltitude aerial networks supported by terrestrial cellular infrastructures The joint design is formulated as a mixed-integer linear program (MILP), which can be optimally solved using off-the-shelf solvers. p. 0

System Model Description

The considered low-altitude aerial network involves unmanned aerial vehicles (UAVs) operating within a structured air corridor planned over a grid-based airspace, and each BS serves selected corridor grid cells by sharing its communication resources among all UAVs therein. The airspace is discretized into an N ×N arrangement of cuboidal grid cells with dimensions ∆x, ∆y, and ∆z along the x, y, and z directions. p. 2 An air corridor connecting two points is represented by a sequence of face-adjacent grid cells according to the four-neighbor definition. p. 2

Channel Representation and Service Model

The channel knowledge map (CKM) is employed to characterize the channel gains between BSs and candidate airspace cells, exploiting prior knowledge of the local propagation environment. The effective SINR for service by BS k at grid cell (i, j) is defined using locally averaged channel gains from the CGM entries. p. 3 The spectral efficiency from BS k to grid cell (i, j) is defined as log2(1 + γk,i,j). p. 3

Joint Optimization Formulation

The joint planning problem aims to minimize a weighted objective balancing corridor length and the supportable average UAV density along the entire corridor subject to minimum communication rate requirements for all UAVs. The problem is formulated as (P1) which seeks to optimize Ai,j, xk,i,j, u, uk. p. 0 An exact MILP reformulation (P2) is developed by linearizing the products between traffic density and binary coverage variables using auxiliary variables. p. 1

Scalable Algorithm for Large Maps

For large maps, a scalable suboptimal algorithm based on Lagrangian relaxation and A∗ search is developed to reduce computational cost. This involves relaxing the original rate constraints into additive BS resource constraints and using Lagrangian multipliers to incorporate resource consumption into the A∗ search costs. p. 1 The algorithm proceeds by generating candidate corridors and then solving a recovery MILP (P4) for each corridor to jointly optimize BS coverage and traffic density. p. 5

Numerical Results and Performance

Numerical results demonstrate the benefit of jointly planning corridor layouts and BS coverage for increasing the supportable UAV density. The proposed algorithm achieves lower objective values than the considered benchmarks and comparable traffic densities to the full MILP on tested small maps. p. 1 The method consistently achieves the highest traffic density over a range of minimum per-UAV rate requirements R0, showing that the benefit of jointly optimizing the corridor and BS coverage persists under different rate requirements. p. 0 The proposed method maintains competitive solution quality while substantially reducing the computational burden as the problem size increases. p. 9

Conclusion

The paper concludes that the proposed design jointly determines the corridor layout and the corridor cells served by each BS to balance corridor length and supportable UAV density while accounting for BS resource sharing. The results further demonstrated that corridors with the same geometric length can support markedly different UAV densities, highlighting the importance of jointly planning corridor geometry and BS coverage under resource sharing. p. 3

Appendix A Proof of Theorem 1

The proof for Theorem 1 shows that for a globally optimal solution, the grid cells selected by every solution form exactly one connected simple path from the departure to destination cells. This property is ensured by the objective term in (P1) which includes a positive corridor length term, excluding isolated cycles at optimality. p. 4

Appendix B Proof of Theorem 2

The proof for Theorem 2 shows that the best-dual-value obtained after T iterations asymptotically attains the optimal value of the corresponding Lagrangian dual problem. The gap between the best dual value and the optimal value is shown to be O log(T + 1) / sqrt(T) as T approaches infinity. p. 7

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. Scalable Suboptimal Optimization for Large-Scale Maps: The system can efficiently plan air corridors and BS coverage for high-density UAV traffic on large maps by using a Lagrangian relaxation based A star search to generate candidate corridors, significantly reducing computation time compared to solving the full Mixed Integer Linear Program.

  2. Joint Corridor and Coverage Optimization: The improved system can jointly optimize the corridor layout, BS coverage, and normalized aerial traffic density to minimize a weighted objective balancing corridor length and supportable traffic density while ensuring each UAV achieves its required minimum communication rate.

  3. Resource Sharing Awareness: The system explicitly accounts for resource sharing among all UAVs served by the same BS by incorporating load-based rate estimates like rk,i,j = B uk ηk,i,j into the optimization constraints and objective function.

Sources

Related papers