Source-linked AI summary
Enhancing Vehicular Network Performance Through Integrated RSU and UAV Deployment
Muhammad Ismail, Syed Luqman Shah, Fazal Muhammad, Zeeshan Shafiq
TL;DR
Increasing vehicle density can exceed the finite resources of fixed roadside infrastructure. The paper evaluates on-demand auxiliary UAV assistance across terrestrial and heterogeneous baselines using coverage, aerial-link LoS probability, and SINR-based service decisions. Simulations show increased service capacity under high loads, including when an aerial node is already part of the primary infrastructure.
Problem
Growing vehicular demand can exceed fixed infrastructure resources, creating a need to improve service availability and throughput during temporary traffic surges.
Method
The paper evaluates UAVa in two configurations and uses coverage, UAV–vehicle LoS probability, and SINR threshold constraints to determine service.
Results
Auxiliary UAV assistance increases available service capacity under growing vehicular loads in both terrestrial and heterogeneous terrestrial–aerial configurations.
Takeaways & Limitations
On-demand aerial assistance can flexibly accommodate temporary increases in communication demand, including when the primary infrastructure already contains an aerial node.
Abstract
from arXiv · showhide
The increasing density of connected vehicles can place substantial pressure on fixed roadside infrastructure, particularly when the available communication resources become insufficient to accommodate temporary traffic surges. This paper investigates auxiliary unmanned aerial vehicle (UAV) assistance as a flexible mechanism for improving service availability and throughput in vehicular networks. Two representative network configurations are considered. In the first, an auxiliary UAV (UAVa) supplements two fixed roadside units (RSUs), whereas in the second, UAVa assists a heterogeneous infrastructure comprising one RSU and one UAV. Vehicle service is determined according to node coverage, the line-of-sight (LoS) probability of aerial links, and a prescribed signal-to-interference-plus-noise ratio (SINR) requirement. The resulting framework enables UAVa to accommodate eligible vehicles that cannot be adequately served by the primary infrastructure as the network load increases. Simulation results show that, under the considered configurations, the aerial nodes benefit from more favorable propagation conditions and achieve higher throughput than the fixed terrestrial RSU. Moreover, the introduction of UAVa increases the available service capacity under high vehicular loads, both for a purely terrestrial baseline and for a network already supported by an aerial node. These results demonstrate the potential of auxiliary UAV assistance as a flexible load-relief mechanism for capacity-constrained vehicular networks.
I. INTRODUCTION
Growing vehicular demand challenges stationary RSU-based networks, motivating UAV integration to provide flexible coverage and improve throughput and QoS. The paper considers auxiliary UAV assistance as a dynamic response to changing traffic conditions.
- Motivation: Stationary RSUs can underperform as traffic conditions, vehicle densities, and bandwidth-intensive applications increase.The motivation includes reliable connectivity, effective data transfer, interference minimization, and ultra-low latency.
- Motivation: UAV-assisted networks offer LoS links, dynamic deployment, on-demand coverage, scalability, and adaptability to high-traffic or emergency locations.The paper presents these properties as reasons to supplement existing RSUs with aerial base stations.
- Related work: Existing heuristic and data-driven RSU deployment approaches face scaling or adaptability challenges in dynamic vehicular networks.The related-work discussion identifies genetic and particle-swarm methods as potentially suboptimal at scale, while data-driven methods may not fully capture network dynamics.
- Motivation: Dynamic UAVa deployment is proposed to assist existing infrastructure during emergencies and reduce packet loss and communication disruptions.The passage associates suitable UAVa deployment with more reliable and stable vehicular communication.
- Contributions: The study evaluates two configurations: UAVa assisting two fixed RSUs or assisting one RSU and one UAV.The contribution framework also includes a coverage- and channel-aware association procedure and throughput evaluation under varying vehicle loads.
II. SYSTEM MODEL
The system models a time-slotted vehicular network with finite infrastructure resources and introduces UAVa as an on-demand node for vehicles that primary infrastructure cannot adequately serve. Two configurations differ in whether the primary infrastructure is purely terrestrial or heterogeneous.
- System model: Vehicles arrive according to a Poisson process, while finite communication resources can prevent infrastructure from serving all active vehicles at high loads.The active-vehicle set at time slot t is denoted V_t, with λ_v representing the mean arrival rate.
- Auxiliary assistance: UAVa provides additional service on demand when initially deployed nodes cannot adequately serve eligible vehicles.This auxiliary role produces the two network configurations examined in the paper.
- Network configurations: Scenario I contains two fixed RSUs and UAVa, whereas Scenario II contains RSU1, UAV1, and UAVa.The figures identify the corresponding infrastructure compositions for the two scenarios.
A. Scenario I
Scenario I uses two fixed RSUs as primary infrastructure and deploys UAVa when terrestrial resources become insufficient. Its elevated and flexible aerial position supports additional service and traffic redistribution during congestion, while Scenario II provides the heterogeneous comparison case.
- Scenario I: Scenario I comprises RSU1, RSU2, and UAVa, with the RSUs serving vehicles within their respective coverage regions under nominal traffic.Increasing vehicle density can make terrestrial demand exceed the finite resources of both RSUs.
- Scenario I: UAVa serves vehicles that the two RSUs cannot accommodate when terrestrial service demand exceeds available resources.Its elevated position and deployment flexibility enable it to complement fixed infrastructure during congestion.
- Scenario II: Scenario II uses RSU1 and UAV1 as primary nodes, so one aerial serving node is already present before UAVa is employed.The scenario tests whether additional on-demand aerial assistance remains beneficial in a heterogeneous terrestrial–aerial baseline.
III. PROPOSED SCHEME
Both scenarios use the same association principle: candidate nodes must cover a vehicle, after which propagation conditions and SINR determine channel admissibility. UAVa is invoked when the primary infrastructure cannot adequately serve an eligible vehicle.
- Association procedure: The two scenarios differ only in their primary serving nodes: RSU1 and RSU2 in Scenario I, versus RSU1 and UAV1 in Scenario II.UAVa is included as the auxiliary node in both candidate serving sets.
- Auxiliary activation: UAVa is invoked when the primary infrastructure cannot adequately serve an eligible vehicle.This rule applies to both network configurations.
- Coverage filtering: A node can serve a vehicle only when the vehicle lies within that node’s coverage region.The coverage test uses the node–vehicle distance and coverage radius to eliminate geographically unsuitable candidates.
- Channel admissibility: After coverage filtering, propagation conditions and the resulting SINR determine channel admissibility.This sequential procedure first removes geographically unable nodes, then evaluates communication-channel suitability.
A. Air-to-Ground Link Model
The model distinguishes terrestrial and aerial propagation, using geometry-derived LoS likelihood for UAV links and separate large-scale gain models for RSU and aerial links.
- For UAV U and vehicle v, the LoS model uses the elevation angle and a prescribed minimum LoS-probability threshold.The threshold is denoted pth and lies in [0, 1].
- The elevation angle is determined by UAV and vehicle heights together with their horizontal separation.The geometry uses ρt_U,v, hU, and hv.
- Increasing elevation angle generally increases the probability of establishing an LoS connection.
- Terrestrial RSU links use a separate large-scale channel-gain model based on distance, path-loss exponent, and reference attenuation parameters.The ground-to-ground path-loss exponent is ηR, while βR collects reference-distance gain and other attenuation factors.
- Aerial links distinguish LoS and NLoS propagation states, with NLoS incurring stronger attenuation than LoS.
B. Channel Selection and SINR Constraint
Channel selection evaluates each candidate link’s received SINR on available orthogonal channels and admits a channel only when the SINR requirement is satisfied.
- The network provides a set C of Nch orthogonal channels for candidate serving nodes and vehicles.
- Received SINR is computed for candidate node ℓ, vehicle v, and channel c using signal power, receiver noise, and aggregate co-channel interference.
- The interference term includes simultaneously active transmitters using the same channel during the time slot.
- Transmissions on different orthogonal channels do not contribute to co-channel interference.
- A channel is admissible only if its instantaneous SINR meets γth; otherwise, the vehicle attempts another available channel.The threshold tests resulting signal quality rather than removing interference.
C. Instantaneous Throughput
The framework converts admissible link SINR into instantaneous throughput, aggregates successfully served vehicles by node and channel, and sums across active nodes for network-wide throughput.
- Once association and channel admissibility are established, link throughput is computed from SINR over the bandwidth B allocated to one channel.
- A vehicle contributes to served throughput only when coverage, any required aerial-link condition, and SINR conditions are simultaneously satisfied.
- Aggregate instantaneous throughput is formed from vehicles successfully served by each node on each channel.
- Network-wide throughput sums the per-node throughput across all active serving nodes to compare load handling and quantify UAVa’s additional service.
IV. SIMULATION RESULTS AND DISCUSSION
The simulations vary active-vehicle load and determine service using coverage, LoS probability, and SINR conditions. They use fixed channel-access and reception parameters, with settings summarized in Table I.
- Simulation setup: Network load is varied by changing the number of active vehicles, while successful service requires coverage, LoS-probability, and SINR conditions.For each successfully served vehicle, instantaneous throughput is computed according to (10).
- Simulation setup: The considered network uses Nch = 10 orthogonal channels and q = 0.2 channel-access probability.
- Simulation setup: Successful reception requires a minimum SINR of γth = 10 dB.
- Simulation setup: Aerial links require a minimum LoS probability of pth = 0.8, consistently with (4).Node locations and remaining simulation parameters are summarized in Table I.
A. Scenario I
In Scenario I, UAVa supplements two fixed RSUs as load increases. Its elevated position supports more eligible links and enables additional service and load redistribution when terrestrial resources become insufficient.
- A. Scenario I: RSU1 and RSU2 show similar throughput behavior, but their throughput becomes increasingly constrained as active-vehicle load grows.Each RSU serves only vehicles meeting its coverage and link-quality requirements.
- A. Scenario I: UAVa achieves higher throughput over the considered load range because its elevated position provides more favorable propagation geometry.A larger fraction of candidate UAVa–vehicle links can satisfy the LoS requirement p_LoS,U,v ≥ pth.
- A. Scenario I: UAVa creates additional service opportunities for vehicles that fixed RSUs cannot accommodate.These opportunities arise when candidate aerial links satisfy the required LoS condition.
- A. Scenario I: UAVa provides an additional degree of freedom for load redistribution while fixed RSUs continue primary terrestrial service.This complementary behavior is the operating principle of Scenario I.
B. Scenario II
In Scenario II, UAVa assists a primary infrastructure containing one RSU and one UAV. The results show that aerial assistance remains beneficial by adding service capacity beyond the existing aerial node.
- B. Scenario II: RSU1 exhibits the lowest throughput among serving nodes, while elevated UAV1 supports a larger traffic load under the considered network geometry.UAV1 benefits from more favorable propagation conditions for a larger subset of vehicles.
- B. Scenario II: UAVa provides an additional gain beyond that already obtained from UAV1.It accommodates vehicles satisfying coverage, LoS-probability, and SINR requirements when RSU1 and UAV1 are insufficient.
- B. Scenario II: Aggregate throughput increases through the contribution of RSU1, UAV1, and UAVa.
- B. Scenario II: Across both scenarios, UAVa is associated with additional service availability under increasing network load, including when UAV support is already present.
V. CONCLUSION AND FUTURE WORK
The paper evaluates auxiliary UAV assistance across two infrastructure configurations, using coverage, aerial-link LoS characteristics, and SINR requirements to determine service eligibility. Simulations indicate that UAVa increases throughput and service opportunities as vehicular load grows, while the study remains bounded by fixed UAV locations and a predefined service procedure.
- Two configurations evaluate UAVa assistance: two fixed RSUs, and a heterogeneous infrastructure comprising one RSU and one UAV.
- Vehicle service is determined using coverage constraints, aerial-link LoS characteristics, and an SINR requirement.
- UAVa provides additional service opportunities and increases available network throughput as vehicular load grows.The reported benefit is attributed to favorable propagation geometry.
- Auxiliary UAV assistance remains useful when an aerial node is already part of the primary infrastructure, not only with a purely terrestrial baseline.
- The study assumes fixed UAV locations and a predefined service procedure, motivating joint optimization of placement, trajectory, and resource allocation.The proposed extension would also account for spatial and temporal traffic distributions.