P2C2: Peer-to-Peer Car Charging

Abstract

With the rising concerns of fossil fuel depletion and the impact of Internal Combustion Engine(ICE) vehicles on our climate, the transportation industry is observing a rapid proliferation of electric vehicles (EVs). However, long-distance travel withEV is not possible yet without making multiple halt at EV charging stations. Many remote regions do not have charging stations, and even if they are present, it can take several hours to recharge the battery. Conversely, ICE vehicle fueling stations are much more prevalent, and re-fueling takes a couple of minutes. These facts have deterred many from moving to EVs. Existing solutions to these problems, such as building more charging stations, increasing battery capacity, and road-charging have not been proven efficient so far. In this paper, we propose Peer-to-Peer Car Charging (P2C2), a highly scalable novel technique for charging EVs on the go with minimal cost overhead. We allow EVs to share charge among each others based on the instructions from a cloud-based control system. The control system assigns and guides EVs for charge sharing. We also introduce mobile Charging Stations (MoCS), which are high battery capacity vehicles that are used to replenish the overall charge in the vehicle networks. We have implemented P2C2 and integrated it with the traffic simulator, SUMO. We observe promising results with up to 65% reduction in the number of EV halts with up to 24.4% reduction in required battery capacity without any extra halts.

Full text

P2C2: Peer-to-Peer Car Charging
Prabuddha Chakraborty, Robert C. Parker, Tamzidul Hoque, Jonathan Cruz, and Swarup Bhunia
Department of Electrical & Computer Engineering
University of Florida, Gainesville, FL, USA
Abstract— With the rising concerns of fossil fuel
depletion and impact of Internal Combustion Engine
(ICE) vehicles on our climate, the transportation
industry is observing a rapid proliferation of Electric
Vehicles (EVs). However, long-distance travel with
EV is not possible yet without making multiple halts
at EV charging stations. Many remote regions do
not have charging stations, and even if they are
present, it can take several hours to recharge the
battery. Conversely, ICE vehicle fueling stations are
much more prevalent, and re-fueling takes a couple of
minutes. These facts have deterred many from moving
to EVs. Existing solutions to these problems, such as
building more charging stations, increasing battery
capacity, and road-charging have not been proven
efficient so far. In this paper, we propose Peer-to-
Peer Car Charging (P2C2), a highly scalable novel
technique for charging EVs on the go with minimal
cost overhead. We allow EVs to share charge among
each other based on the instructions from a cloud-
based control system. The control system assigns and
guides EVs for charge sharing. We also introduce
Mobile Charging Stations (MoCS), which are high
battery capacity vehicles that are used to replenish
the overall charge in the vehicle networks.We have
implemented P2C2 and integrated it with the traffic
simulator, SUMO. We observe promising results with
up to 65% reduction in the number of EV halts with
up to 24.4% reduction in required battery capacity
without any extra halts.
I. Introduction
Electric vehicles have existed for a while, but have
never enjoyed mainstream adoption. Now, with the need
to reduce our carbon footprint and companies like Tesla,
Nissan, and Chevrolet in the picture, the electric vehicle
has become more appealing and affordable. Nevertheless,
the adoption of EVs remains slow, mainly due to con-
sumer concerns regarding battery life, battery range, and
limited access to charging stations [1]. Inefficient charg-
ing cycles or complete discharge of a battery reduces its
life, making it imprudent to travel the full range provided
by the battery without any recharging in the middle [2].
Even though major cities in developed countries have
charging stations, the amount is still unable to support
a large EV population. Charging stations in remote
regions are few and far between. Most of the existing
charging stations are Level-2 (220V) which require long
waiting periods to charge a vehicle. Level-3 charging
stations or DC fast chargers (DCFC) (440V) are a faster
alternative; however, they are limited and very expensive
to build[3]. With these concerns in mind, researchers have
been looking into several potential solutions. Andwari
et al. surveyed innovations in EV battery technologies
[1], but concluded that the battery range and charging
time remains the most critical barrier. Novel solutions
Fig. 1: P2C2 enabled charge sharing among EVs and
MoCS-based charge distribution for charging on the go.
like charging via solar-powered roads are not applicable
across the geography [4].
In this paper, we propose a scalable peer-to-peer vehi-
cle charging solution that is both low cost and easily to
implement with minimal changes to the EVs. As shown
in Fig. 1, vehicles will share charge and sustain each other
to reach their respective destinations. A set of cloud-
based schedulers decides charge providers and receivers.
Based on the charge transaction and subsequent reroute
decisions, the cloud-based control system instructs the
EVs to carry out the charge transfer operations. With
this scheme in place, the total charge in the EV network
will eventually spread out across all the EVs. However,
even in a dynamic network with EVs entering and leav-
ing, we observe through simulation, the total charge of
the network will slowly deplete. To keep the EVs in a
state of perpetual motion, we introduce Mobile Charging
Stations (MoCS), to bring in a considerable amount
of outside charge into the EV network. The EVs are
then responsible for the fine-grained distribution of the
outside charge deposited by the MoCS. We have devel-
oped a scheduling algorithm that controls the charge
transactions and decides when and where to insert a new
MoCS. We quantitatively analyze its effectiveness using
SUMO [5] as our traffic simulator. We demonstrate that
our algorithm is fast, scalable, and efficient in dealing
with battery-related problems present in modern EVs. In
particular, we make the following major contributions:
1) We introduce a novel solution to address the electric
vehicle charging issue by proposing an on-the-go
peer-to-peer charge sharing scheme.
2) We formalize a complete framework to enable elec-
tric vehicles for sharing charges guided by a cloud-
based control system.
3) We introduce the concept of mobile charging sta-
tions, that seamlessly fit into our framework.
4) We propose an algorithm for charge transaction
scheduling and MoCS insertion that also controls the
EVs for optimal rerouting and charge sharing.
5) We quantitatively analyze the effectiveness of our

solution with an extensive simulation in SUMO [5].
II. Motivation and Background
A. Impact of charging issues in EV adoption
1) Limited range: The limited battery capacity re-
stricts long-distance driving in EVs. Even with enough
charging stations, the travel time is impacted due to
frequent, long halts for charging. The price per kilowatt-
hour of lithium-ion battery is reducing at a meager rate,
making it difficult to increase the battery capacity of EVs
without a drastic price increase [1]. High-end EVs such
as Tesla Model S and Model X with maximum range of
300 to 370 miles, suffer from high charging times. Even
with a 220V charging station, it takes about 10 hours for
a full charge [6]. Although 440V stations can reduce the
charging time, the amount of charging stations required
to support a large EV fleet will be enormous.
2) Limited stationary charging stations: The overall
number of stationary charging stations are few compared
to ICE refueling stations and mostly limited to urban
areas. High-end EVs will suffer long charging time from
level-1 or level-2 stations. DCFC (Level-3) stations are
very few, making it infeasible to sustain a big EV Fleet.
Creating large number of DCFC station is financially
infeasible as each charging unit costs $10,000-$40,000 [3].
3) Battery life: Most of the modern high-end EVs
are using Lithium-ion batteries [7]. Complete discharging
and charging, or inefficient charging cycles cause the
Lithium-ion batteries to age at an accelerated rate [8].
Hence, a long-distance drive without recharging the EV
is undesirable for the battery.
B. Existing solutions to address charging Issues
1) Better access to fast charging stations: A brute
force solution to the battery range and charging problem
is to build a high concentration of very high speed (Level-
3) charging stations to allow fast charging anywhere in
the world. However, dense and uniformly placed Level-3
stations costing $100,000 each is not feasible. Further-
more, the local power grids must be able to handle the
large amount of power that must be transferred in a short
amount of time for these stations [9].
2) Improving battery capacity: While improving the
battery capacity is undoubtedly helpful, it could sig-
nificantly increase the price of the EV [1]. Besides, it
does not solve the core problem of having to stop at a
designated station to recharge.
3) Charging from road: Charging from the road is
an exciting solution to the core problem. However, the
solar panels fitted road in Normandy, France produced
only 80,000kWh in 2018 and around 40,000kWh by the
end of July 2019 due to its inherent dependency over
the weather [4]. Converting every road in the world
into electric/solar road is a big financial undertaking,
rendering the solution infeasible.
III. Peer-to-Peer Charging Methodology
A. System overview
To allow efficient charge sharing, we design a cloud-
based control system containing a charge transaction
scheduling unit, a rerouting unit, and a database for
storing the information from EVs. The EVs will interact
with each other and the control system, as shown in
Fig.2(a). The control system (1) instructs some EVs to
share charge with some other EVs, (2) reroutes specific
EVs to bring charge providers and receivers together,
(3) speed lock EVs to allow seamless charge sharing, (4)
detaches a charge provider/receiver for overall network
charge optimization. To allow the charge scheduler to
operate, the EVs send information to the control system
periodically. An example EV-to-EV synchronization for
charge sharing is shown in Fig. 2(b) and Fig. 2(c). A big
road system can be divided into sections having separate
control systems doing the micromanagement. The overall
framework will have a global control system that will help
in seamless transfer of EVs from one region to another.
Sharing charge between EVs can distribute the total
charge in the network among all the entities. But we
observe through simulation in Fig. 6(c) that without
an outside-the-network charge source, the network will
experience a slow overall charge decay increasing the per-
centage of EV halts. To avoid this problem, we introduce
Mobile Charging Stations (MoCS) which introduces a
high volume of charge into the network. Fig.1, shows
an MoCS charging a set of EVs in a lane. To identify
charge deprived regions, the control unit maintains a
charge distribution map that is updated at a regular
interval. MoCS are inserted in charge deprived regions
if the constraints permit.
B. Scheduler optimization goals
While computing the reroute, the charge transaction,
and the MoCS insertion schedules, the scheduler takes
into account certain factors. In our embodiment of the
scheduler, we consider the following optimization goals:
1) Maximize effective charge usage by analyzing
the charge distribution map.
2) Minimize charging station halts by sustaining
low battery vehicles.
3) Minimize travel time of all EVs by limiting the
number of rerouting.
4) Maximize battery life by taking into account the
depth of discharge of each EV.
5) Prioritize MoCS as provider over normal EVs.
The final decision of the scheduler is a function of all
the optimization parameters, where each parameter can
be weighted differently depending on which goals the user
wishes to prioritize. Note that the optimization goals we
propose are by no means exhaustive and more goals can
be added to the scheduler depending on the scenario.
C. Scheduling Algorithms
The core algorithm for P2C2 scheduling is presented
in Algo. 1. The scheduler takes in the charge distribution
map (Charge Dist Map) as input and generates a list
of instructions (Instruction List) to be followed by the
EVs, MoCS, and MoCS depots. The scheduler acts as
an intelligent decision function. We design a method
find critical evs for identifying the EVs in the network
with critical battery capacity using the charge distribu-
tion map maintained in the cloud control system. In line
3, we use this method to generate the critical EV list
(Crit EV s List). The method find prov ev is designed
to identify the best provider EV (P rov EV ) for a given
critical EV from all nearby EVs within a user-specified
range. This method uses a greedy search algorithm based
on a linear weighted function of all the optimization

Fig. 2: (a)A system view of P2C2 showing the interaction between the control system and EVs. The control system is
located in the cloud facilitates EV-to-EV charge sharing. (b)The paired EVs are being guided by the control system
to move closer and come on the same lane. (c)The donor EV is sustaining the EV with critical battery condition.
Algorithm 1 P2C2 Context-Aware MoCS Scheduler
1: procedure generate Schedule(Charge Dist M ap)
2: Instruction List =∅�Initialized to empty set.
3: Crit EV s List =f ind critical ev s(Charg e Dist M ap)
4: i= 0
5: while i < length(C rit EV s List)do
6: P rov EV =f ind prov ev(Crit E V s List[i])
7: inst =gen charge tr an inst(P rov E V, C rit E V s List[i])
8: Instruction List.append(inst)
9: i=i+ 1
10: Charg e D R List =f ind char ge dr(C harge D ist Map)
11: i= 0
12: while i < length(C harge DR List)do
13: ins pt =f ind best mocs ins pt(C harge D R List[i])
14: mocs ins num =f ind M oCS num(Charg e DR List[i])
15: MoC S I nst =gen mocs ins inst(ins pt, mocs ins num)
16: Instruction List.append(MoC S I nst)
17: i=i+ 1
18: return Instruction List
goals mentioned earlier. In line 7, we generate the charge
transaction instruction (inst) required to facilitate the
charge transfer. The instruction (inst) is appended to the
Instruction List in line 8. The instructions are targeted
towards helping the EVs to come nearby and speed
lock. We define a method f ind charge dr which finds all
the charge deprived regions in the network using linear
search. In line 10, the method find charge dr finds the
regions in the road system with a high density of critical
EVs. We define two methods f ind best mocs ins pt and
find M oC S num to find out the best MoCS insertion
point and amount of MoCS that should be spawned
to deal with a particular charge deprived region, re-
spectively. The MoCS insertion point (ins pt) is se-
lected based on the predicted trajectory of the low
battery charge EVs such that the MoCS can easily
converge with them. The number of MoCS to be inserted
(mocs ins num) is based on the severity (number of
critical EVs) of the charge deprived region and the
MoCS quota remaining. The function gen mocs ins inst
generates the instruction (MoCS Inst) specifying the
amount of MoCS and MoCS insertion location to be sent
to the MoCS depot. The complete Instruction List is re-
turned in line 18 from the GEN E RAT E SC HE DU LE
method. The instructions generated are sent to the re-
spective MoCS depots and EVs. For the purpose of our
Fig. 3: A physical embodiment of EV-to-EV on-the-go
charging mechanism.
simulation, we modify SUMO to emulate MoCS depots
and the whole EV network.
D. Car-to-Car charging mechanism
We envision a safe, insulated, and firm telescopic arm
carrying the charging cable. After two EVs lock speed
and are in range for charge sharing, they will extend their
charging arms, as shown in Fig. 3. The arms heads will
contain the charging ports, and they will latch together
using either magnetic pads or other means. The arms
and the overall charging operation will be coordinated
by the respective arm controllers of each EV. This is just
one possible realization of the charge transfer mechanism.
The entire charging operation can be safely orchestrated
if the EVs involved follow a certain predefined proto-
col. For autonomous/semi-autonomous EVs, the pairing
mechanism can be further streamlined. Wireless charging
is also possible in the future.
E. Battery chemistry
Battery to battery charging required for charge sharing
is feasible and being actively explored. Products like
[10] allow EVs to share charge. Efforts are also being
made towards faster-charging batteries. In particular,
lithium plating-free charging allows quick recharge at
all temperatures without sacrificing the durability of
battery cells [11]. Emerging battery technologies such as
the aluminum dual-ion battery have been shown to have
impressive charging rates and high energy density [12].
IV. Simulation Results Peer-to-Peer Charging
A. Simulation Setup & Fundamental Observations
To analyze the effectiveness of our cloud control system
and the scheduling algorithm, we use an open-source
traffic simulator, SUMO (Simulation of Urban Mobil-
ity) [5] and integrate the P2C2 scheduler with it. We
made modifications to SUMO to support peer-to-peer car
charging and MoCS. The P2C2 scheduler communicates
with SUMO periodically to gather traffic information and

send instructions. We use a 240 km highway to test our
method. We run each simulation instance for 5 hours in
real-time. We ensure that each EV travels at least 50
km. Each EV weighs 2109 kg with a battery capacity
of 75 kWh. Unless otherwise mentioned, the EVs and
MoCS enter the simulation with full charge. The weight
of each MoCS is 11793 kgs which is the gross vehicle
weight rating for a class 6 truck [13]. Each MoCS carries
850 kWh charge and are battery powered themselves. We
observe the effect of other parameters such as (1) MoCS-
to-EV charge transfer rate, (2) amount of MoCS in the
network, and (3) battery capacity reduction of the EVs
in later sections.
We test most of our observations on three different
traffic scenarios. The internal parameters defining each
of these scenarios are as follows:
1) Light Traffic: Initially 500 EVs are inserted with a
new EV entering the simulation every 4 seconds. A
total of 5000 EVs will be inserted over 5 hours.
2) Medium Traffic: Initial traffic of 1000 EVs with a
new EV entering the simulation every 3 seconds. A
total of 7000 EVs will be inserted over 5 hours.
3) High Traffic: Initially 2000 EVs are inserted with
a new EV entering the simulation every 2 seconds.
A total of 11000 EVs will be inserted over 5 hours.
We use a charging rate of 1kW/min for simulation
based on a realistic EV-to-EV charging estimate pro-
vided in [10]. We consider an EV to be halted when its
charge reaches zero. All charge transfer is carried out
with 95% efficiency (i.e., 5% loss during transfer).
Fig. 4 illustrates the overall charge distribution in the
highway. Each point on the plot indicates the average
charge of vehicles in the region. In the charge distribution
map shown, we can observe a potential charge deprived
region. A few MoCS will probably be inserted in the
region depending on the scheduler’s decision.
In Fig. 5, we can see the battery charge trend for
6 sampled EVs (red) on the left and 2 sampled MoCS
(blue) on the right from the network. The EVs generally
experience an initial drop in the battery charge before
they are assigned another EV as a provider. After that
point, most of the EVs maintain a particular battery level
and continue to move perpetually. The purpose of MoCS
is to deposit a huge amount of charge in the network
quickly; hence, they constantly lose charge as can be seen
from the blue plots.
B. Effect of MoCS Charge Transfer Rate on Halt
We observe the effect of different MoCS-to-EV charge
transfer rate on the percentage of EV halts. 1x charge
Fig. 4: The charge distribution map maintained by the
cloud application at a particular time instance.
rate is 1kWh per minute based on [10]. Note that we only
change the charge transfer rate between an MoCS and an
EV. The EV-to-EV charge transfer rate remains 1kWh
per minute throughout the experiment. In Fig. 6(a), we
observe that the percentage of halts for all the three traf-
fic scenarios decreases as we increase the MoCS charge
transfer rate. If fast charge transfer batteries can be used
in the EVs/MoCS, then the effectiveness of P2C2 will
be increased. P2C2 charging scheme appears to be more
effective in denser traffic scenarios. As can be seen in
Fig. 6(a), the percentage of halts for high traffic is least.
With more EVs in the network, less amount of rerouting
is needed, and an EV with a critical battery state can be
quickly assigned to a provider EV which is close by.
C. Effect of Number of MoCS on Percentage of Halt
To observe the effect of the number of MoCS in the
network on the percentage of EV halts, we set the MoCS-
to-EV charge transfer rate to 2x (2kWh per minute),
and EV-to-EV charging rate to 1x (1kWh per minute)
and vary the limit on the percentage of MoCS in the
network. The percentage of MoCS refers to the maximum
allowable MoCS for every 100 EVs in the network. In
Fig. 6(c), we observe that as we increase the limit of
the percentage of MoCS, the percentage of EV halts
decreases. So a higher quantity of charge influx also helps
in halt reduction.
D. Charging Time Reduction Analysis
Based on the battery capacity of the cars used in the
simulation, it should take approximately 10 hours to fully
charge on the NEMA 14-50 plugs through a 240v outlet
[6]. By multiplying the average time charging for each
halt with the total number of halts from Table I, we
obtain the total charge time for all traffic scenarios. As
shown in Table I, the total time spent for stationary
charging reduces significantly due to the charge sharing
scheme proposed. The % of reduction for P2C2 is calcu-
lated compared to the required charge time results for
no P2C2 (without). We use MoCS-to-EV charging rate
of 2x and 5% MoCS amount limit for obtaining the P2C2
results in Table I.
E. Battery Capacity Reduction and MoCS Tradeoff
High capacity batteries in EVs lead to an increase in
EV weight and cost. In Fig. 6(b), we observe the effect
of reducing the battery capacity of the EVs on the per-
centage of halts for the medium-traffic scenario. We see
the percentage of faults increase as the battery capacity
is reduced. There is a trade-offpossibility between the
amount of MoCS and the battery capacity of EVs. If the
amount of MoCS inserted is within 15% of the total EVs
in the network, we can reduce the battery capacity of all
EVs by 24.4% and still achieve the same amount of halts
compared to not using the P2C2 scheme. Therefore, we
can reduce the battery capacities of all EVs and have
more MoCS running in the system. In future work, we
will look into incorporating cost in this trade-off.
V. Conclusion
We have presented a novel framework, called P2C2,
for charging EVs on the go to address the issue of
EV charging infrastructure. P2C2 relies on EV-to-EV
coordination as well as a cloud-based guidance system for

Fig. 5: Change of battery charge level over time for sampled EVs(red) and MoCS(blue) in the network.
Fig. 6: (a) The percentage of EV halts reduces as the MoCS-to-EV charge transfer rate increases. (b) The percentage
of halt increases as we decrease the battery capacity. The halt percentage is less in presence of more percentage of
MoCS. (c) The percentage of EV halts reduces as the limit on the percentage of MoCS in the network is increased.
TABLE I: The percentage of halt induced charging time
reduction in different traffic scenarios.
Light Traffic Medium Traffic High Traffic
Baseline P2C2 Baseline P2C2 Baseline P2C2
% of Halts 19.68 12.62 19.02 10.18 18.25 9.16
Num of EVs 5000 5000 7000 7000 11000 11000
Halt Time (hrs.) 9840 6310 13314 7126 20075 10076
Halt Time Cut % - 35.87 - 46.48 - 49.81
creating a real-time charge distribution map of the net-
work of EVs and making informed decisions about charge
transactions. We incorporate the concept of MoCS - a
mobile charging vehicle with a large battery - that can
be dispatched to recharge a network of EVs. We have
developed a system, algorithm, and method to enable
EV-to-EV charging. Using a popular traffic simulator,
SUMO, we have realized the P2C2 framework with real-
istic charging parameters and observed a reduction in
the number of halts, and reduction in battery capac-
ity requirements of EVs (thus, leading to reduced cost
and weight). Future work will involve augmenting our
solution for a heterogeneous network of battery operated
entities, like drones and utility robots.
References
[1] A. M. Andwari, A. Pesiridis, S. Rajoo, R. Martinez-Botas,
and V. Esfahanian, “A review of Battery Electric Vehicle
technology and readiness levels,” Renewable and Sustainable
Energy Rev., vol. 78, pp. 414–430, 2017.
[2] F. Hoffart, “Proper care extends li-ion battery life,” Power
Electronics, p. 25, 2008.
[3] “Costs Associated With Non-Residential Electric Vehicle
Supply Equipment,” 2015, U.S. Department of Energy. [On-
line]. Available: https://afdc.energy.gov/files/u/publication/
evse cost report 2015.pdf
[4] “France’s Solar Road Dream May be Over After
Test Fails,” 2019, Global Construction Review. [Online].
Available: http://www.globalconstructionreview.com/news/
frances-solar- road-dream- may-be-over-after- test-fa/
[5] D. Krajzewicz, J. Erdmann, M. Behrisch, and L. Bieker,
“Recent development and applications of SUMO-Simulation
of Urban MObility,” Int. Jour. On Advances in Syst. and
Measurements, vol. 5, no. 3&4, 2012.
[6] “Mobile Connector Owners Manual,” 2012, Tesla.
[Online]. Available: https://www.tesla.com/sites/default/
files/downloads/gen 2 mobile connector owners manual 32
amp en US.pdf
[7] L. Lu, X. Han, J. Li, J. Hua, and M. Ouyang, “A review on
the key issues for lithium-ion battery management in electric
vehicles,” Jour. of Power Sources, vol. 226, pp. 272–288, 2013.
[8] B. Xu, A. Oudalov, A. Ulbig, G. Andersson, and D. S.
Kirschen, “Modeling of lithium-ion battery degradation for
cell life assessment,” IEEE Tran. on Smart Grid, vol. 9, 2016.
[9] M. Etezadi-Amoli, K. Choma, and J. Stefani, “Rapid-charge
electric-vehicle stations,” IEEE Tran. on Power Delivery,
vol. 25, no. 3, pp. 1883–1887, 2010.
[10] “ORCA Inceptive data sheet,” 2019, Andromeda Power LLC.
[Online]. Available: http://andromedapower.com/uploads/
ORCA Inceptive Datasheet.pdf
[11] X.-G. Yang, G. Zhang, S. Ge, and C.-Y. Wang, “Fast charging
of lithium-ion batteries at all temperatures,” Proceedings of
the National Academy of Sciences, vol. 115, no. 28, 2018.
[12] E. Zhang, W. Cao, B. Wang, X. Yu, L. Wang, Z. Xu, and
B. Lu, “A novel aluminum dual-ion battery,” Energy Storage
Materials, vol. 11, pp. 91–99, 2018.
[13] R. A. Rahman, M. N. Tamin, and O. Kurdi, “Stress analysis of
heavy duty truck chassis as a preliminary data for its fatigue
life prediction using fem,” Jurnal Mekanikal, vol. 26, 2008.