Low-Altitude Wireless Networks: The Next Horizon of Wireless Infrastructure Yuanhao Cui, Jiali Nie, Weijie Yuan, Fan Liu, Ziye Jia, Jie Xu, Zhiyong Feng, Mohamed-Slim Alouini
arXiv:2605.24368v1 [cs.NI] 23 May 2026
Abstract Low-altitude airspace, roughly defined as the region up to 3000 meters above ground level, is envisioned as a new spatial domain for daily human and machine activities. This article introduces the concept of the Low-Altitude Wireless Network (LAWN), which represents a paradigm shift from the current ground-based communication-only network to a three-dimensional (3D) multifunctional network. We analyze the key driving forces, network architecture, and limiting factors of LAWN, with a particular focus on the tight integration of communication, sensing, and control in highly dynamic airspace environments. By establishing the coupling between airspace capacity and wireless channel capacity, we reveal the intrinsic limits of airspace management and identify the fundamental challenges and opportunities associated with its evolution.
The success of wireless networks, as one of the most impactful technological advances in human history, has significantly revolutionized humankind’s daily life and led to worldwide infrastructure expansion. Today, more than 70% of terrestrial areas and 90% of the global population are supported by wireless infrastructure offering uninterrupted communication service, yet these systems generally maintain a ground-based approach in both design and execution [1,2] . This omission reflects a historical focus on human-centric connectivity, overlooking the imminent demand for three-dimensional (3D) wireless services to support emerging activities in low-altitude airspace, e.g., electric vertical takeoff and landing (eVTOL) aircraft and unmanned aerial vehicles (UAVs) [3–5] . However, the widespread deployment of UAVs and the rapid maturation of eVTOL platforms now amplify this urgency. For example, worldwide regulatory agencies are expected to grant more than 100 airworthiness certifications by 2030, paving the way for the commercialization of low-altitude transportation systems [6] . Driven by demand for drone delivery and automated cargo transport, industry projections estimate that the low-altitude logistics market could reach 60 billion dollars by 2035. Consequently, terrestrial cellular networks are not well suited to meet the communication and sensing demands created by the projected growth of low-altitude vehicle traffic over the next decade [7] . Therefore, we are witnessing a paradigm shift in wireless infrastructure from a ground-based communication-only network to a skyward multifunctional network, namely the low-altitude wireless network (LAWN) [8] . Beyond providing reliable communication connectivity, LAWN is envisioned as the
wireless infrastructure underpinning low-altitude air transport through a unified operational framework that combines surveillance, navigation, and traffic orchestration. This transformation would fundamentally redefine cellular networks, shifting them from passive utilities to active enablers of a digitized 3D “skyway”, in which the infrastructure autonomously supports low-altitude transportation safety and efficiency [5,9] . Closing this gap between existing cellular networks and LAWN requires innovation at both the physical and architectural levels. At the physical level, networks require integrated sensing and communication (ISAC) to merge environmental monitoring with data transmission, enabling base stations (BSs) to track aircraft while maintaining connectivity [10] . At the architectural level, interoperability with digitized airspace management platforms is non-negotiable. By integrating ISAC-enabled BSs with air traffic management systems, wireless networks can dynamically map airspace, construct air corridors, enforce collisionavoidance protocols, and automate routing, thereby transforming raw bandwidth into structured mobility governance [11] . LAWN is therefore expected to become an important component of next-generation wireless infrastructure.
1 Drivers and Vision of LAWN Low-altitude airspace, roughly defined as the region up to 3000 meters above ground level, as shown in Fig. 1, is poised to become a critical frontier for both human and machine activity in the coming decade. This space will support a diverse set of applications, from parcel-delivery drones and urban air taxis to precision-agriculture UAVs and public-safety 1
GNSS 3000m
Large UAV
Large UAV
ADS-B
T2: 3D Wireless Coverage and Handover in a Large Area 5G
Industrial UAVs
T4: Air Traffic Capacity limited by Communication and Sensing Air Corridor
eVTOL
>60m
Beam Handover
10~100km
Consumer UAV
Communication and Control
300m
BS1
BS2
Data Stream 4G
20~50m
0 1 0 1
1 0 1 0
Collision Detection
BS Sensing
BS
T1: UAV and eVTOL as Majority Terminals
BS
Drone Data Center
T3: Enhanced Situational Awareness for Air Traffic Management
Figure 1 The digitized low-altitude skyway and its evolutionary trends. This conceptual figure illustrates the multi-layer 3D architecture of LAWN as a structured skyway that integrates ISAC-enabled ground infrastructure, data centers, and heterogeneous airborne terminals to provide connectivity, high-resolution sensing, and autonomous traffic orchestration. The four main drivers highlighted in the figure are (T1) UAVs and eVTOLs as major aerial terminals with massive uplink data streams, (T2) 3D wireless coverage and handover across a large service area, (T3) enhanced situational awareness for air traffic management, and (T4) air traffic capacity constraints that are jointly shaped by communication and sensing performance within structured air corridors.
altitudes, including near-ground, rooftop, and severalhundred-meter levels, without signal gaps or weak zones [18,19] . For instance, a delivery drone or eVTOL should maintain network connectivity throughout its flight path, which requires seamless handover among BSs in both the vertical and horizontal planes. Therefore, additional rooftop BSs that fill vertical coverage gaps are often required in LAWN design, ensuring that low-altitude vehicles remain connected throughout the airspace corridor [20] . Trend 3: Enhanced Situational Awareness for Air Traffic Management: Real-time situational awareness, high-precision terrain information, and nearby weather awareness are crucial for flying vehicles, which in turn require LAWN to function as a reliable situational-awareness infrastructure [21,22] . Flying vehicles must not operate “blind”; instead, the network and associated management systems should continuously monitor the surroundings of each aircraft to prevent accidents. In addition, the feasibility of dynamic airspace management, route planning, and collision avoidance in low-altitude airspace is highly dependent on the sensing and communication capacity of the infrastructure. For example, NASA envisions predefined routes or corridors and “dynamic geofences” around no-fly zones, with networked warnings to ensure that a drone does not stray into a dangerous area. The width and height of air corridors are therefore tightly constrained by the sensing and communication precision available from the infrastructure. In this context, LAWN would effectively augment onboard sensors with widely distributed,
surveillance. This exponential growth in airborne connectivity and sensing services demands a paradigm shift in network design.
1.1 Communication and Sensing Trends Trend 1: UAVs and eVTOLs as Major Aerial Terminals: The proliferation of drones and eVTOL aircraft introduces a new class of heterogeneous wireless terminals operating in the sky [3,12] . By 2030, the number of cellular-connected UAVs is projected to reach tens of millions globally, rivaling smartphones in scale. These airborne devices require enormous data traffic volumes and ultra-reliable low-latency links for safety-critical control at scale [13–15] . For instance, a small fleet of drones can produce roughly 150 TB of data per day, so tens of millions of UAVs could collectively generate hundreds of petabytes of data daily. Unlike relatively homogeneous mobile phones, UAVs vary widely in size, altitude, and mission profile, resulting in highly diverse sensing and communication requirements [16] . This heterogeneity leads to complex connectivity challenges such as intermittent signal blockage (e.g., by buildings or terrain), extremely frequent handovers due to high mobility, and the need for adaptive protocols that maintain control links and target tracking for fast-moving aerial vehicles [17] . Trend 2: 3D Wireless Coverage and Handover in a Large Area: Providing seamless and robust coverage in three dimensions is a fundamental requirement for LAWN. In this setting, BSs (or other access nodes) must reliably serve users at various 2
2.1 Altitude-Specific Layered Operation
high-resolution, all-weather, day-and-night “eyes and ears” on the ground [23] . Trend 4: Air Traffic Capacity Constrained by Communication and Sensing: Air traffic capacity is jointly limited by self-positioning, sensing, and control precision [24] . For example, a drone travelling at 15 m/s can cover 1 to 2 m in 0.1 s, which may be the difference between stopping short of another drone and colliding with it. Moreover, drones flying near the ground are subject to urban-canyon effects [25] , in which signals such as GPS and radio links can be blocked or reflected by tall structures. A drone flying between high-rise buildings may lose line of sight to GPS satellites or BSs; even moderate-height buildings or terrain can cause intermittent GPS reception, which further limits the usable airspace and its traffic capacity [26] . On the other hand, sensing precision also constrains flying speed. For example, a small quadcopter cruising slowly can hold its position within a few centimeters and navigate around objects carefully, whereas at speeds above 20 m/s the position error or overshoot may become orders of magnitude larger, increasing collision risk.
2.1.1 Below 300 m Low-altitude airspace below 300 m is typically used for urban air mobility and local drone services. Representative applications include eVTOL air taxis within cities and last-mile delivery drones dropping off packages in neighborhoods [27] . Public-safety and infrastructure-monitoring missions are also common in this layer, such as inspection drones for utilities like power lines and railroads. Terrestrial BSs are expected to play a major role in connectivity for aerial vehicles below 300 m [28] . At these heights, drones often maintain line of sight to multiple cell towers, which yields strong received signals but also increases interference from neighboring cells. Therefore, proper interference-management techniques are required for LAWN evolution, such as adjusting antenna tilt or using sectors optimized for sky coverage. With these enhancements, an urban eVTOL or delivery drone can be served by LAWN for telemetry, control, and payload communications. In addition, the network must interface with UTM systems by providing real-time tracking and identification of each drone over the cellular link [29] . Overall, a terrestrial-network-centric architecture is expected to handle the dense connectivity demand, provided that it is tightly integrated with airspace traffic management for access control and safety [30] .
1.2 Design Objectives of LAWN The overarching goal of LAWN is to serve as the wireless infrastructure underpinning low-altitude air transport, thereby ensuring the efficiency and safety of flying vehicles such as UAVs and eVTOLs. To meet this objective, LAWN must go beyond traditional cellular networks and seamlessly integrate communication, sensing, and air traffic management functions, as illustrated in Fig. 2b. Accordingly, LAWN should provide not only reliable and low-latency connectivity but also become an integral part of digitized airspace [9] . In this vision, each BS can act as a basic airspace-management unit, providing high-resolution situational-awareness services to surrounding aerial vehicles.
2.1.2 Between 300 m and 1,000 m The 300–1000 m altitude layer primarily supports medium-range aerial operations, including cross-city cargo drones and inter-city eVTOL air taxis. UAVs operating in this range can cover distances of tens of kilometers, serving regional logistics, medical-supply delivery, and inter-district transportation [31,32] . In this intermediate airspace, vehicles fly above building heights but below traditional commercial airplanes, typically within uncontrolled airspace except near airports. Therefore, to ensure safe and efficient operations, LAWN-based UTM systems must evolve to handle higher flight speeds, extended mission durations, and seamless oversight across broader regions. Maintaining continuous, wide-area connectivity becomes particularly challenging at these altitudes, as UAVs frequently move beyond the coverage of standard terrestrial networks. While terrestrial BSs can still provide connectivity along parts of these routes, a purpose-designed network is often required to support long-distance airborne operations. One possible approach is deploying high-power BSs on tall towers to extend cell range for aerial users. However, ground BSs may be sparse and cannot cover the entire 3D space, and UAV relays may be employed to fill this gap.
2 Network Architecture and Limiting Factors The transition from ground-centric cellular systems to a fully 3D LAWN architecture explicitly reflects the vertical diversity of low-altitude airspace. To provide both reliable connectivity and real-time traffic management, LAWN must be decomposed into altitudespecific layers, each with its own mix of terrestrial, aerial, and satellite links, as shown in Fig. 1. In this section, we first discuss the layered network architecture of LAWN and then examine its key limiting factors. 3
a
U U U 0 1 0 1
Uplink
S
D
b Medium UAV
0 1 0 1
Image Data Sensor Data Telemetry Data
Downlink Control
1 0 1 0
Urban Traffic Management
Navigation
D D D
Large UAV ADS-B
1 0 1 0
S
U
Instruction Approval
Commu n
Air Route Planning
60
Regional Monitoring Field
Data Analysis
Sensing Beam
Base Station
Sensing
Infrastructure Inspection Downlink Data
ication Beam
Beam
UTM Cloud Platform
UTM Center Controllers
Data Feedback
Center Controllers
Web Server
Database Server
Communication Beam
120
0
Administrator
Precision Agriculture
Main Server
Logistics Delivery
Service Providers
Remote Operators
…… Strategy Making
Uplink Data
Figure 2 System architecture of LAWN with tightly integrated communication, sensing, and unmanned traffic management (UTM). a, Comparison between a conventional downlink-dominant terrestrial frame structure (“DDDSU”) and an uplink-oriented LAWN frame structure (“UUUSD”), reflecting heavy image, sensor, and telemetry uploads together with lighter control and navigation downlinks. b, ISAC-enabled BSs provide communication and sensing beams for mission services such as regional monitoring, infrastructure inspection, logistics delivery, precision agriculture, and air-route planning, while a cloud UTM platform coordinates data analysis, feedback, strategy making, and instruction approval for administrators, controllers, service providers, and remote operators.
that can cause link unreliability and dead zones [40] . High mobility can also introduce pronounced Doppler shifts [41] . These physical-layer issues demand advanced link adaptation, beamforming, and potentially integrated sensing to continuously track and maintain robust links in 3D space [42] .
2.1.3 Above 1,000 m Above 1000 m, low-altitude UAV networks begin to intersect with the lower bounds of traditional aviation airspace. This category includes large UAVs for surveillance and remote sensing [33–35] and highflying drone swarms for wide-area monitoring [36,37] . For example, military or border-surveillance drones often cruise at altitudes of a few kilometers to gain a broad view, and environmental-monitoring UAVs may operate at similar heights to cover large regions. The operational airspace above 1000 m is often regulated, which requires UAVs to obtain flight clearance and continuous tracking similar to that of manned aircraft, especially if they share air corridors with piloted planes. Thus, interoperability with the conventional air traffic management system is a key requirement for LAWN. At these altitudes, aerial and satellite communications become the primary connectivity solutions. LEO satellite networks (such as Starlink) can offer broadband data, enabling drones to stay connected to the internet or control centers even over oceans or remote regions [38,39] . In addition, experimental high-altitude long-endurance (HALE) drones, originally designed for Earth observation, are now being explored as airborne communication platforms.
2.2.2 Network Layer Challenges LAWN must handle extreme mobility in three dimensions. A drone can travel through multiple cell sectors in minutes, leading to very frequent handovers. Without intelligent handover management, UAVs may drop connections or suffer high latency during handoff. Moreover, current network routing and cell association algorithms, optimized for quasi-static ground users, may perform poorly for aerial nodes. Supporting seamless mobility requires multi-cell connectivity that allows drones to connect to multiple towers simultaneously or predictive handover techniques based on trajectory [43] . The network layer also faces scalability concerns as millions of airborne devices demand routing, addressing, and QoS management in the sky, significantly increasing load on network infrastructure.
2.2.3 Asymmetric Downlink/Uplink Data Traffic
2.2 Key Limiting Factors for LAWN Despite rapid progress, several factors still limit the deployment and scaling of LAWN.
Fig. 2a highlights a fundamental asymmetry between uplink (UAV-to-BS) and downlink (BS-to-UAV) traffic, with LAWNs exhibiting significantly higher service demands in the uplink. Aerial nodes typically generate large volumes of data that must be transmitted from air to ground, including high-definition video streams and real-time sensor data [44] . Conversely, downlink transmissions from ground to air mostly consist of brief control commands and navigation updates, which are comparatively lightweight. This imbalance
2.2.1 Propagation & Channel Characteristics The wireless channel in low-altitude airspace can vary from dense multipath near the ground to nearly freespace propagation at higher altitudes. Near the ground in urban areas (< 300 m), signals are obstructed by buildings, leading to shadowing and multipath fading 4
3.1 Airspace-Channel Coupling and Capacity Limits
exists consistently across all altitude layers. Accordingly, substantial changes are required in resource allocation, antenna configuration, and protocol design relative to ground-based networks [45] . Antenna systems require vertical beamforming adaptations or upward-oriented coverage patterns to better receive and manage aerial transmissions.
The fundamental challenge of LAWN lies in the intrinsic coupling between airspace management and wireless channel capacity [9] . Unlike terrestrial cellular systems, where users are largely confined to a twodimensional plane, low-altitude airspace introduces a 3D, mobile, and congestion-prone environment, in which spatial conflicts directly translate into communication interference [52,53] . To establish a unified analytical foundation, it is necessary to model lowaltitude airspace capacity and channel capacity within a common framework [54] .
2.2.4 Spectrum Management Communication links in LAWN often share spectrum with ground networks [46] . Using existing cellular bands for drones can therefore cause mutual interference. In particular, drone signals transmitted from altitude can travel farther and potentially interfere with distant cell sites. The issue of spectrum allocation for UAVs is only beginning to be addressed by regulators. Careful spectrum planning, dynamic frequency coordination, and possibly new frequency bands (e.g., millimeter-wave for high-capacity drone-to-drone links or optical communications at higher altitudes) will be needed to accommodate LAWN growth.
3.1.1 Airspace and Channel Capacity Definition The 3D airspace D is discretized into N uniform grid cells, indexed by G = {1, . . . , N }. The global traffic state is represented by a position matrix X ∈ {0, 1}K×N , where K denotes the total number of active UAVs. The airspace capacity (Cair ) is defined as the macroscopic traffic volume supported by the system, corresponding to the L1 -norm of the traffic state: K X N X Cair (X) ≜ Xk,n = K,
2.2.5 Regulations
k=1 n=1
The deployment of LAWN is tightly coupled with evolving regulations [47] . Aviation authorities impose strict rules on unmanned aircraft operations. For example, routine flights are commonly limited to altitudes around 120 m. Expanding LAWN operations to medium or higher altitudes will require new airspace classifications. In addition to allocating suitable frequency bands, regulators may need to constrain transmit power to minimize interference. Furthermore, privacy and cybersecurity requirements will significantly shape network design, including encryption protocols and authentication mechanisms.
(1)
s.t. X1N = 1K ,
XT 1K ≤ cgeo . where the constraint X1N = 1K ensures that each UAV occupies exactly one grid cell, while XT 1K ≤ cgeo imposes the geometric occupancy limit of each cell [9] . The channel capacity characterizes the achievable spectral efficiency for user k given the spatial configuration X and precoding W:
Cchan,k (X) = log2 (1 + γk (W, X))
(2)
where γk is the Signal-to-Interference-plus-Noise Ratio (SINR). Under a practical regime of coherent intrabeam interference, the SINR is governed by the Beam Occupancy Level (µk ), which counts the active UAVs falling within the interference subspace of user k :
3 Digitized Airspace Governance: Modeling, Structuring, and Co-Design Airspace management is foundational to scaling lowaltitude applications because it encompasses the systematic organization and dynamic control of airspace to ensure the safe, efficient, and equitable integration of users such as UAVs and eVTOLs [48] . The design of structured air corridors is particularly important because it enhances operational safety and efficiency by segregating traffic flows, minimizing cross-path conflicts, and enabling predictable routing. In addition, conflict-resolution mechanisms must be embedded within such corridors, leveraging real-time sensing and communications to support collision avoidance and automated geofencing against incursions by unauthorized objects or non-cooperative flights [49–51] .
γk (X) ≈
ρ (µk (X) − 1)ρ + 1
(3)
where ρ ≜ Prx /σ 2 denotes the raw Signal-to-Noise Ratio (SNR).
3.1.2 Phase Transition of Airspace–Channel Coupling The relationship between Cair and Cchan exhibits a nonlinear phase transition governed by the interaction between thermal noise and spatial interference. By identifying the critical capacity threshold Ccrit = L(1 + 1/ρ), where L is the number of orthogonal 5
a
b
Regime Ⅰ Noise-Masked
Regime Ⅱ Linear Trade-off
Regime Ⅲ Saturation
No-Fly Zone
Spectral Efficiency Per UAV (bps/Hz)
7
Power Gain Effective
Power Gain Ineffective
Ground Surveillance Station
Transmit Power(SNR) Transmit Power: SNR = 0dB Transmit Power: SNR = 10dB Transmit Power: SNR = 20dB
6 5
Operational Phases Regime Ⅰ (Stable) Regime Ⅱ (Trade-off) Regime Ⅲ (Degraded)
4
Power Gain
3 2
Curves Converge
1 0
0
25
50
75
100
125
Inclined Corridor
Horizontal Corridor
Total Number of Active UAVs
150
175
200
c
Passenger Aircraft
Controlled Airspace
Aerial-Route
Safety Boundary
Vertical Corridor
3000m
Class D/E airspace
North
· Fixed-Wing Aircraft · Large Helicopters
1000m
South Buffer Zone
Hovering
Class D/E airspace · Light Helicopters · eVTOL Aircraft
Uncontrolled Airspace
East
West
300m
No-Fly Layer North
Class G airspace · Industrial UAVs · Manned Aircraft
120m
South Horizontal Corridor
Buffer Zone
East
West
Class W airspace · Consumer UAV · Industrial UAV
0m
Vertical Corridor Arrival Landing Point
Vertical Corridor
Geofenced No-Fly Zone
Departure Takeoff Point
Figure 3 Air-corridor-aware airspace structuring and the airspace–channel coupling limit in LAWN. a, Beamoccupancy illustration in a discretized airspace: as concurrent UAV density increases, some users remain in high-gain sectors whereas others become interference-limited near no-fly constraints. b, Spectral efficiency per UAV versus total active UAVs under different SNR conditions, revealing three operating regimes (noise-masked, linear trade-off, and saturation/degraded) and convergence at high load. The shaded regime boundaries are illustrated using the low-SNR case (ρ = 0 dB) for visual clarity and should be interpreted as conceptual operating regions rather than exact thresholds for all SNR curves. c, Multi-layer corridor design from 0 to 3000 m, including controlled and uncontrolled airspace classes, vertical, horizontal, and inclined corridors, buffer zones, and geofenced no-fly areas for heterogeneous users such as consumer and industrial UAVs, eVTOLs, helicopters, and passenger aircraft.
• The Saturation Regime: When Cair ≫ Ccrit , the system becomes strictly interference-limited. The channel capacity degrades inversely with airspace 1 density: Cchan ≈ log2 (1 + Cair /L−1 ). In this regime, power control is ineffective, and spatial separation or deconfliction becomes the dominant strategy.
beams, the airspace capacity can be divided into three distinct operational regimes, as shown in Fig. 3a,b:
• The Noise-Masked Regime: Occurring when Cair ≤ L, the system is noise-limited. In this sparse state, µk = 1, and Cchan is decoupled from Cair , bounded only by transmit power. • The Linear Trade-off Regime: In the region L < Cair ≤ Ccrit , beam collisions occur, but the interference power is comparable to the noise floor. A linear trade-off exists where joint power-space optimization is most effective.
3.1.3 QoS-Constrained Capacity Boundary To guarantee safe operation, the system must maintain a minimum spectral-efficiency requirement Rmin 6
spatial density of UAVs. This structuring limits the beam-occupancy level µk , preventing the system from exceeding the critical threshold Ccrit and entering a performance-collapse regime. Furthermore, the corridors are strategically aligned with the high-gain regions of BS beamforming patterns to support uninterrupted telemetry and control links. This mapping between 3D airspace and communication resources is therefore a prerequisite for ultra-reliable aerial operations [58] .
(in bit/s/Hz). The maximum allowable airspace capacity is thus theoretically bounded by: 1 1 max Cair ≤ L 1 + Rmin − (4) 2 −1 ρ This derivation provides a theoretical foundation for LAWN airspace management, highlighting that airspace throughput is fundamentally limited by beam occupancy, which is in turn shaped by sensing precision, interference management, and available communication resources.
3.2.3 Mission-Specific Architectures LAWN implements a multi-tier strategy for the design of dedicated air corridors tailored to specific mission requirements and operational priorities. For example, medical transport corridors may prioritize shortestpath routing with emergency right-of-way privileges, whereas cargo delivery corridors could focus on noise mitigation and industrial-zone alignment, following rigorous risk stratification. This mission-specific optimization is further bolstered by dynamic geofencing capabilities, which allow the network to enforce automated boundaries around no-fly zones [59] . By leveraging networked warnings and real-time situational awareness, the infrastructure ensures that UAVs do not stray into restricted or hazardous areas, thereby transforming the airspace into a managed, task-aware mobility layer [60] .
3.2 Structured Air Corridors: Organizing 3D Mobility Air corridors are reserved 3D volumes designed to structure aerial traffic, replacing free-flight regimes with predefined pathways and thereby enhancing both operational safety and traffic efficiency [55,56] . Unlike uncontrolled airspace, air corridors enforce predetermined flight trajectories to minimize collision risks, reduce communication and sensing overhead, and prevent incursions into restricted zones [56] .
3.2.1 Geometric Layering and Spatial Discretization The fundamental concept of the air corridor involves discretizing and layering the continuous 3D low-altitude airspace based on altitude, orientation, and mission attributes. As illustrated in Fig. 3c, this architectural shift replaces free-flight regimes with a multi-layered model in which the airspace is partitioned into altitude-specific corridor units to enhance operational safety and traffic efficiency. In a typical implementation, the airspace is divided into three functional layers: the top and bottom layers serve as directional conduits, accommodating southbound/northbound and eastbound/westbound traffic, respectively, thereby segregating flows to minimize mid-air cross-path conflicts. A specialized middle layer acts as a transition zone, providing a reserved volume for aircraft to hover or adjust altitude when changing direction, thereby supporting fluid and coordinated traffic orchestration. The width and height of these corridors are tightly constrained by the sensing and communication precision provided by LAWN infrastructure [57] . By aligning these physical dimensions with network capability, the system enforces predetermined flight trajectories and reduces the likelihood of incursions into restricted zones.
3.3 Physics-Aware CSC Co-Design: Closing the Loop 3.3.1 Communication Model Most existing systems treat communication, sensing, and control as loosely coupled functional modules, which struggle to operate effectively in highly dynamic LAWN environments [61] . Conventional approaches often rely on idealized or quasi-static communication assumptions, thereby overlooking the stochastic latency, packet loss, and time-varying reliability inherent in wireless channels [62] . In the context of closed-loop control over wireless links, the actual execution of the control command is conditioned on successful packet delivery. Specifically, the discrete-time dynamics of the UAV can be formulated as:
qt+1 = Aqt + αt But + nt ,
(5)
where qt denotes the system state vector, A is the state transition matrix characterizing the inherent physical evolution of the plant, B is the control input matrix, ut is the control command generated by the BS based on the latest estimates, and nt represents the process noise. The binary variable αt ∈ {0, 1} characterizes the communication packet loss at time slot t, which follows a Bernoulli distribution: ( 1, with probability pt (wt ), αt = (6) 0, with probability 1 − pt (wt ),
3.2.2 Communication-Space Resource Mapping Each air corridor implicitly corresponds to a finite communication resource budget, including the number of available beams, tolerable interference levels, and minimum communication-rate constraints. By confining UAV flight paths within designated corridors, the infrastructure can effectively control the 7
a
c Communication and Sensing
Communication
Delivery Drone
Sensing/Radar
Control
Sensing Data Fusion & Analysis
Surveillance Drone
Passenger eVTOL
Multi-Source Data Input
Radar
Cloud Platform Data Fusion Center
Camera
LiDAR
AI-Driven Situational Awareness & Prediction Radar
3D Dynamic Map
Control Command Generation
ISAC Base Stations
Trajectory Planning & Decision Making
b Application Scenarios
Logistics Needs Coverage
Logistics & Delivery
Cost Efficiency
Edge Processing Unit
Emergency Needs
Emergency Rescue Urban Air Mobility (UAM)
Latency
Secure & Reliable Requirements
Reliability
Local Control
UAM Needs Bandwidth
Safety Integrity
Figure 4 CSC co-design framework for closed-loop LAWN operations. a, Delivery drones, passenger eVTOLs, surveillance UAVs, radar, and ISAC BSs exchange communication and sensing links, and multi-source observations are fused at a cloud data platform. b, Representative applications (logistics, emergency rescue, and urban air mobility) map to differentiated requirement vectors in coverage, latency, reliability, bandwidth, cost efficiency, and safety integrity. c, Fused radar/camera/LiDAR data drive AI situational awareness, 3D dynamic mapping, trajectory planning, and control-command generation, while edge processing units provide local feedback control for real-time and safety-critical actuation.
where et denotes the state estimation error. In our framework, this error is modeled as a zero-mean Gaussian random variable with covariance Σt = E[et eTt ]. Based on estimation theory, the lower bound of this covariance is characterized by the Cramér-Rao Bound (CRB), which is the inverse of the Fisher Information Matrix (FIM). Specifically, when estimating the spatial parameter θ (e.g., Angle-of-Arrival), the FIM J(θ) is intrinsically coupled with the beamforming design wt . For a system with L snapshots and Nr receive antennas, the CRB for angle estimation is given by:
where pt (wt ) is the packet success probability (PSP). Unlike static models, the PSP in our framework is intrinsically coupled with the beamforming design wt . To capture the nonlinear relationship between the Signal-to-Interference-plus-Noise Ratio (SINR) and link reliability, we employ a sigmoid-based mapping function:
pt (wt ) =
1 , 1 + exp (−a (SINRt (wt ) − γth ))
(7)
where a is the steepness coefficient and γth represents the target SINR threshold. Any failure of the beamforming gain to maintain SINRt above the threshold γth results in a sharp escalation of the packet loss probability. For an unstable system, such communicationinduced outages can lead to the exponential growth of the state error, potentially violating the topological stability requirements discussed in the sequel.
CRB(θ) =
σs2 , H 2 r · |ȧ (θ )wt |
2L|β|2 N
(9)
From the sensing perspective, the BS simultaneously performs target sensing by processing the echoes of the transmitted signal. Let qt be the true kinematic state of the UAV. The BS obtains the state estimate q̂t through echo processing, which is defined as:
where σs2 is the noise variance, L represents the number of snapshots used for sensing accumulation, Nr is the number of receive antennas, β is the complex channel gain, and ȧ(θ) represents the derivative of the steering vector a(θ) with respect to θ. This bound reveals a fundamental trade-off, indicating that minimizing sensing uncertainty requires the beamforming vector wt to be aligned with the derivative of the steering vector ȧ(θ) rather than with the steering vector itself. The sensing uncertainty directly propagates into the control domain. By applying the state-feedback control law ut = −Kq̂t , the control input becomes:
q̂t = qt − et ,
ut = −K(qt − et ) = −Kqt + Ket .
3.3.2 Sensing Model
(8) 8
(10)
where λi (A) are the unstable eigenvalues of the state transition matrix. This necessitates an entropy reduction constraint on the beamforming design:
where K represents the state-feedback gain matrix. Then, the control law can be expanded as:
qt+1 = Aqt + αt But + nt , = (A − αt BK)qt + αt BKet + nt ,
(11a)
H(A)
SINR(wt ) ≥ 2 BW − 1, | {z }
(11b)
(14)
γcritical
To ensure the stability of the closed-loop system under stochastic packet loss and sensing uncertainty, we define a quadratic Lyapunov function V (qt ) = qTt Pqt , where P ≻ 0 is a positive definite matrix [63] . The stability of the system is guaranteed if the onestep conditional expectation of the Lyapunov function satisfies the following drift constraint [64] :
E[V (qt+1 ) | qt ] − ηV (qt ) ≤ 0,
where BW is the communication bandwidth and γcritical delineates the absolute survival threshold. Any misalignment that causes SINR(wt ) < γcritical renders the system unstabilizable, as the information flow becomes insufficient to neutralize physical entropy production [69] . Beyond mere survival, the CSC architecture must guarantee stability and state convergence. Consequently, the CSC co-design problem transitions into a survival-constrained joint optimization [70] :
(12)
where η ∈ (0, 1) denotes the target decay rate. This formulation defines the safety boundary of the system, where E[·] is intrinsically coupled with both the packet success probability p(wt ) and the sensingerror covariance represented through CRB(wt ). To satisfy (12), the beamforming vector wt must be judiciously designed to navigate the fundamental trade-off between enhancing communication reliability p(wt ) and suppressing the state-estimation error et .
min ∥wt ∥2 s.t.
To navigate the complexities of the emerging lowaltitude economy, LAWN must transcend the conventional siloed design paradigm and evolve toward a deeply integrated communication, sensing, and control (CSC) co-design architecture [65] . As illustrated in Fig.4, this architecture establishes a closedloop ecosystem where multi-source data streams from ISAC-enabled BSs, terrestrial radars, and UAV platforms are aggregated at cloud-edge platforms. These fused data streams drive AI-powered situational awareness modules, enabling precise trajectory planning and real-time navigation command generation [66] . Furthermore, the CSC framework supports adaptive resource scheduling to accommodate the heterogeneous mission requirements inherent in LAWN [67] . These range from the wide-area coverage and moderate latency needs of logistics delivery to the stringent safety-integrity and ultra-reliable lowlatency constraints of urban air mobility. Unlike traditional wireless systems where reliability is primarily defined by the ability to decode signals against noise, reliability in CSC co-design is intrinsically linked to the physical stability of the UAV. According to the data rate theorem, a linear unstable system is stabilizable only if the feedback communication rate R(wt ) exceeds the system’s inherent uncertainty expansion rate, known as the topological entropy [68] :
X
log2 |λi (A)|,
SINR(wt ) ≥ γcritical ,
(15a)
E[V (qt+1 ) | qt ] ≤ ηV (qt ),
(15b)
where (15a) ensures topological stabilizability, while (15b) enforces kinematic convergence. By aligning communication reliability and sensing precision with the entropy-based stability requirement, the CSC architecture ensures that the infrastructure dynamically prioritizes information fidelity and latency based on the instantaneous kinematic risk of aerial platforms [71–73] . This approach transforms LAWN from a “best-effort” data pipe into a physics-aware resilient network capable of supporting safety-critical urban air mobility.
3.3.3 Entropy-Based CSC Co-Design
H(A) =
(P1)
wt
4 Challenges and Opportunities in LAWN The differences between LAWN and traditional cellular networks are summarized in Table 1. To fully harness the potential of LAWN, critical technical challenges spanning both the physical and network layers must be addressed [74] . These challenges not only underscore the complexity of deploying robust LAWN systems but also present unique research opportunities across communication and sensing protocols, resource management, and network scalability. In this section, we identify key obstacles in physicallayer enhancement, sensing services, and network and airspace management, and propose practical directions to advance the feasibility and performance of LAWN architectures.
4.1 Physical-Layer Enhancement in LAWN
(13)
As established in prior sections, the physical layer of LAWN exhibits intrinsic operational contrasts with
|λi (A)|>1
9
terrestrial systems, such as the asymmetric downlink/uplink data traffic. This inversion of traditional traffic dynamics, where downlink services typically dominate, stems from LAWN-specific use cases, such as UAV-based data collection and aggregation or realtime telemetry, necessitating tailored physical-layer design strategies [75] . Therefore, several physical-layer enhancements are required in LAWN.
4.2 Sensing and ISAC
4.1.1 Dedicated Frame Structure Design
4.2.1 Sensing Coverage Enhancement
The conventional 5G NR frame structure (e.g., “DDDSU”) is no longer suited to LAWNs because of their inherent uplink-centric demands. In this configuration, as shown in Fig. 2a, “D” denotes downlink slots, “U” represents uplink slots, and “S” signifies a flexible special slot comprising mixed downlink and uplink symbols together with a guard period (GP), which prioritizes downlink-heavy traffic in typical terrestrial networks [76] . In contrast, LAWNs require dynamic slot allocation that favors uplink capacity. This mismatch underscores the need for adaptive frame structures tailored to LAWNs’ asymmetric traffic patterns.
Conventional terrestrial networks are conceived for 2D ground coverage, typically offering reliable communication service only up to roughly 100 m in vertical height from sub-6 GHz macro BSs. However, aerial platforms in LAWN can operate from near-ground altitudes to around 3000 m, and potentially beyond in some suburban or rural scenarios. This vertical disparity creates sensing-coverage gaps that must be addressed. One effective strategy is to expand the vertical beamforming angle of the transmit antenna array. Recent experiments within the framework of 5G-Advanced (5G-A) networks have demonstrated that increasing the vertical beamforming angle to 65◦ can substantially enhance integrated ground-to-air sensing and communication performance [79] . Specifically, a well-designed monostatic ISAC-enabled active antenna unit (AAU) operating at sub-6 GHz has been shown to support reliable sensing at altitudes up to 600 m. Additionally, cooperative sensing and beam coordination among multiple BSs can be employed to realize seamless sensing in complex urban environments.
ISAC, which integrates radio sensing and wireless communication into a unified system [4] , plays a foundational role in enabling the full capabilities of LAWN. This convergence equips wireless infrastructure with real-time sensing, dynamic target tracking, and situational-awareness capabilities in low-altitude airspace.
4.1.2 Task-Oriented Uplink Transmission In LAWN applications such as environmental monitoring, infrastructure inspection, and media broadcasting, UAVs collect substantial sensor data that require real-time transmission to BSs under minimal distortion. To optimize this process, semantic communication techniques may be employed to compress data in a content-aware and task-oriented manner, ensuring that only important and relevant information is delivered [77] . For example, during infrastructure inspections of power lines or pipelines, UAVs can use semantic compression to focus on transmitting data related to potential anomalies or defects. This approach reduces bandwidth usage and accelerates issue identification.
4.2.2 Tradeoff between UAV Sensing and Terrestrial User Communication One of the key challenges in LAWN arises from integrating wireless sensing functionality into the network infrastructure, which introduces an inherent tradeoff between UAV sensing and communication service delivery to ground users [80] . To achieve the desired sensing performance, certain BSs must adjust the tilt angles of their antenna arrays to steer spatial beams toward UAVs [81,82] . However, such beam steering often diminishes radiated signal power toward the ground, resulting in a substantial decline in communication rates for terrestrial users. A potential solution is to deploy dedicated BSs at elevated positions that exclusively provide ISAC services to low-altitude UAVs. While effective, this approach incurs significant costs in terms of infrastructure deployment and network planning. As a more practical alternative, advanced ISAC signal processing and resource-allocation strategies can be implemented in a “best-effort” manner using existing BSs [83] . These strategies aim to minimize sensing loss functions, such as the mean squared error (MSE) in UAV localization and tracking, while simultaneously upholding the quality-of-service (QoS) requirements of ground users [84,85] .
4.1.3 Multi-Functional URLLC Downlink Transmission Unlike terrestrial networks, LAWN downlinks operate as multifunctional platforms rather than simple data pipes, converging sensing, communication, and control into a unified framework. This integration equips BSs with continuous, ubiquitous, and all-weather airspace surveillance via downlink wireless sensing, enabling early detection of unauthorized UAV intrusions [78] . Simultaneously, BSs leverage sensing-derived spatial information to dynamically adjust UAV flight paths through downlink navigation commands, which is a critical capability in GNSS-denied environments. To realize real-time, seamless, and robust control of large fleets of UAVs, downlink data must be delivered at very low latency (typically below 10 ms), where URLLC techniques may play a vital role. 10
Table 1 Requirement Metrics: LAWN vs. Traditional Cellular Networks
Category
Traditional Cellular Network
Low-Altitude Wireless Network (LAWN)
Primary Service Targets Coverage Dimension Typical Mobility
Smartphones, IoT devices 2D ground coverage Ground-based (<120 km/h)
Uplink/Downlink Balance Frequency Bands Latency Requirement
Downlink-dominant (>70%) Sub-6 GHz, partial mmWave 10–100 ms (eMBB)
Throughput Requirement Reliability Requirement Communication Range Sensing Capability Interference Pattern QoS Management Deployment Topology Sensing Accuracy
Downlink: 10–100 Mbps average ∼99.9% (standard QoS) ∼200–500 m (urban macro cell) Weak or absent Horizontal inter-cell interference RRM, slicing, HARQ, MCS Fixed macro/micro BSs ∼10 meters (GNSS-based)
Handover Complexity User/Device Density Energy Constraints Service Scenarios
Horizontal handover (2D) ∼1 million devices/km2 Moderate (smartphones/IoT) Video streaming, IoT, VoLTE
UAVs, eVTOLs, aerial sensors, airborne radars 3D integrated ground-air coverage (0–3000 m) Mission-dependent 3D aerial mobility (tens to hundreds of km/h) Uplink-dominant (>60%) due to telemetry and sensing Sub-6 GHz, mmWave, potential THz <10 ms (control); millisecond-level for sensing and safety-critical functions Uplink: 50–200 Mbps average; up to Gbps level >99.999% (critical UAV control) Several km in LoS; dynamic 3D coverage Integrated sensing and communication (ISAC) Vertical cross-domain interference (UAV ↔ UE) Task-aware dynamic scheduling, trajectory-based QoS Hybrid deployment with flying relays and fixed BSs Meter-level to sub-meter-level (target detection and tracking) Frequent 3D handovers, CoMP support Fewer users but high dynamics and spatial complexity High (energy-constrained UAVs) Remote control, real-time sensing, HD video uplink
4.2.3 Low-Slow-Small Target Detection and Recognition
vantage points. A key advantage lies in their ability to augment terrestrial perception through crossperspective fusion [88] . In a coordinated LAWN setting, UAVs operating at altitudes of 100–200 m can help identify blind spots in ground-based sensing, such as alleyways or overpasses. For instance, when aerial LiDAR depth maps are fused with mmWave radar returns from ground BSs, it becomes feasible to reconstruct partial occlusions and estimate pedestrian trajectories with higher confidence. Nevertheless, the altitude and speed of aerial platforms cause rapid geometric changes in observation angle, resulting in significant variation in elevation angle and slant range, which in turn leads to complex backscattering effects and geometric distortion. Airborne mmWave radar mounted on low-flying UAVs can detect ground vehicles in urban canyons with sub-meter accuracy, but performance degrades significantly when targets are shielded by metallic structures or dense foliage.
Low-altitude airspace is increasingly populated with low, slow, and small (LSS) aerial targets, such as consumer drones, birds, and micrometeorological probes, which pose challenges for both civilian applications and airspace security. Most such targets are characterized by low radar cross-section (RCS, often below –20 dBsm), slow and irregular motion (e.g., 5–20 m/s), and limited RF emissions, making them difficult to detect or recognize using conventional radar platforms [86] . Tests conducted by the UK Civil Aviation Authority showed that small quadcopters, such as the DJI Phantom series with an RCS as low as –23 dBsm, could not be reliably tracked beyond 500 m using standard airport-surveillance radar. Tackling this issue necessitates close collaboration among large numbers of BSs in LAWN. By exploiting the spatial diversity of a dense BS deployment, the network can correlate faint signal anomalies across multiple observation points and thereby improve detectability.
5 Outlook 4.2.4 Air-to-Ground Sensing
In this article, we have presented a vision of lowaltitude wireless networks as a new wireless infrastructure for digitized airspace. Starting from the driving trends and design objectives of LAWN, we discussed its layered network architecture, key limiting factors, and the intrinsic coupling among airspace capacity, wireless channel capacity, and closed-loop control stability. Compared with traditional cellular networks, LAWN represents a paradigm shift from
Air-to-ground sensing serves as a critical enabler for situational awareness in LAWN, especially as UAVs increasingly function not only as communication relays but also as mobile sensing platforms [87] . Unlike terrestrial sensing systems constrained by obstructions and horizon limits, aerial platforms benefit from broader LoS coverage and flexible mobility, enabling real-time environmental mapping from elevated 11
[12] Namuduri, K. et al. Advanced air mobility: Research directions for communications, navigation, and surveillance. IEEE Vehicular Technology Magazine 17, 65–73 (2022). [13] Xia, Y. et al. Characterization of low-power wireless links in uav-assisted wireless-sensor network. IEEE Internet of Things Journal 10, 5823–5842 (2023). [14] Mohsan, S. A. H., Othman, N. Q. H., Li, Y., Alsharif, M. H. & Khan, M. A. Unmanned aerial vehicles (uavs): Practical aspects, applications, open challenges, security issues, and future trends. Intelligent service robotics 16, 109–137 (2023). [15] Zhang, W. et al. Energy transfer and data collection from batteryless sensors in low-altitude wireless networks. arXiv preprint arXiv:2507.07481 (2025). [16] Song, Y. et al. An overview of cellular isac for low-altitude uav: New opportunities and challenges. IEEE Communications Magazine 63, 88–95 (2025). [17] Mohsan, S. A. H., Khan, M. A., Noor, F., Ullah, I. & Alsharif, M. H. Towards the unmanned aerial vehicles (uavs): A comprehensive review. Drones 6, 147 (2022). [18] Guo, Y. & Zhang, H. 3d boundary modeling and handover analysis of aerial users in heterogeneous networks. IEEE Transactions on Vehicular Technology 72, 13523–13529 (2023). [19] Zhou, S., Liu, X., Tang, B. & Tan, G. Handover and coverage analysis in 3-d mobile uav cellular networks. IEEE Internet of Things Journal 11, 29911–29925 (2024). [20] Shayea, I. et al. Handover management for drones in future mobile networks—a survey. Sensors 22, 6424 (2022). [21] Shrestha, R., Bajracharya, R. & Kim, S. 6g enabled unmanned aerial vehicle traffic management: A perspective. IEEE Access 9, 91119– 91136 (2021). [22] Chi, Y., Nie, J., Zhong, L., Wang, Y. & Delahaye, D. A review of situational awareness in air traffic control. IEEE Access 11, 134040–134057 (2023). [23] Hamissi, A. & Dhraief, A. A survey on the unmanned aircraft system traffic management. ACM Computing Surveys 56, 1–37 (2023). [24] Bulusu, V., Polishchuk, V., Sengupta, R. & Sedov, L. AIAA (ed.) Capacity estimation for low altitude airspace. 17th AIAA Aviation Technology, Integration, and Operations Conference (AIAA ed.), 4266 (2017). [25] Cui, Y. & Ge, S. S. Autonomous vehicle positioning with gps in urban canyon environments. IEEE Transactions on Robotics and Automation 19, 15–25 (2003). [26] Fei, Z., Wang, X., Wu, N., Huang, J. & Zhang, J. A. Air-ground integrated sensing and communications: Opportunities and challenges. IEEE Communications Magazine 61, 55–61 (2023).
communication-oriented connectivity to a 3D multifunctional infrastructure that integrates communication, sensing, and control. In this vision, BSs are expected to serve not only as communication access points, but also as basic airspace-management units that support situational awareness, traffic orchestration, and safety-critical control. These new requirements and applications also introduce a range of research challenges and opportunities, spanning physical-layer enhancement, sensing service provision, network resource management, and airspace governance. Addressing these challenges will be essential for transforming LAWN from a conceptual extension of cellular networks into a reliable air-transport infrastructure for future low-altitude economies.
References [1] Lin, X., Rommer, S., Euler, S., Yavuz, E. A. & Karlsson, R. S. 5g from space: An overview of 3gpp non-terrestrial networks. IEEE Communications Standards Magazine 5, 147–153 (2021). [2] You, X. et al. Towards 6g wireless communication networks: Vision, enabling technologies, and new paradigm shifts. Science China information sciences 64, 110301 (2021). [3] Zaid, A. A., Belmekki, B. E. Y. & Alouini, M.-S. evtol communications and networking in uam: Requirements, key enablers, and challenges. IEEE Communications Magazine 61, 154–160 (2023). [4] Liu, F. et al. Integrated sensing and communications: Toward dual-functional wireless networks for 6G and beyond. IEEE J. Sel. Areas Commun. 40, 1728–1767 (2022). [5] Yuan, W. et al. From ground to sky: Architectures, applications, and challenges shaping low-altitude wireless networks. arXiv preprint arXiv:2506.12308 (2025). [6] Zhu, W. Low-altitude economy on cusp of takeoff. China Daily Global (2025). [7] Tang, J. et al. Cooperative isac-empowered low-altitude economy. IEEE Transactions on Wireless Communications (2025). [8] Wu, J. et al. Low-altitude wireless networks: A comprehensive survey. arXiv preprint arXiv:2509.11607 (2025). [9] Huang, H., Su, J. & Wang, F.-Y. The potential of low-altitude airspace: The future of urban air transportation. IEEE Transactions on Intelligent Vehicles (2024). [10] Cui, Y., Liu, F., Jing, X. & Mu, J. Integrating sensing and communications for ubiquitous iot: Applications, trends, and challenges. IEEE Network 35, 158–167 (2021). [11] Jiang, Y. et al. Integrated sensing and communication for low altitude economy: Opportunities and challenges. IEEE Communications Magazine 63, 72–78 (2025). 12
[27] Ha, N. T., Akbari, M. & Au, B. Last mile delivery in logistics and supply chain management: A bibliometric analysis and future directions. Benchmarking: An International Journal 30, 1137–1170 (2023). [28] Abdalla, A. S. & Marojevic, V. Communications standards for unmanned aircraft systems: The 3gpp perspective and research drivers. IEEE Communications Standards Magazine 5, 70–77 (2021). [29] Prevot, T. et al. AIAA (ed.) Uas traffic management (utm) concept of operations to safely enable low altitude flight operations. 16th AIAA aviation technology, integration, and operations conference (AIAA ed.), 3292 (2016). [30] Honghong, Z. et al. Risk assessment framework for low-altitude uav traffic management. Journal of Intelligent & Fuzzy Systems 42, 2775–2792 (2022). [31] Geraci, G., López-Pérez, D., Benzaghta, M. & Chatzinotas, S. Integrating terrestrial and nonterrestrial networks: 3d opportunities and challenges. IEEE Communications Magazine 61, 42–48 (2022). [32] Geraci, G. et al. What will the future of uav cellular communications be? a flight from 5g to 6g. IEEE communications surveys & tutorials 24, 1304–1335 (2022). [33] Huang, Y., Reddy, K. N., Fletcher, R. S. & Pennington, D. Uav low-altitude remote sensing for precision weed management. Weed technology 32, 2–6 (2018). [34] Guo, Y. et al. IEEE (ed.) A drone-based sensing system to support satellite image analysis for rice farm mapping. IGARSS 2019-2019 IEEE International Geoscience and Remote Sensing Symposium (IEEE ed.), 9376–9379 (IEEE, 2019). [35] Mathews, A. J., Singh, K. K., Cummings, A. R. & Rogers, S. R. Fundamental practices for drone remote sensing research across disciplines. Drone Systems and Applications 11, 1–22 (2023). [36] Asaamoning, G., Mendes, P., Rosário, D. & Cerqueira, E. Drone swarms as networked control systems by integration of networking and computing. Sensors 21, 2642 (2021). [37] Saffre, F., Karvonen, H. & Hildmann, H. Springer (ed.) Wild swarms: autonomous drones for environmental monitoring and protection. International conference on FinDrones (Springer ed.), 1–32 (Springer, 2023). [38] Liu, J., Shi, Y., Fadlullah, Z. M. & Kato, N. Space-air-ground integrated network: A survey. IEEE Communications Surveys & Tutorials 20, 2714–2741 (2018). [39] Zhang, P. et al. Uav communication in space–air– ground integrated networks (sagins): Technologies, applications, and challenges. Drones 9, 108 (2025).
[40] Khawaja, W., Guvenc, I., Matolak, D. W., Fiebig, U.-C. & Schneckenburger, N. A survey of airto-ground propagation channel modeling for unmanned aerial vehicles. IEEE Communications Surveys & Tutorials 21, 2361–2391 (2019). [41] Sandri, M., Pagin, M., Giordani, M. & Zorzi, M. Implementation of a channel model for nonterrestrial networks in ns-3. Proceedings of the 2023 Workshop on ns-3, 28–34 (2023). [42] Giordani, M. & Zorzi, M. Non-terrestrial networks in the 6g era: Challenges and opportunities. IEEE network 35, 244–251 (2020). [43] Rinaldi, F. et al. Non-terrestrial networks in 5g & beyond: A survey. IEEE access 8, 165178–165200 (2020). [44] Zhou, H., Hu, F., Juras, M., Mehta, A. B. & Deng, Y. Real-time video streaming and control of cellular-connected uav system: Prototype and performance evaluation. IEEE Wireless Communications Letters 10, 1657–1661 (2021). [45] Zhang, Z. et al. 6g wireless networks: Vision, requirements, architecture, and key technologies. IEEE vehicular technology magazine 14, 28–41 (2019). [46] Lee, H.-W., Medles, A., Chen, C.-C. & Wei, H.Y. Feasibility and opportunities of terrestrial network and non-terrestrial network spectrum sharing. IEEE Wireless Communications 30, 36–42 (2023). [47] Azari, M. M. et al. Evolution of non-terrestrial networks from 5g to 6g: A survey. IEEE communications surveys & tutorials 24, 2633–2672 (2022). [48] Li, A., Hansen, M. & Zou, B. Traffic management and resource allocation for uav-based parcel delivery in low-altitude urban space. Transportation Research Part C: Emerging Technologies 143, 103808 (2022). [49] Seo, J., Kim, Y., Kim, S. & Tsourdos, A. Collision avoidance strategies for unmanned aerial vehicles in formation flight. IEEE Transactions on aerospace and electronic systems 53, 2718–2734 (2017). [50] Yasin, J. N. et al. Unmanned aerial vehicles (uavs): Collision avoidance systems and approaches. IEEE access 8, 105139–105155 (2020). [51] Rezaee, M. R., Hamid, N. A. W. A., Hussin, M. & Zukarnain, Z. A. Comprehensive review of drones collision avoidance schemes: Challenges and open issues. IEEE Transactions on Intelligent Transportation Systems 25, 6397–6426 (2024). [52] Shoewu, O. O., Akinyemi, L. A. & Edozie, R. in Uav cellular communication in 5g new radio wireless standards Unmanned Aerial Vehicle Cellular Communications (Shoewu, O. O. ed.) 25–45 (Springer, 2022). [53] Luo, G. et al. Toward low-altitude airspace management and uav operations: Requirements, architecture and enabling technologies. IEEE
13
Wireless Communications (2025). [54] Bulusu, V., Polishchuk, V., Sengupta, R. & Sedov, L. AIAA (ed.) Capacity estimation for low altitude airspace. 17th AIAA Aviation Technology, Integration, and Operations Conference (AIAA ed.), 4266 (2017). [55] Aggarwal, S. & Kumar, N. Path planning techniques for unmanned aerial vehicles: A review, solutions, and challenges. Computer communications 149, 270–299 (2020). [56] Muna, S. I. et al. Air corridors: Concept, design, simulation, and rules of engagement. Sensors 21, 7536 (2021). [57] Wang, S., Liu, H., Rinaldi, M. & Tsang, Y. P. Evaluating the impact of air corridors on the environment and public interests. Transportation Research Part D: Transport and Environment 143, 104732 (2025). [58] Karimi-Bidhendi, S., Geraci, G. & Jafarkhani, H. Optimizing cellular networks for uav corridors via quantization theory. IEEE Transactions on Wireless Communications (2024). [59] Haddad, J., Mirkin, B. & Assor, K. Traffic flow modeling and feedback control for future low-altitude air city transport: An mfd-based approach. Transportation Research Part C: Emerging Technologies 133, 103380 (2021). [60] Munir, A., Aved, A. & Blasch, E. Situational awareness: techniques, challenges, and prospects. AI 3, 55–77 (2022). [61] Jin, H. et al. Advancing the control of lowaltitude wireless networks: Architecture, design principles, and future directions. npj Wireless Technology 2, 2 (2026). [62] Wang, Z. et al. Qos-oriented sensing– communication–control co-design for uav-enabled positioning. IEEE Transactions on Green Communications and Networking 7, 497–511 (2023). [63] Polycarpou, M. M. & Ioannou, P. A. IEEE (ed.) A robust adaptive nonlinear control design. 1993 American control conference (IEEE ed.), 1365– 1369 (IEEE, 1993). [64] Freeman, R. & Kokotovic, P. V. Robust nonlinear control design: state-space and Lyapunov techniques (Springer Science & Business Media, 2008). [65] Jin, H., Wu, J., Yuan, W., Liu, F. & Cui, Y. Codesign of sensing, communications, and control for low-altitude wireless networks. IEEE Transactions on Mobile Computing 24, 12035–12048 (2025). [66] Xu, Y., Zhang, T., Liu, Y. & Yang, D. Uavenabled integrated sensing, computing, and communication: A fundamental trade-off. IEEE Wireless Communications Letters 12, 843–847 (2023). [67] Liang, T., Zhang, T., Cao, B. & Zhang, Q. Sensing, communication, and control co-design for energy-efficient uav-aided data collection. IEEE Wireless Communications Letters (2024).
[68] Minero, P., Franceschetti, M., Dey, S. & Nair, G. N. Data rate theorem for stabilization over time-varying feedback channels. IEEE Transactions on Automatic Control 54, 243–255 (2009). [69] Wan, S. & Qiu, L. IEEE (ed.) Stabilization of networked control systems with finite data rate. 2009 7th Asian Control Conference (IEEE ed.), 1067–1072 (2009). [70] Eisen, M. et al. Control aware radio resource allocation in low latency wireless control systems. IEEE Internet of Things Journal 6, 7878–7890 (2019). [71] Gatsis, K., Ribeiro, A. & Pappas, G. J. Optimal power management in wireless control systems. IEEE Transactions on Automatic Control 59, 1495–1510 (2014). [72] Zheng, M., Zhang, L. & Liang, W. Controlaware resource scheduling method for wireless networked control systems. IEEE Sensors Journal 23, 21946–21955 (2023). [73] Scheuvens, L. et al. State-aware resource allocation for wireless closed-loop control systems. IEEE Transactions on Communications 69, 6604–6619 (2021). [74] Zeng, Y., Zhang, R. & Lim, T. J. Wireless communications with unmanned aerial vehicles: Opportunities and challenges. IEEE Communications magazine 54, 36–42 (2016). [75] Zeng, Y., Zhang, R. & Lim, T. J. Wireless communications with unmanned aerial vehicles: opportunities and challenges. IEEE Communications Magazine 54, 36–42 (2016). [76] 3GPP, W. Study on new radio (nr) to support non-terrestrial networks. 3rd Generation Partnership Project (3GPP), Technical Report (TR), TR 38.811 (2020). [77] Xie, H., Qin, Z., Li, G. Y. & Juang, B.-H. Deep learning enabled semantic communication systems. IEEE Transactions on Signal Processing 69, 2663–2675 (2021). [78] Cui, Y. et al. Sensing-assisted high reliable communication: A transformer-based beamforming approach. IEEE Journal of Selected Topics in Signal Processing 18, 782–795 (2024). [79] Bayesteh, A. et al. Integrated sensing and communication (isac)—from concept to practice. Communications of Huawei Research 4–25 (2022). [80] Cui, Y., Cao, X., Zhu, G., Nie, J. & Xu, J. Edge perception: Intelligent wireless sensing at network edge. IEEE Communications Magazine 63, 166– 173 (2025). [81] Lyu, Z., Zhu, G. & Xu, J. Joint maneuver and beamforming design for uav-enabled integrated sensing and communication. IEEE Transactions on Wireless Communications 22, 2424–2440 (2022). [82] Meng, K. et al. Uav-enabled integrated sensing and communication: Opportunities and challenges. IEEE Wireless Communications 31,
14
97–104 (2023). [83] Zhang, J. A. et al. Enabling joint communication and radar sensing in mobile networks—a survey. IEEE Communications Surveys & Tutorials 24, 306–345 (2021). [84] Zhang, X., Peng, M. & Liu, C. Sensingassisted beamforming and trajectory design for uav-enabled networks. IEEE Transactions on Vehicular Technology 73, 3804–3819 (2023). [85] Dong, F. et al. Joint beamforming design for dual-functional mimo radar and communication systems guaranteeing physical layer security. IEEE Transactions on Green Communications and Networking 7, 537–549 (2023). [86] Khawaja, W. et al. A survey on detection, classification, and tracking of uavs using radar and communications systems. IEEE Communications Surveys & Tutorials (2025). [87] Jing, X., Liu, F., Masouros, C. & Zeng, Y. Isac from the sky: Uav trajectory design for joint communication and target localization. IEEE Transactions on Wireless Communications 23, 12857–12872 (2024). [88] Tong, P., Yang, X., Yang, Y., Liu, W. & Wu, P. Multi-uav collaborative absolute vision positioning and navigation: A survey and discussion. Drones 7, 261 (2023).
15