PaperlayerView exact revision on arXiv

Source-linked AI summary

Multi-tier Drone Architecture for 5G/B5G Cellular Networks: Challenges, Trends, and Prospects

Silvia Sekander, Hina Tabassum, Ekram Hossain

arXiv:1711.08407v1cs.NI

TL;DR

Existing work largely focuses on single-tier drone deployment, leaving multi-tier feasibility and its role in complementing terrestrial networks insufficiently studied. This paper reviews relevant challenges and innovations, then numerically evaluates multi-tier downlink spectral efficiency, finding gains under environment- and load-specific conditions.

  • Problem

    Existing research mainly optimizes single-tier drone deployment, leaving multi-tier feasibility and its trade-off between line-of-sight connectivity and path loss insufficiently analyzed.

  • Method

    The paper reviews drone-network research and challenges, then numerically evaluates multi-tier downlink spectral efficiency across drone tiers, altitudes, proportions, and urban environments.

  • Results

    Approximately 133% higher spectral efficiency is observed in high-rise urban environments with an optimal proportion versus single-tier small or big drones.

  • Takeaways & Limitations

    Multi-tier deployment can improve user spectral efficiency, but optimal drone proportions and altitudes vary substantially across urban environments.

  • Takeaways & Limitations

    More precise air-to-ground modeling incorporating temperature, wind, foliage, near-sea, and urban environments remains a future direction.

Abstract

from arXiv · show

Drones (or unmanned aerial vehicles [UAVs]) are expected to be an important component of fifth generation (5G)/beyond 5G (B5G) cellular architectures that can potentially facilitate wireless broadcast or point-to-multipoint transmissions. The distinct features of various drones such as the maximum operational altitude, communication, coverage, computation, and endurance impel the use of a multi-tier architecture for future drone-cell networks. In this context, this article focuses on investigating the feasibility of multi-tier drone network architecture over traditional single-tier drone networks and identifying the scenarios in which drone networks can potentially complement the traditional RF-based terrestrial networks. We first identify the challenges associated with multi-tier drone networks as well as drone-assisted cellular networks. We then review the existing state-of-the-art innovations in drone networks and drone-assisted cellular networks. We then investigate the performance of a multi-tier drone network in terms of spectral efficiency of downlink transmission while illustrating the optimal intensity and altitude of drones in different tiers numerically. Our results demonstrate the specific network load conditions (i.e., ratio of user intensity and base station intensity) where deployment of drones can be beneficial (in terms of spectral efficiency of downlink transmission) for conventional terrestrial cellular networks.

INTRODUCTION

The introduction motivates multi-tier drone networks as a way to support or complement 5G/B5G terrestrial cellular networks under diverse operational and network conditions. It frames the study around architecture feasibility, drone-tier design, deployment scenarios, associated challenges, and downlink spectral efficiency.

  • Motivation: Drones can support 5G/B5G broadcasting, point-to-point, and point-to-multipoint communication requirements.Applications include satellite news-gathering, live sports coverage, portable field monitoring, and video streaming.
  • Motivation: Drones can substitute for or complement terrestrial networks by serving shadowed, interfered, overloaded, damaged, idle-cell, and rural users.They can also offload users when new terrestrial infrastructure is not immediately economically or practically feasible.
  • Architecture: Different drones’ SWAP constraints create distinct altitude, communication, coverage, computation, and endurance capabilities, motivating multi-tier architectures.Low altitude platforms have low power and capacity, with payloads ranging from a few dozen grams to 5-7 kilograms and autonomy of 10 to 40 minutes.
  • Research questions: The study examines whether multi-tier drone networks are preferable to single-tier networks and how tier intensities and altitudes should be selected.It also asks when drones can improve existing terrestrial cellular networks.
  • Contributions: The article reviews drone-network innovations, identifies multi-tier and drone-assisted cellular challenges, and evaluates downlink spectral efficiency.Identified challenges include energy consumption optimization, interference management, and energy-aware and interference-related design.
  • Results: Optimal drone intensities vary across high-rise urban, suburban, and dense-urban environments, while deployment benefits depend on specific network-load conditions.The comparison concerns multi-tier versus single-tier drone networks and terrestrial cellular networks.

Deployment of Drone-Aided Cellular Network and Air Traffic Control Systems for Drones · Trajectory Planning and Mobility Control for Drones · Operational Altitude of Drones

Drone deployment should be matched to user load and active terrestrial base stations because offloading is not universally beneficial. Drone mobility and altitude must also account for operational constraints, regulatory limits, connectivity needs, and urban propagation conditions.

  • Deployment of Drone-Aided Cellular Network and Air Traffic Control Systems for Drones: Future traffic growth may overload cellular networks, while massive terrestrial BS deployment could alleviate overload.The paper frames drone deployment against simultaneous growth in user traffic and terrestrial infrastructure.
  • Deployment of Drone-Aided Cellular Network and Air Traffic Control Systems for Drones: Drone deployment should be opportunistic, considering both user load and active BSs.The appropriate deployment decision depends on the prevailing network load and terrestrial base-station activity.
  • Deployment of Drone-Aided Cellular Network and Air Traffic Control Systems for Drones: Offloading users from terrestrial BSs to drones may be ineffective because user-to-drone communication distances are typically large.The passage cautions that turning off terrestrial BSs does not always improve network operation.
  • Deployment of Drone-Aided Cellular Network and Air Traffic Control Systems for Drones: Offloading mechanisms can target spectral or energy efficiency based on interference, desired QoS, and terrestrial-BS traffic load.Determining the user-to-drone load ratio is identified as crucial for positive cellular-network assistance.
  • Trajectory Planning and Mobility Control for Drones: Optimal drone trajectories must satisfy air-corridor, connectivity, fuel, collision, terrain-avoidance, and aviation-regulation constraints.Regional, military, and civil-aviation regulations can limit drone altitude and coverage.
  • Trajectory Planning and Mobility Control for Drones: Rotary-wing drones can hover over coverage areas as static aerial BSs, eliminating the need for dedicated trajectory planning in typical cellular applications.This operating mode treats hovering drones as stationary aerial infrastructure.
  • Operational Altitude of Drones: SWAP constraints can restrict drone types to operational altitudes that may not suit every urban environment.High-rise users may need higher LoS connectivity, whereas sub-urban users may benefit more from path-loss reduction.
  • Operational Altitude of Drones: Higher drone altitude improves LoS connectivity by reducing reflection and shadowing, while lower altitude reduces path loss.The favorable altitude therefore depends on the environment and users’ connectivity requirements.

Terrestrial-Drone Interference … Backhauling Cellular Communication

The paper identifies interference, energy and endurance limitations, deployment cost, and backhaul feasibility as major challenges for integrating multi-tier drones with terrestrial cellular networks. It highlights tier-specific spectrum use, energy harvesting, and drone-enabled LoS backhaul as potential mitigation directions.

  • Terrestrial-Drone Interference: Interference management is crucial because drones operating in single or multiple tiers affect desired drone-user communications, with AtG channels and mobility intensifying the challenge.Mobility-induced Doppler shift can cause severe inter-carrier interference.
  • Terrestrial-Drone Interference: Using different frequency spectra across tiers, such as mm-wave for LAPs and RF for HAPs, can mitigate interference.
  • Energy Consumption of Drones: Drones can reduce terrestrial BS energy consumption by serving lightly loaded active users and allowing some BSs to become idle, although drone energy use may also be high.Energy-efficient mobility patterns should account for both hovering and continuous movement using appropriate energy consumption models.
  • Limited Endurance: LAP endurance is limited by SWAP constraints and rechargeable-battery dependence, making energy harvesting important without substantially increasing fuel-system mass or size.HAPs may harvest solar energy because of their larger dimensions, whereas LAPs may use ambient RF energy harvesting.
  • Cost, Security and Lack of Regulations: Deploying drone networks is costly because of drone purchases, accessories, maintenance, battery requirements, and potentially higher costs after aerial collisions.
  • Backhauling Cellular Communication: Backhauling all BSs may be infeasible at outdoor, remote, or hard-to-reach sites, while existing RF, microwave, and wired solutions incur interference, congestion, or higher cost.Drones can exploit LoS connectivity, but reliable backhaul requires high-capacity links with terrestrial wireless backhaul hubs.
  • Backhauling Cellular Communication: Cost-effective drone backhaul may combine unlicensed-spectrum technologies such as FSO and mmWave with traditional RF, while solar-powered unmanned balloons are another option.

EXISTING STATE-OF-THE-ART TECHNIQUES … Optimization of Drone Intensity

The paper reviews state-of-the-art optimization techniques for drone-assisted cellular networks, covering altitude, three-dimensional placement, energy efficiency, and drone intensity. Prior studies optimize coverage, revenue, congestion support, transmit power, throughput, and coverage performance through analytical and algorithmic approaches.

  • EXISTING STATE-OF-THE-ART TECHNIQUES: The review classifies prior drone-assisted cellular-network studies by objectives including altitude/location, energy-efficiency, and intensity optimization, comparing applications, solution approaches, and drone types.The review summary is provided in Table I.
  • Altitude Optimization for Drones: Prior work derives a UAV altitude that maximizes user coverage while accounting for environmental statistics, path loss, shadowing, and scattering.Man-made structures contribute additional path loss, modeled with a Gaussian distribution approximated by its mean.
  • Optimization of 3-D Placement of Drones: A 3-D UAV placement problem maximizes network revenue, defined as proportional to the number of drone-cell users, using bisection search.The algorithm jointly determines UAV altitude and coverage area.
  • Optimization of 3-D Placement of Drones: Three-dimensional drone placement is also optimized for macro-cell assistance during congestion, using an algorithm designed to serve users effectively.The cited work studies drones assisting macro cells by serving users during congestion.
  • Optimization of 3-D Placement of Drones: For mobile UAVs, prior work provides expressions for coverage probability and achievable rates while optimizing stop-point counts to minimize transmit power.The optimization concerns the number of stops made by a mobile UAV.
  • Optimization of Drone Intensity: In drone small-cell networks sharing spectrum with traditional cellular BSs, researchers calculate the optimal drone small-cell density for maximum DSC throughput.The drone small cells are placed at a limited height and coexist with traditional cellular base stations.
  • Optimization of Drone Intensity: For multiple UAVs with directional antennas, prior work maximizes coverage over a target area while ensuring minimum required transmit power.The study frames coverage optimization around a specified target area.

Traffic Offloading/User Association · Power Minimization of Drones · Summary

The paper reviews drone-assisted traffic offloading and power-minimization approaches, then examines whether multi-tier drone architectures are feasible and useful compared with single-tier and terrestrial networks. It focuses on balancing LoS connectivity and path loss across drone tiers under specified cellular-network assumptions.

  • Traffic Offloading/User Association: UAV traffic offloading is optimized through demand-dependent placement in heterogeneous macro- and small-cell networks.A neural cost model incorporates coverage, capacity, delay, and achievable LOS.
  • Power Minimization of Drones: Power minimization derives UAV locations within a given cell boundary using a facility location framework.The framework treats UAV transmit powers as transportation costs and adjusts altitudes and locations to reduce them.
  • Summary: Existing research mainly optimizes drone intensity and altitude for single-tier networks, leaving multi-tier feasibility largely unanalyzed.The paper identifies a trade-off between increasing LoS connectivity and path loss across different drone tiers.
  • Summary: The study investigates multi-tier drone architectures across urban environments and identifies scenarios where they may outperform single-tier drones or terrestrial cellular networks.This investigation directly evaluates when multi-tier deployment may be useful over both alternatives.
  • Summary: The modeled terrestrial network contains base stations and users distributed as homogeneous PPPs with intensities λt and λu, respectively.Each terrestrial base station transmits with fixed power Pt.
  • Summary: The two-tier drone network models big and small drones as homogeneous PPPs with intensities λm and λs at fixed altitudes hm and hs.Big and small drones transmit with powers Pm and Ps, respectively.
  • Summary: All drone tiers and terrestrial base stations share spectrum, while users associate according to maximum received signal power.Users sharing a serving transmitter receive channels with equal probability, and transmitters without associated users are turned off.

Path-Loss Model · Terrestrial Network: · Multi-tier Drone Network:

The terrestrial model uses distance-dependent path loss with unit-mean exponential small-scale fading. The multi-tier drone model accounts for free-space propagation, environment-dependent excessive loss, and altitude- and distance-dependent LoS/NLoS conditions.

  • Terrestrial Network:: Terrestrial links use channel power gain proportional to hd−α, where d is user–BS distance and α is the path-loss exponent.Small-scale fading power gain h follows an exponential distribution with unit mean.
  • Multi-tier Drone Network:: Drone RF signals incur free-space path loss before reaching urban environments, where foliage and buildings introduce additional excessive path loss.Excessive path loss is random and is commonly represented by its mean value ηϵ.
  • Multi-tier Drone Network:: For the first two propagation groups, ηϵ is treated as a constant obtained by averaging samples within a propagation group.Its values vary with carrier frequency and urban environment.
  • Multi-tier Drone Network:: Air-to-ground path loss combines free-space path loss with propagation losses associated with LoS and NLoS reception.Free-space path loss is evaluated using the standard Friis equation, with carrier frequency, light speed, and drone–user distance as inputs.
  • Multi-tier Drone Network:: LoS probability depends on drone altitude hk and horizontal drone–user distance ri, with k ∈{m, s}.The horizontal distance is defined from the drone and user coordinates.
  • Multi-tier Drone Network:: The LoS probability uses a modified Sigmoid S-curve whose parameters a and b depend on the urban environment.The model covers high-rise urban, dense urban, sub-urban, and urban environments.
  • Multi-tier Drone Network:: NLoS probability is defined as 1 − PLOS(hk, ri), complementing the LoS probability for each drone–user geometry.The considered air-to-ground model is described as simple, general, and applicable across multiple network settings.

Results and Discussions

The section defines downlink spectral efficiency through the SINR of a typical user and examines multi-tier drone-network performance as the proportion of small drones varies. The analysis considers macro- and small-drone tiers with distinct altitudes and a fixed total drone intensity.

  • Spectral Efficiency: Downlink spectral efficiency is defined as R = log2(1+SINR) for a typical user.The SINR expression incorporates received signal power, interference, and noise.
  • SINR Model: A typical user’s SINR is specified for association with either a macro-drone or small-drone base station.The formulation distinguishes drone tiers m and s and includes interference from other transmitters.
  • Multi-Tier Evaluation: λ = 10 evaluates downlink SE as a function of the proportion of small drones, with hm = 3000 m and hs = 150 m.The figure studies a multi-tier drone network under fixed drone intensity and tier-specific altitudes.

Multi-tier vs Single-tier Drones:

Multi-tier drone networks outperform single-tier deployments when drone proportions are optimized, balancing big drones’ stronger signals and interference against small drones’ connectivity advantages. The optimal proportions and altitudes depend on the urban environment and can substantially improve spectral efficiency.

  • Multi-tier vs Single-tier Drones:: Optimal proportions of big and small drones yield higher spectral efficiency than single-tier big- or small-drone deployments across all chosen environments.Big drones provide stronger signals through higher power, while small drones offer relatively better LoS connectivity; combining them balances signal strength and path loss.
  • Multi-tier vs Single-tier Drones:: 133% improvement is observed in high-rise urban environments, while dense urban multi-tier networks improve spectral efficiency by 75% versus small-drone tiers and 250% versus big-drone tiers.These improvements occur when the proportion of drone types is optimized.
  • Multi-tier vs Single-tier Drones:: Increasing the proportion of big drones improves signal quality but also raises interference, making the optimal big-drone proportion essential.Big drones contribute through high LoS probability and large transmission power, but their interference cost must be controlled.
  • Multi-tier vs Single-tier Drones:: Most environments favor a higher proportion of small drones, whereas high-rise urban environments require more big drones for stronger high-altitude LoS connectivity.The preferred mix therefore varies with the urban environment.
  • Multi-tier vs Single-tier Drones:: With an optimal drone proportion, dense urban environments provide better typical-user performance than high-rise environments, whose shadowing, reflections, and scattering are more severe.For any chosen small-drone altitude, proper small-drone proportions improve multi-tier spectral efficiency over single-tier networks by balancing interference; higher small-drone altitudes are preferable in high-rise environments.

Multi-tier Drone-Assisted Terrestrial Cellular Networks:

Multi-tier drones can improve terrestrial network spectral efficiency when user density is high relative to terrestrial base-station density and drone proportions are chosen correctly. Their benefits arise from offloading users despite added drone interference, whereas unsuitable single-tier proportions may not outperform terrestrial networks.

  • Performance conditions: High user-to-BS density enables multi-tier drones to improve terrestrial network performance by reducing traffic load per terrestrial BS.Higher user intensity reduces each terrestrial BS user’s channel share; drones provide additional associations and offload traffic.
  • Performance conditions: Higher LoS probability increases user association with drones, while terrestrial users gain channel share at the cost of increased drone interference.The passage identifies both user association and terrestrial-BS load reduction as mechanisms behind the improvement.
  • Performance conditions: With drones deployed in the correct proportion, interference-related performance loss for terrestrial-BS users is not significant.The result specifically concerns users associated with terrestrial BSs in the drone-assisted network.
  • Deployment trade-offs: Multi-tier drone-aided cellular networks outperform terrestrial networks when the small-drone proportion is chosen correctly, but single-tier co-channel deployments may not be beneficial.The cited examples identify small-drone proportions of 0.2 and 0.9 for big-drone and small-drone tiers, respectively, as potentially non-beneficial single-tier scenarios.

Extension to Multicast Systems

The proposed architecture can be viewed as a multicast system, where one source communicates simultaneously with multiple locations and capacity scales linearly with receiver count.

  • Extension to Multicast Systems: The proposed architecture can be visualized as a multicast system.
  • Extension to Multicast Systems: Point-to-multipoint communication uses multiple connecting paths from one location to multiple locations simultaneously.
  • Extension to Multicast Systems: System capacity increases linearly with the number of receivers; with transmission rate r and K users, achievable capacity becomes Kr.The passage states this relationship for multicast channels.

CONCLUSION AND FUTURE DIRECTIONS

The study finds that multi-tier drone architectures can improve user spectral efficiency over single-tier designs, while optimal drone proportions depend on urban environments. Future work targets cost, propagation accuracy, application-specific operation, sharing economics, and UAV-network virtualization.

  • Conclusion: Multi-tier drone architectures improve user spectral efficiency by balancing line-of-sight probability and path loss across different drone tiers.The optimal proportion of drones depends on the type of urban environment.
  • Future Directions: Heterogeneous drone deployment and spectrum selection should be optimized to reduce deployment cost and energy consumption.
  • Future Directions: Realistic propagation studies should investigate unlicensed FSO and mm-wave bands, including opportunistic mm-wave/RF selection to maximize spectral efficiency.
  • Future Directions: More precise air-to-ground channel models should incorporate temperature, wind, foliage, nearsea environments, and urban environments for accurate performance analysis.
  • Future Directions: Application-specific optimization of drone deployment, mobility, and operation is needed for converged broadband scenarios such as online video streaming and multimedia broadcasting.
  • Future Directions: Future research should examine third-party drone sharing economics and UAV virtualization integrated with cloud, web, and service-oriented network resources.Sharing should optimize network throughput against drone usage costs, while virtualization methods are needed for UAV-enabled 5G networks.
Loading 1711.08407v1…