2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

Urban Air Mobility:

Vision, Challenges and Opportunities
2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR) | 978-1-6654-7640-9/23/$31.00 ©2023 IEEE | DOI: 10.1109/HPSR57248.2023.10148014

Debjyoti Sengupta and Sajal K. Das

Department of Computer Science
Missouri University of Science and Technology
Rolla, MO 65409, USA
E-mail:{dsry8, sdas}@[Link]

Abstract—Urban Air Mobility (UAM) involving piloted or This calls for innovation in alternative modes of transport
autonomous aerial vehicles, is envisioned as emerging disruptive that will not only enhance existing transportation capacity in
technology for next-generation transportation addressing mobil- congested urban areas but also provide economically feasible
ity challenges in congested cities. This paradigm may include
aircrafts ranging from small unmanned aerial vehicles (UAVs) solutions and services with reliable operations. Focusing on
or drones, to aircrafts with passenger carrying capacity, such as urban air mobility (UAM) as an emerging technology to help
personal air vehicles (PAVs). This paper highlights the UAM realize such a vision, this paper attempts to highlight some of
vision and brings out the underlying fundamental research the underlying research challenges and opportunities from the
challenges and opportunities from computing, networking, and networking and computing perspectives.
service perspectives for sustainable design and implementation of
this promising technology providing an innovative infrastructure The contributions of this paper are summarized as follows.
for urban mobility. Important research questions include, but are First, we describe various types of aircrafts envisioned for
not limited to, real-time autonomous scheduling, dynamic route urban air mobility. Next, we discuss the infrastructure and
planning, aerial-to-ground and inter-vehicle communications, operations for managing the airspace and air-traffic in UAM
airspace traffic management, on-demand air mobility, resource networks. This is followed by challenges and opportunities
management, quality of service and quality of experience, sensing
(edge) analytics and machine learning for trustworthy decision in the underlying communications, networking, and on-board
making, optimization of operational services, and socio-economic computations (edge analytics) on UAM. We then discuss ways
impacts of UAM infrastructure on sustainability. to minimize adverse environmental impacts on sustainable
Index Terms—Urban air mobility, vertical take-off and landing design and usage of UAM technology. Different services
vehicles, space-air-ground integrated networks. including on-demand air mobility are also presented. Security
and privacy issues are highlighted for safe operations of this
I. I NTRODUCTION
new mode of transport. Finally, we outline the socio-economic
With the ubiquity of the Internet, wireless mobile commu- and policy impacts, and conclude the paper.
nications, smart sensors (including smartphones), internet of
things (IoT) and pervasive technologies, citizens from diverse II. UAM A IRCRAFTS AND AUTONOMY
demographic regions around the world have access to vast The UAM scenario may encompass vehicles ranging from
opportunities than ever. Such development has been instru- small unmanned aerial vehicles (UAVs) or drones, to personal
mental for better quality of life, increased job employments, air vehicles (PAVs) with larger payload capacity and more
and marketability of essential commodities and services in robust autonomous capabilities [2]. Autonomous capabilities
daily life, thus resulting in unprecedented economic growth and features developed for terrestrial vehicles are discussed as
to the society at large [1]. However, significant challenges are to their applicability to the airspace. Challenges and oppor-
posed for sustainable development and maintenance of proper tunities in developing novel methods of object detection and
infrastructures that ensure smooth and reliable operations and collision avoidance are also briefly discussed.
services catering to the growing needs of modern lifestyles. A wide variety of commercial aircrafts exists today with
To this end, a key challenge is to provide sustainable mobility varying payloads. This works well for the aviation industry
and transportation solutions, particularly in urban areas. where flight schedules, routes and demands are quite well
Increased urbanization and shift to urban lifestyles bring defined, determining the allocation of aircrafts for operations.
forth a unique set of fundamental challenges and opportunities However the lack of specific schedules in the anticipated UAM
in computing and networking research. Some of the basic market to pre-determine the type of flights to be undertaken
needs are the availability of affordable and sustainable housing leads to the ‘on-demand’ nature of flight scheduling, where the
and the development of a stable, manageable, safe and efficient popularity of flights and/or resulting congestion might change
transportation network to aid the movement of the resident frequently over short intervals (days or weeks), making vertical
and commuter population. With the urban sprawl of a city, take-off and landing (VTOL) and short take-off and landing
the existing transportation capabilities are often saturated or (STOL) vehicles the main contenders for the UAM market
insufficient to support the increased load on mobility needs [2]. (see Figure 1 for illustration).

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
978-1-6654-7640-9/23/$31.00 ©2023 IEEE 204
 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

A recent surge has been noted to the UAM airspace to fully realize the benefits of this new
on research and development of transportation paradigm [2]. Two most common terms used in
VTOL (eVTOL) vehicles involv- the literature are Vertiports and Vertistops, catering to VTOL
ing major companies like Airbus, aircrafts effective in congested urban areas. Vertiports have
Volocopter, and Terrafugia. Dif- much larger capacity to handle multiple UAM vehicles, anal-
ferent types of vehicles that can ogous to the existing airports, whereas Veristops are similar
potentially be used in UAM are to single helipads, catering only to one aircraft at a time.
described in [3]. Vehicles with The design and construction of Vertistops and Vertiports will
VTOL capabilities include City not not only depend on the precise locations for passenger on-
Airbus and ALIA-250, a joint ef- and-off boarding, but also how the UAM vehicle airspace is
fort by Blade and Beta [4], which designed and managed. Since most air taxis are expected to
conducted the first test flight in be VTOL enabled, the landing units can be built as docking
February 2023 in New York City. stations with multiple ports stacked on top of one another, for
The general consensus is that ini- better utilization of vertical space. The design will also depend
Fig. 1. From top: PAV (Volo-
copter), STOL (Just Aircraft), tially the UAM market will be on the altitude of the airspace in which the UAM vehicle will
VTOL (Joby Aviation) dominated by piloted aircrafts, operate. The capacity of each station depends on the nature of
eventually leading to fully autonomous vehicles with collision UAM flights, requiring efficient allocation and management of
avoidance and real-time route maintenance as the top priority. the airspace (resource), discussed next.
Collision avoidance by terrestrial autonomous vehicles are
based on object detection algorithms relying mostly on vision- IV. UAM A IRSPACE AND T RAFFIC M ANAGEMENT
based techniques, such as YOLOv2 used in self-driving cars A critical issue in developing UAM ecosystem is efficient
with the help of monocular cameras and LiDAR sensors [5]. utilization of available airspace. This section discusses if
Monocular cameras provide shape and texture information some of the traditional methods for airspace and air-traffic
for identifying elements that are important to observe while management are applicable to UAM scenario. An important
driving on road, but will be absent in UAM flight scenarios. challenge is how to choose efficient routes in real-time from an
It is also known that object detection capabilities are heavily origin to a destination given such constraints as airspace sector
dependent on the resolution of the camera and ambient condi- capacity, adverse weather, traffic congestion, or blockage of
tions, and the model does not perform well during night time some routes. Efficiently solving this multi-objective problem
nor rainy conditions. Moreover, YOLOv2 have limitations in will help design optimal scheduling and routing algorithms.
adverse weather conditions and reduced visibility [6]. The first step towards operating a UAM fleet is to efficiently
Therefore, reliance on other modalities than only vision allocate suitable airspace. One approach is to segregate the
must be a fundamental design consideration for autonomous available airspace into different regions (sectors) for different
UAM flights, leading to better perception of the environment. types of aerial vehicles. Since air taxis are typically expected
This implies redesign of sensors and algorithms for air taxi to fly at a lower altitude than conventional airlines or drones,
(PAV) operations. While other modalities like seismic and efficient planning of limited airspace is crucial. The Federal
acoustic signature (e.g., vibration and noise) can be used for Aviation Authority (FAA) is developing Unmanned Aircraft
object detection in terrestrial vehicles [7], [8], the aeroacoustic System Traffic Management (UTM), while EUROCONTROL
signature [9] can be considered for sensing operation in UAM. is developing U-Space [2], [11]–[13]. The U-Space Concept
Accurate weather predictions are also extremely important for of Operations (ConOps) divides the airspace into three types
safe operations [10] of UAM. X, Y, and Z based on the services offered and the entry/access
From computing viewpoint, creation of reliable datasets requirements. Type X does not provide conflict resolution and
is important for accurate object detection and tracking in the remote pilot has full responsibility of flight operation. Type
airspace. Existing datasets are mostly based on classifying Y supports only pre-flight or offline conflict resolution by
objects (humans, lane markings) encountered in ground trans- coordinating flight plans ahead of time. Type Z offers both
portation. Datasets with elements encountered in the air like pre-flight and in-flight conflict resolution using the position
other vehicles, birds, and clouds are essential. For example, and motion of other aircrafts.
identifying clouds as obstacles by air taxis will require con- In addition to planning and allocation of airspace, efficient
siderable training of machine learning models as clouds are path planning of air taxis from source to destination depends
not easily distinguishable from the sky background. on a variety of factors that are often dynamic in nature. In
traditional air traffic controller (ATC) concept, for safety every
III. UAM I NFRASTRUCTURE AND O PERATIONS flight is manually monitored by air traffic personnel for run-
This section presents the infrastructure design aspects for time operations, such as re-routing due to weather, elevation
UAM traffic network, which depends on the types of aircrafts adjustment, or emergency actions [14]. If UAM is realized
and their take-off and landing capabilities. Key positioning of with autonomous (or piloted) air taxis, safe flight operations
the take-off and landing areas and localization of passengers and minimum separation between air taxis can be assured by
and vehicles are essential for smooth and convenient access employing techniques proposed for UAVs in [14], where the

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
205
 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

entire usable airspace is divided into sectors, similar to cellular at varying altitudes ensuring safe transition of air vehicles from
wireless communication networks. This will help reduce the one layer to the other. The mobility management aims to de-
current load of air traffic personnel as the expected volume velop novel techniques for efficiently switching tracks within
of flights in the UAM scenario will be much more, making it the same layer or across layers in the airspace, analogous to
impossible to manually handle all run-time flight requests. horizontal/vertical hand-offs in wireless networks [21], [22].
This model can further be extended to include additional For autonomous vehicles, sensors and IoT will detect obstacles
parameters like the closure of certain sectors (e.g., due to high and other elements in upper and lower strata of 3D airspace.
priority passengers passing or flight route reservation for air With the advent of any new technology like UAM, there
ambulance), as in ground transportation. Likewise, in the aerial is tremendous opportunity for interdisciplinary collaborations
mode of transport, there will be designated skyways similar to across research community and industry with a goal to pro-
roads, for proper monitoring and surveillance of the airspace, vide affordable and robust solutions with improved customer
thereby decreasing accidents due to uncertainty related to the experience. As autonomous air taxis continue daily commutes,
unplanned and unrestricted air vehicle paths. the model will learn over time. Vehicles belonging to different
service providers can collaborate and share data for improving
V. UAM C OMMUNICATIONS AND N ETWORKING the model and refinement of the technology. This is where
When the UAM airspace becomes sufficiently crowded with federated learning (one kind of distributed machine learning)
vehicles, they will not only serve as a new mode of passenger can be useful to preserve privacy of customer’s sensitive data.
mobility or cargo transport, but also as access points or relays Thus, individually trained models from different companies
to aid in better integration and adoption of space-air-ground could be aggregated to create a more robust model, offering
integrated networks (SAGIN), as illustrated in Figure 2. better value to the customers without compromising their data.
Similar to static roadside units (RSUs) in autonomous
terrestrial vehicular networks, the UAM scenario will support
air taxi to ATC infrastructure (V2I) communications and
air taxi to air taxi (V2V) communications. Such networked
environment is vulnerable to failures and malicious attacks
(see Section IX for details). Due to the heterogeneous nature
of 5G networks, fair allocation and efficient reuse of resources
(wireless bandwidth) will play important roles [23], [24].
VI. O NBOARD C OMPUTATION – E DGE A NALYTICS
Similar to IoT, UAV and vehicular networks, how much
onboard computation and edge analytics are feasible in UAM
vehicles with limited resources (energy, bandwidth) for real-
time scheduling and route planning [25]? For a comprehensive
survey on edge computing driven IoT, see [26].
Computation offloading using federated learning (FL) is
used in [27] to jointly optimize latency and energy con-
sumption in vehicular edge computing. A similar approach
can be used in UAM where autonomous vehicles requiring
Fig. 2. UAM as part of space-air-ground integrated network (SAGIN)
huge computations while minimizing both latency and energy.
Demanded by the volume of UAM users and data generated Like RSUs on roads, the UAVs supporting inter-air taxi
by them, cellular wireless communications are expected to communications will act as edge computers, thereby reducing
dominate UAM operations, spawning new research ranging energy consumption of air taxis. The objective is not only to
from adaptation of 5G/6G signals for aeronautical use [15], to perform computations but also aggregate the models of air
securing such communications [16]. Aerial vehicles will act as taxis belonging to different service providers to generate a
servers or communication relays, thereby reducing the shadow more robust model. The updated aggregate model can later
effects of buildings [17]–[19]. Similar to UAVs, air vehicles be dispatched to all air-taxis coming in proximity of the
may also act as sensor data collectors from the ground [20]. edge units at pre-scheduled times essentially performing an
As regard to 5G networking, one may explore if peer-to- update of the model on-board the vehicle. Following [28], an
peer communication is effective between the air taxis. Another important research problem is to apply distributed learning to
question is: how to optimize the minimum separation distance connected vehicles to acquire data at scale.
between these aerial vehicles for efficient communications yet Furthermore, the UAVs acting as substitutes for RSUs in the
guaranteeing safe operations? For smooth traffic movement aerial domain can potentially be used to mitigate important
within the airspace, the UAM paradigm of commuter transport information about the current state of the airspace they are
offers the opportunity of utilizing different layers within the placed in (see Fig. 3). In case of a skyway closure due to
same airspace. They can be used as parallel routes for transport reasons mentioned in Section IV, this information can be sent

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
206
 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

from a central server like ATC to all the UAVs relaying the For eVTOLs, efficient
information to the air taxis as and when they pass by the route planning must not
units. It can also be directly transmitted to the vehicles through only depend on conven-
services like GPS, but the inclusion of supporting UAVs adds tional power sources (bat-
fault tolerance in potentially critical situations (see Figure 3). teries), but also consider al-
Edge analytics not only ternate energy (e.g., solar)
improves computation ef- given the vehicles will be in
ficiency but also flight the air most of the time. To
path planning for a trip. compute an optimal path
Besides sector constraints from source to destination,
mentioned earlier, data of- factors like weather and
floading and uploading to sector blockage can be con-
and from the UAV edge sidered, in addition to the
units require energy. Hence number of hops and time. Fig. 4. Calculating optimal path to the
the flight path can be op- Assuming sunlight is a destination to maximize solar charging
and minimize distance: shortest path
timized not only in terms factor, under good weather takes 7 sector hops, increasing to 9 us-
of shortest path measured Fig. 3. UAVs acting as relays to pass in all sectors and no satu- ing sectors with sunlight while charging
in the number of sectors, on important updates the from central ration in any of the sectors, the VTOL aircraft at the same time.
server (Computing Cloud) on ground
but also considering energy the multi-objective optimization goal will be to cover maxi-
required for edge computation at various waypoints [25]. Due mum sectors with proper sunlight while selecting the shortest
to the dynamic nature of communication (as vehicles are path. To this end, dynamic route planning algorithm proposed
mobile), data offloading to a computing cloud can employ in [14] for UAVs can be suitably adapted. This approach will
deep reinforcement learning and deep meta Q-learning in such allow the air vehicle to be charged en route, and reduce the
a large scale mobile edge computing scenario [29], [30]. frequency of electric charging. thus saving power (see Fig. 4
To improve positioning or localization accuracy of air taxis for illustration). Additionally, distributed V2V charge sharing
in UAM, range-free localization algorithms can be employed as in electric vehicles can also be potentially exploited [36].
by using UAVs as mobile anchors [31], which relay informa-
tion received from air taxis passing by them to the ground VIII. UAM S ERVICES - O N D EMAND M OBILITY
units for better tracking. Thus, even if GPS does not function
properly due to reasons like bad weather, a direct beacon from This section summarizes pros and cons of different services
the air taxi sent to UAV as mobile anchor can be relayed to to be offered for UAM along with the computing and net-
the ground stations keeping track of vehicle movements in the working issues. Similar to ride-share services like Uber for
airspace. In [32], we developed methods for reducing localiza- ground transportation, the UAM service will be available in
tion errors in some range-based algorithms that can prove to be future through various platforms. On demand mobility (ODM)
useful for UAM vehicle localization and relative positioning can offer shorter travel times with increased carrying capacity.
with respect to other vehicles during flight. Approaches similar Existing ride-share services mostly follow the transportation
to federated learning for collaborative positioning of internet network companies (TNC) model [2], which may not be
of vehicles [33], and localization methods for IoT using suitable for UAM scenario initially, as it will require an
compartmental model [34] can be explored for air taxis. abundance of eVTOL aircraft owners and qualified pilots.
The concept of transportation service providers (TSPs) will
VII. S USTAINABLE D ESIGN OF UAM offer a more convenient solution by involving companies
providing air taxi services with a fleet of aircrafts and certified
This section discusses how to reduce adverse environmental pilots. This approach is scalable with autonomous aircrafts,
threat and ecological footprint of specific technologies (e.g., ultimately bringing down the service cost and making them
propulsion systems, efficient batteries) in UAM design and widely accessible. The timeline for realizing this vision de-
usage. Electric air taxis will significantly reduce harmful emis- pends on socio-economic factors (see Section X) and the pace
sions compared to combustion-based commercial aircrafts. of research addressing domain-specific challenges.
Although most of the UAM aircrafts are expected to be of To make ODM a viable and useful concept to the users,
eVTOL types, the power requirements are expected to be much proper integration of aerial services with existing ground
higher than ground vehicles due also to the energy required to transport networks is needed, involving efficient planning of
hover at a particular place. Therefore, an active research area sky routes in tandem with ground vehicular routes, thereby op-
is to develop more energy efficient power sources (batteries) at timizing the source to destination transit. Thus, UAM must be
reduced costs. The industry standard has mostly been lithium- viewed as an alternate mode of transport and a complementary
ion batteries for electric vehicles. Recently, the JAC Group in service provider toward overall solution for urban mobility.
China unveiled the world’s first sodium-ion battery powered While ODM may lead to a decrease in car ownership due to
car, the Hua Xianzi meaning ‘Flower Fairy’ [35]. cost and journey time, major automobile manufacturers are

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
207
 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

likely to eventually provide air mobility solutions, leading to communication channels and UAM applications providing on-
novel pricing and revenue models. demand mobility services must be robust enough to handle
To support ODM needs of such intrusions. This is particularly important during early
users, the aerial vehicles must stages of UAM deployment when the aircrafts may not be
be on standby at strategically fully autonomous. If a pilot is compromised or has malicious
placed hubs. The hub nearest intent, proper tracking will ensure safety and apprehend rogue
to the customer requesting a individuals upon landing. The UAM service apps may have an
ride can deploy an aircraft to option for passengers to hit a panic button that will transfer the
minimize the waiting time. In vehicle control to a remote operator for smooth landing at a
Fig. 5, the hubs H3 and H5 secured location. High standards of security is instrumental for
and the optimal flight path wider acceptance and market penetration of UAM applications.
between A and B must be
selected prior to the deploy- X. S OCIO -E CONOMIC AND P OLICY I MPACTS
ment of air taxi. Due to en-
vironment dynamics (wind or Although UAM is a promising technology for air mobility,
Fig. 5. Air taxi deployed from hub
H3 as distance d3=min(d1,d2,d3). rain) and other factors, the
full realization of this vision will depend on various socio-
After dropping customer at B, the path planning algorithm will economic factors. At the social front, the most challenging
vehicle proceeds to hub H5 for dock- be dynamic, employing simi- hurdle to overcome is the acceptance by the customers. Riding
ing as d5=min(d4,d5,d6)
lar approaches to [10], [14]. on an autonomous air vehicle (without human pilot) is a major
Machine learning methods can be applied to increase service concern for many. There exist some studies on customers’
provider’s revenue by allocating variable number of vehicles to willingness to avail air taxi services [42]–[44]. Safety is
the hubs based on local demands during the day. A model for revealed as the most important factor for which proper policies
allocating vehicles to the hubs will learn over time which hub must be in place to harness the full potential of this technology.
locations need to serve more customers at specific times of the With gradual progress of UAM services, the technology can
day, and accordingly divert fleets on standby mode. Such an become a popular alternative to current mobility solutions in
action and reward scenario can be formulated as reinforcement busy urban areas. From economic viewpoint, there is a huge
learning, in which the state of the intelligent agent can be business potential. Intelligent algorithms can help improve rev-
modeled with location and time of the day, while an action enue generation under UAM transport scenarios, in presence
will be to reach a certain hub at a given time of the day. The of uncertainty. For example, the selection of optimal sites for
destination hub location and time will constitute the resulting passenger pick-up based on aerial vehicle’s payload capacity,
(terminating) state having a utility (e.g., the number of flight can be formulated as a Knapsack problem.
requests received). The utility value will change over time, and Computing technology can further aid in easier adoption of
the agent will identify the hubs (states) with maximum utility. this paradigm by reducing privacy concerns like blocking in-
flight passengers from viewing people’s homes when flying
IX. S ECURITY AND P RIVACY I SSUES over urban settings using augmented reality in aircraft win-
dows. Trustworthy and explainable AI will also have a positive
Security is extremely important for communications, net- impact on UAM services.
working, and computing in UAM services. A malicious agent
can launch manipulation-based attacks by gaining control of XI. C ONCLUSION
communication channels [37], sensors [38] or other critical
components of airborne vehicles, and injecting false data Moving into a new era of urban air mobility integrated
to impede proper functioning of flight operations. Malicious with satellite, ground and UAV networks presents a plethora
agents can cause further damage by sharing sensitive infor- of unique challenges and opportunities in terms of new
mation to unauthorized entities, thus posing severe privacy technological innovation and computing and communications
concerns for UAM customers as well as service providers. research. This paper provides a snapshot of key research
Analogous to authentication and secure V2V communi- issues that must be tackled at the design and operational
cations in the Internet of vehicles [39], [40], developing phases towards the realization of UAM technology. Impor-
secure UAM communication protocols will lead to safe and tant research topics include autonomous aircraft design and
reliable smart air transportation. Recent advances in securing infrastructure development, real-time scheduling and dynamic
autonomous terrestrial vehicles using collaborative AI and route planning, aerial-to-ground and inter-vehicle communi-
Blockchain techniques [41] can potentially be deployed to cations, airspace traffic management and on-demand air mo-
improve the security of unmanned personal aerial vehicles. bility, resource management, quality of service and quality of
There also exist possibilities of location trailing attacks by experience, edge (sensing) analytics and machine learning for
tracking customer flight paths and potentially creating profiles trustworthy decision making, optimization of services offered,
based on the places visited frequently and other personal data. socio-economic and policy impacts of UAM technology on
This can eventually lead to targeted attacks, implying the environmental sustainability and user acceptance.

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
208
 2023 IEEE 24th International Conference on High Performance Switching and Routing (HPSR)

R EFERENCES [23] A. Pratap, R. Misra, and S. K. Das, “Maximizing fairness for resource
allocation in heterogeneous 5G networks,” IEEE Transactions on Mobile
Computing, vol. 20, no. 2, pp. 603–619, 2021.
[1] J. Hjort and C. Sacchetto, “Can internet access lead to improved eco- [24] A. Pratap and S. K. Das, “Stable matching based resource allocation for
nomic outcome?,” [Link] service provider’s revenue maximization in 5G networks,” IEEE Trans.
internet-access-lead-improved-economic-outcomes, 2022. on Mobile Computing, vol. 21, no. 11, pp. 4094–4110, 2022.
[2] J. Pons-Prats, T. Živojinović, and J. Kuljanin, “On the understanding [25] A. Khochare, Y. Simmhan, F. B. Sorbelli, and S. K. Das, “Heuristic
of the current status of urban air mobility development and its fu- algorithms for co-scheduling of edge analytics and routes for UAV fleet
ture prospects: Commuting in a flying vehicle as a new paradigm,” missions,” in IEEE INFOCOM, 2021. (Extended version to appear in
Transportation Research Part E: Logistics and Transportation Review, IEEE/ACM Transactions on Networking, 2023).
vol. 166, p. 102868, 2022. [26] K. Kong, J. Tan, J. Huang, G. Chen, S. Wang, X. Jin, P. Zeng, M. K.
[3] S. Gupta, M. Ghonge, and P. Jawandhiya, “Review of Unmanned Khan, and S. K. Das, “Edge-computing-driven internet of things: A
Aircraft System (UAS),” International Journal of Advanced Research survey,” ACM Comput. Surveys, vol. 55, no. 8, pp. 174:1–174:41, 2023.
in Computer Engineering and Tech., vol. 9, pp. 1646–1658, 04 2013. [27] S. S. Shinde, A. Bozorgchenani, D. Tarchi, and Q. Ni, “On the design
[4] C. Alcock, “Beta & Blade Air Mobility demonstrate ALIA-250 eVTOL of federated learning in latency and energy constrained computation
aircraft in NY area,” [Link]/news-article/2023-02-14/beta-and- offloading operations in vehicular edge computing systems,” IEEE
blade-air-mobility-demonstrate-alia-250-evtol-aircraft-new-york, 2023. Transactions on Vehicular Tech., vol. 71, no. 2, pp. 2041–2057, 2022.
[5] H.-H. Jebamikyous and R. Kashef, “Autonomous vehicles perception [28] A. Qayyum, M. Usama, J. Qadir, and A. Al-Fuqaha, “Securing con-
(AVP) using deep learning: Modeling, assessment, and challenges,” nected autonomous vehicles: Challenges posed by adversarial machine
IEEE Access, vol. 10, pp. 10523–10535, 2022. learning and the way forward,” IEEE Communications Surveys Tutori-
[6] M. Gluhaković, M. Herceg, M. Popovic, and J. Kovačević, “Vehicle als, vol. 22, no. 2, pp. 998–1026, 2020.
detection in the autonomous vehicle environment for potential collision [29] Z. Hu, J. Niu, T. Ren, B. Dai, Q. Li, M. Xu, and S. K. Das, “An
warning,” in Zooming Innovation in Consumer Technologies Conference efficient online computation offloading approach for large-scale mobile
(ZINC), pp. 178–183, 2020. edge computing via deep reinforcement learning,” IEEE Transactions
[7] D. Roy, Y. Li, T. Jian, P. Tian, K. Roy Chowdhury, and S. Ioannidis, on Services Computing, vol. 15, no. 2, pp. 669–683, 2022.
“Multi-modality sensing and data fusion for multi-vehicle detection,” [30] N. Sharma, A. Ghosh, R. Misra, , and S. K. Das, “Deep meta q-learning
IEEE Transactions on Multimedia, pp. 1–1, 2022. based multi-task offloading in edge-cloud systems,” IEEE Transactions
[8] A. Nooralahiyan, H. Kirby, and D. McKeown, “Vehicle classification on Mobile Computing, to appear, 2023.
by acoustic signature,” Mathematical and Computer Modelling, vol. 27, [31] F. B. Sorbelli, S. K. Das, C. M. Pinotti, and G. Rigoni, “A comprehensive
no. 9, pp. 205–214, 1998. analysis of range-free localization algorithms with mobile anchors at
[9] J. H. Stephenson, “Acoustic signature measurements and modelling of different altitudes,” Pervasive and Mobile Computing, vol. 73, 2021.
sUAS vehicles,” U.S. Army Combat Capabilities Development Com- [32] F. B. Sorbelli, C. M. Pinotti, S. Silvestri, and S. K. Das, “Measurement
mand, Aviation and Missile Center, MS 461, Hampton, VA, USA 23681. errors in range-based localization algorithms for UAVs: Analysis and
[10] F. B. Sorbelli, F. Coro, S. K. Das, and C. M. Pinotti, “Energy-efficient experimentation,” IEEE Transactions on Mobile Computing, vol. 21,
delivery of goods with drones in varying wind conditions,” IEEE Trans. no. 4, pp. 1291–1304, 2022.
Intelligent Transportation Systems, vol. 22, no. 9, pp. 6048–6060, 2021. [33] X. Kong, H. Gao, G. Shen, G. Duan, and S. K. Das, “FedVCP: A feder-
[11] FAA, “Concepts of operations, foundational principles, roles and respon- ated learning based cooperative positioning scheme for social internet of
sibilities, scenarios and operational threads, Unmanned Aircraft Systems vehicles,” IEEE Transactions on Computational Social Systems, vol. 9,
(UAS) Traffic Management (UTM) concept of operations, v2. 0,” 2020. no. 1, pp. 197–206, 2022.
[34] S. Kumar and S. K. Das, “Target detection and localization methods
[12] C. Barrado, M. Boyero, L. Brucculeri, G. Ferrara, A. Hately, P. Hullah,
using compartmental model for internet of things,” IEEE Transactions
D. Martin-Marrero, E. Pastor, A. P. Rushton, and A. Volkert, “U-space
on Mobile Computing, vol. 19, no. 9, pp. 2234–2249, 2020.
concept of operations: A key enabler for opening airspace to emerging
[35] “The JAC Hua Xianzi is the world’s first sodium-ion battery ve-
low-altitude operations,” Aerospace, vol. 7, no. 3, p. 24, 2020.
hicle,” [Link]/motoring/news/2023-03-07-the-jac-hua-xianzi-is-
[13] Single European Sky ATM Research 3 Joint Undertaking, U-space : the-worlds-first-sodium-ion-battery-vehicle/, 2023.
blueprint. EU Publications Office, 2017. [36] P. Chatterjee, P. Majumder, A. Debnath, and S. K. Das, “Distributed
[14] S. Vaidhun, Z. Guo, J. Bian, H. Xiong, and S. K. Das, “Dynamic decision making for V2V charge sharing in intelligent transportation
path planning for unmanned aerial vehicles under deadline and sector systems,” in IEEE SECON, pp. 506–514, 2022.
capacity constraints,” IEEE Transactions on Emerging Topics in Com- [37] M. Tiloca, D. De Guglielmo, G. Dini, G. Anastasi, and S. K. Das,
putational Intelligence, vol. 6, no. 4, pp. 839–851, 2022. “Jammy: A distributed and self-adaptive solution against selective jam-
[15] L. Ma and C. Zhang, “5G waveforms design for aeronautical communi- ming attack in tdma wsns,” IEEE Transactions on Dependable and
cations,” in IEEE/AIAA Digital Avionics Systems Conf., pp. 1–7, 2016. Secure Computing, vol. 14, no. 4, pp. 392–405, 2017.
[16] B. Li, Z. Fei, Y. Zhang, and M. Guizani, “Secure uav communication [38] J.-W. Ho, M. Wright, and S. K. Das, “Zonetrust: Fast zone-based node
networks over 5G,” IEEE Wireless Communications, vol. 26, no. 5, compromise detection and revocation in wireless sensor networks using
pp. 114–120, 2019. sequential hypothesis testing,” IEEE Transactions on Dependable and
[17] N. Hosseini, H. Jamal, J. Haque, T. Magesacher, and D. W. Matolak, Secure Computing, vol. 9, no. 4, pp. 494–511, 2012.
“UAV command and control, navigation and surveillance: A review [39] H. Vasudev, V. Deshpande, D. Das, and S. K. Das, “A lightweight mutual
of potential 5G and satellite systems,” in IEEE Aerospace Conference, authentication protocol for V2V communication in Internet of vehicles,”
pp. 1–10, 2019. IEEE Trans. Vehicular Technology, vol. 69, no. 6, pp. 6709–6717, 2020.
[18] R. Gopal and N. BenAmmar, “Framework for unifying 5g and next [40] T. Limbasiya, D. Das, and S. K. Das, “MComIoV: Secure and energy-
generation satellite communications,” IEEE Network, vol. 32, no. 5, efficient message communication protocols for Internet of vehicles,”
pp. 16–24, 2018. IEEE/ACM Trans. on Networking, vol. 29, no. 3, pp. 1349–1361, 2021.
[19] E. W. Frew and T. X. Brown, “Airborne communication networks for [41] G. Bendiab, A. Hameurlaine, G. Germanos, N. Kolokotronis, and S. Shi-
small unmanned aircraft systems,” Proceedings of the IEEE, vol. 96, aeles, “Autonomous vehicles security: Challenges and solutions using
no. 12, pp. 2008–2027, 2008. blockchain and artificial intelligence,” IEEE Transactions on Intelligent
[20] W. Xu, T. Xiao, J. Zhang, W. Liang, Z. Xu, X. Liu, X. Jia, and S. K. Transportation Systems, vol. 24, no. 4, pp. 3614–3637, 2023.
Das, “Minimizing the deployment cost of UAVs for delay-sensitive data [42] H. Eißfeldt, V. Vogelpohl, M. Stolz, A. Papenfuß, M. Biella, J. Belz,
collection in iot networks,” IEEE/ACM Transactions on Networking, and D. Kügler, “The acceptance of civil drones in germany,” CEAS
vol. 30, no. 2, pp. 812–825, 2022. Aeronautical journal, vol. 11, no. 3, pp. 665–676, 2020.
[21] S. K. Das, “Mobility management – a personal perspective,” Computer [43] P. Yedavalli and J. Mooberry, “An assessment of public perception of
Communications, vol. 131, pp. 26–31, 2018. urban air mobility (uam),” Tech Report, Airbus, pp. 1–28, 2019.
[22] A. Misra, A. Roy, and S. K. Das, “Information-theory based optimal [44] P. Planing, Y. Pinar, et al., “Acceptance of air taxis-a field study during
location management schemes for integrated multi-system wireless the first flight of an air taxi in a european city,” tech. rep., Center for
networks,” IEEE Trans. Networking, vol. 16, no. 3, pp. 525–538, 2008. Open Science, available online at [Link]/rqgpc, 2019.

Authorized licensed use limited to: Eskisehir Teknik Universitesi. Downloaded on November 13,2025 at [Link] UTC from IEEE Xplore. Restrictions apply.
209