See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/261315863

Smart electric vehicle charging: Security analysis

Conference Paper · February 2013

DOI: 10.1109/ISGT.2013.6497830

CITATIONS READS
28 427

4 authors:

Mustafa A. Mustafa Ning Zhang

The University of Manchester The University of Manchester
26 PUBLICATIONS   159 CITATIONS    115 PUBLICATIONS   805 CITATIONS   

SEE PROFILE SEE PROFILE

Georgios Kalogridis Zhong Fan

Toshiba Research Europe Limited Toshiba Research Europe Limited
36 PUBLICATIONS   1,457 CITATIONS    108 PUBLICATIONS   2,370 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Privacy-friendly peer-to-peer energy trading View project

DiskMan - Dynamic Risk-based Access Management (Frictionless Authentication) View project

All content following this page was uploaded by Mustafa A. Mustafa on 27 May 2017.

The user has requested enhancement of the downloaded file.

 Smart Electric Vehicle Charging: Security Analysis
Mustafa A. Mustafa and Ning Zhang Georgios Kalogridis and Zhong Fan
School of Computer Science Telecommunications Research Laboratory
The University of Manchester Toshiba Research Europe Limited
Manchester, UK Bristol, UK
{mustafm, nzhang}@cs.man.ac.uk {george, zhong.fan}@toshiba-trel.com

Abstract—This paper provides a comprehensive security EVs do not just consume electricity; they can also be used
analysis of the Electric Vehicle (EV) charging service in Smart as Distributed Energy Resources (DERs). Using a Vehicle-to-
Grid (SG) environment (i.e. the “smart” EV charging Grid (V2G) technology [8], EVs can feed electricity back to
application). It first describes three EV charging scenarios, at the grid, thus making them an attractive facility for providing
home, at work and at public places. Based on these use-case ancillary services (e.g. frequency regulation, spinning reserve,
scenarios, the paper presents a model for smart EV charging, etc. [9]). Although providing these services would degrade an
consisted of application entities and interactions among them. It EV’s battery lifetime due to increased number of charging
then illustrates potential message types communicated among cycles, the potential revenue generated from these services
these entities. Based on this model and the exchanged messages,
would compensate the battery degradation cost [8]. However,
the paper analyses security problems and potential security
to be eligible, an ancillary service provider needs to offer at
threats imposed on the entities, which leads to the specification
of a set of security and privacy requirements. These
least a certain amount of flexible demand (e.g. 3MW in the
requirements could be used to guide the future design of UK [10]). As an EV typically has a limited battery capacity of
solutions for secure smart EV charging systems and/or a 10-40kW [5] there is a need for a new entity, EV AGGregator
risk/impact assessment of such systems. (EVAGG) to aggregate the batteries of a number of EVs and
represent their users in the electricity market, i.e. EVAGG will
Index Terms—Electric Vehicle (EV), Security, Smart Charging, act as a middleman between users and grid operators.
Smart Grid (SG), Privacy
In addition to EVAGGs, there will be other stakeholders
I. INTRODUCTION too taking part in smart EV charging activities. These include
Transmission System Operator (TSO), Distribution Network
Smart Grid (SG) is envisioned to take full advantage of Operator (DNO), Data Communications Company (DCC) and
modern technologies in transforming the current electrical grid suppliers (further details in §III). Different stakeholders have
to one that functions more intelligently [1]. The aim is to make different aims and interests. For example, TSO would like to
the electricity industry operate more efficiently and to provide reduce peak loads and have more spare capacity; users would
electricity to society in a more secure, reliable and sustainable like to minimize their costs by recharging EVs at the cheapest
manner. One of the envisioned key elements of SG is the possible price and by using them to provide ancillary services;
Electric Vehicle (EV). In addition to the environmental suppliers would like to maximize their profit by selling more
benefits (i.e. reduced greenhouse gas emissions), EVs also electricity at times when the price is higher; and EVAGGs
have the potential to bring economical benefits to the society would like to maximize their profit by aggregating more EVs
(i.e. reduced operational costs and oil dependency) [2]. and by actively participating in the electricity market. To
However, several studies [3], [4] have shown that letting maximize their respective profits, some of these stakeholders
people recharge EVs in an uncontrolled manner could have may take actions that may conflict with other entities’
negative effects on the grid. The peak load can increase interests. Therefore, solutions designed to support smart EV
significantly requiring more generation capacity and upgrades charging should prevent or minimize the chances of any unfair
in transmission and distribution networks. The gap between play by any of the stakeholders and provide adequate
the base load and the peak load can be even larger resulting in protection against any threats or attacks launched by outsiders.
inefficient utilization of the available generation capacity. There are very few published papers on the smart EV
Specifically, balancing the grid, i.e. matching the supply with charging topic [11]–[14]. These papers largely focus on how
the demand, will be more difficult requiring more spinning to optimize EV charging operations to minimize users’ costs,
reserve. To minimize these negative effects, several studies to maximize users’ profits by offering ancillary services, to
[5]–[7] have suggested an approach of a “smart” EV charging, maximize the use of electricity from RES, and to minimize the
i.e. to charge EVs at times when it is most effective to users overall peak demand, etc. There has not been any prior study
and to the grid (e.g. at off-peak times, at times with surplus of on security analysis of smart EV charging services. This paper
electricity generated by Renewable Energy Sources (RESes)).

This research is supported by the Engineering and Physical Science

Research Council (EPSRC) and Toshiba Research Europe Limited under
Grant [EP/I1501541/1].
focuses on the latter problem. It first describes a generic model installed at work places and connected to the grid via the SMs
for smart EV charging drawn from the various charging use- of the corresponding companies. Initially companies having
case scenarios we have identified. It then analyses security their own fleets (e.g. rent-a-car) may invest in EVSEs. EVs
problems of the applications using this model, before can be recharged from the EVSEs and the corresponding
specifying a set of security requirements necessary to company can pay for the consumed electricity. At later stages,
safeguard the operations of smart EV charging. These companies without their own fleets may also install EVSEs.
requirements could be used to guide a future design of a
secure system to support smart EV charging and/or be used to Furthermore, a company may pass on the charging and
assess risks in such a system. In detail, the rest of the paper is electricity costs to its employees by charging those who use
organized as follows: §II gives smart EV charging use-case the recharging service. In such cases, the EVSEs should be
able to record such information as who has used how much
scenarios; §III describes the generic model drawn from these
scenarios; §IV analyses potential security threats/attacks in the electricity and when. Some companies, especially those which
model; §V specifies a set of security requirements, and, further consume large amount of electricity (e.g. manufacturers), may
buy electricity in bulk directly from generators for a cheaper
discussions are given in§VI. Finally §VII concludes the paper.
price than the price offered by suppliers. In such cases, if the
II. USE-CASE SCENARIOS company resells the electricity to its employees at a higher
The Department for Transport (DfT) in the UK has price than the price it has paid for (but for the employees, this
suggested three different locations where an EV may be price may still be cheaper than retail electricity price), it may
charged [15]: at home, at work and at public places. get the investment for the EVSEs back. This is a win-win
situation for both the company and its employees: the
A. At Home company makes profit by selling electricity and employees
Suggesting the home charging location encourages users to recharge their EVs at a cheaper price than the retail price.
recharge EVs at night. In this way, the recharging may mostly In addition, if a company has a private parking lot large
take place during off-peak times when the price of electricity enough to accommodate a large number of EVs, and if the
is cheaper. The load caused by the recharging can be spread EVs’ batteries are aggregated, then they form a single large
across the distribution network increasing the grid’s reliability. flexible load. This load can play an important balancing role in
To support recharging at home locations, Electric Vehicle the electricity market. The company can aggregate its
Supply Equipment (EVSE) will need to be installed on the employees’ EVs and offer ancillary services to the grid. In
premises of an EV user and connected to the user’s Smart return, the company may be able to offer free recharging
Meter (SM). The electricity consumed by the EV can directly services to its employees as long as they keep their EVs
be added onto the user’s household electricity bill. In addition, plugged into the designated lot. This is another win-win
EVSE could also measure the electricity consumed or fed situation for both the company and its employees: the
back by the EV. Thus, suppliers could offer electricity tariff company enters a new market (balancing market) and its
plans specially designed for EVs and bill them separately. employees may get free electricity at work.
Users may also have detailed information about their EVs’
charging times, durations, electricity consumptions and costs. C. At Public Places
Of course, there may be cases where an EV is recharged Other potential charging places for EVs are public places
on premises different from its user’s home, e.g. a user (UB) such as parking lots of supermarkets, shopping malls, rest
visits a friend (UA) and recharges his EV, EVB, at UA’s home. places between cities, etc. EVSEs installed at these places are
As EVB can consume a considerable amount of electricity, it likely to be used as a top-up option. In the future, there could
would be desirable for UB to bear the cost of the electricity be a fast growing market for these top-up electricity stations
supplied to EVB at UA’s home. There are two ways for UB to (similar to the petrol stations we currently have). Owners of
make this payment: (1) UB pays UA directly by cash; (2) UB these stations will be responsible for providing the electricity;
pays UA via their suppliers. The latter payment method may they will be contracted with an electricity supplier. EV users
be more convenient and socially acceptable, as there will be may go to these stations, recharge their EVs and pay for the
no direct ‘business’ dealings between the two users. If both service used. If these stations are designed for a fast recharge
users are contracted with the same supplier, then it would not (which is likely), ancillary services to the grid are unlikely to
be too difficult for the supplier to simply add the expenses for be offered by EVs or the owners of these stations.
recharging EVB to the account of UB. However, UA and UB Some residential houses/flats or commercial premises may
may use different suppliers, SA and SB, respectively. In such a not have dedicated off-road parking places. In such cases,
case (EV roaming), communication and interactions between EVSEs may have to be installed on streets to offer charging
the two suppliers are necessary. facilities to people who live or work on these premises. As
B. At Work these EVSEs are in public places, they may be accessible to
any EV and they should be able to serve many users.
Parking facilities at employees’ work places may also be Currently, there are some EVSEs installed at public places in
used for charging EVs. Usually EVs are parked for 8 hours a the UK [16], but their use is free as they are part of the
day at these facilities during week days. During these hours government trials. However, in the future, the use of these
EVs can be recharged if they need to top up their batteries. EVSEs may be controlled. Whoever is in charge of such
They can also be discharged here offering ancillary services to EVSEs will need to contract with an electricity supplier,
the grid. To support these services, EVSEs will need to be purchase electricity and sell it to individual EV users.
 Figure 1. A generic model of a smart EV charging.

III. A GENERIC MODEL: ENTITIES AND INTERACTIONS

Based on the above use-case scenarios, we have devised a
generic model for smart EV charging. This model combines
the architecture proposals made by the previous studies [17]–
[19], and takes into account the consultation results of the
UK’s smart metering implementation programme [20]. Figure 2. Interactions in a smart EV charging on private premises.

A. Entities B. Interactions among Entities and Message Types

As shown in Figure 1, the model has the following entities: Potential message types and interactions among the
 EV: a vehicle that is powered by electricity [21]. entities in a smart EV charging activity on private premises
are summarized in Figure 2. The interactions may include:
 EVSE: equipment that connects an EV to the grid [22]
1) Connection: A user plugges the EV into the EVSE.
 User: a person who owns an EV. The user demands, 2) Negotiation: The user negotiates with the EVAGG.
consumes and pays for the electricity used by the EV.
a) User’s preferences: A user sends a charging request
 SM: an advanced meter that measures electricity (REQEV) to EVAGG. REQEV may include information such as
consumption and can perform real-time two-way targeted State-of-Charge (SOC) at time x (e.g. fully recharged
communications with other entities in the SG [20]. battery by 7am). The preferences may include information
such as recharge with electricity from RES, recharge at off-
 Networking facility: it is consisted of Home Area
Networks (HANs) which are private networks located peak time, offer ancillary services, etc. The user’s identity will
on users’ premises, Building Area Networks (BANs) also be included in REQEV, so the EVAGG could incentivize
[23] which aggregate data from several HANs, the correct user for the ancillary services provided by the EV.
Neighborhood Area Networks (NANs) which cover b) Current status: EVAGG obtains information such as
larger areas (streets) and collect data from several the grid’s status from DNOs, the electricity tariff plan at the
BANs, and, finally, Wide Area Networks (WANs) charging location from a supplier, when and how much
[24] which cover even larger areas (towns), collect electricity is expected from RESes, etc.
data from several NANs and forward the data to DCC. c) Optimization: EVAGG analyses the data, works out
 DCC: an organization regulated by the UK an optimal charging schedule for the EV (in accordance with
government. It will be responsible for storing and the user’s preferences), and sends the schedule to the user.
managing data collected from SMs, and transferring d) Confirmation: The user may accept the schedule, ask
the data to other authorized entities [20]. for rescheduling or reject it and start an uncontrolled recharge.
 TSO: it balances the grid using ancillary services. It 3) Control: EVAGG manages the EV’s charging process
also owns and manages the transmission network [25]. in accordance with the agreed schedule by sending control
signals (CSEV) to the EVSE. It also constantly bids at the
 DNOs: they own and manage distribution networks. electricity market to offer ancillary services. If its bid is
 EVAGG: it aggregates EVs, optimizes their charging accepted, it will receive a command to change its overall
processes, and provides ancillary services to TSO. In demand. To comply with the command, the EVAGG may
return, it offers financial incentives to EV users. change the EV’s charging shedule. In spite of these changes,
 Generator (GEN): it generates electricity [25]. the targeted SOC should be reached by the end of the session.
4) Payment: There may be three possible payment flows
 Supplier: it buys electricity from the electricity market related to the user.
and sells it to individual users [25].
  The user pays the supplier for the electricity other parties. For example, battery manufacturers may be
consumed by the EV. interested in knowing how frequently EV batteries are
recharged. With access to such information, the battery
 The user receives payments from the supplier for the manufacturers may be able to identify the most used EVs and
electricity discharged from the EV and fed to the grid. market their products to the users of these EVs.
 The user receives payments from the EVAGG for any D. Denial-of-Service
ancillary services offered by the EV to the grid.
One of the objectives in controlling an EV charging
IV. THREAT ANALYSIS process is to reduce the chances of putting the grid in an
While smart EV charging has a potential to bring financial imbalance state. Ensuring reliable and on-time delivery of
benefits to various entities, it may also offer opportunities for communication messages and reliable operations of EVAGG
some entities to take advantage of the others in order to is essential to achieve this objective. DoS attacks on channels
maximize their benefits. This section analyses how various and services may delay message deliveries and make services
entities in the model may cheat or commit security breaches. inaccessible to legitimate users. One example of such attacks
is tampering with pricing signals sent to users (i.e. modifying
A. Impersonation the price to a very low level). This attack could result in a
Impersonation is the theft of another entity’s identity. In large number of REQEV signals being sent to the EVAGG
the smart EV charging application some entities may try to almost at the same time. A potential consequence of these
launch such attacks. For example, a dishonest user may try to DoS attacks is an unstable grid or even blackouts.
impersonate another user in an attempt to recharge an EV for E. Privacy Breaches
free (by plugging the EV into a socket connected to the SM of
the victim) or to receive any incentives for ancillary services Security problems may also be caused by authorized
offered by the victim’s EV (by intercepting and manipulating insiders. Privacy breaches are an example of such problems.
REQEV sent by the victim); a dishonest supplier may try to Legitimate entities may use the opportunities to build users’
impersonate an EVAGG in an attempt to increase its sales at profiles for purposes that are not directly relevant to smart EV
peak times or an external attacker may try to impersonate the charging. For example, an EVAGG may gather information
EVAGG in an attempt to take control of the charging process about EVs and their charging locations. Using this information
(by forging legitimate CSEV). The aims of these attacks could the EVAGG may build a profile of a user. EVAGG may even
be to seek financial gains, to cause disruptions in the charging sell this information to other interested parties.
processes or to just cause nuisance. F. Disputes
B. Tampering with Communication Messages As the EV charging process has financial implications to
Fraud or harm may also be committed by tampering with all the entities involved, there are incentives for any of the
the messages sent among the participating entities (e.g. entities to fraudulently try to obtain some financial gains or to
REQEV, CSEV, etc.). Tampering may be done by using one of reduce their costs. For example, EVAGG may try to control an
the following ways: modifying, delaying or replaying a EV without having its user’s permission. Similarly, if a user
legitimate message, or inserting illegitimate messages into the sends a REQEV “recharge now” instead of “recharge when
underlying network. Tampering with legitimate messages can the price is below x” by mistake, the user may later deny
cause damages to the grid’s stability and/or inconveniencies to having sent the REQEV and blame the EVAGG for the high
users. For example, if an adversary modifies CSEV content recharging costs. Such dishonest and/or accidental actions
from “recharge” into “discharge”, the user will find his EV may lead to disputes between different entities. Having to deal
with a fully discharged battery. This may force the user to with such disputes may put off users from participating in
recharge the EV at a peak time. In such case, the user will not smart EV charging programs. Therefore, for the success of
only suffer from inconvenience, but also some financial these programs necessary technical measures should be in
losses. A similar situation may arise if a CSEV, assigned to a place to ensure that any potential disputes between the entities
specific EV, is modified to control another EV. Delaying can be resolved promptly and fairly.
important CSEV messages could affect the grid’s stability and V. SECURITY REQUIREMENTS
cause brownouts. Replay attacks may also cause damages. For
example, if an adversary captures a REQEV “start recharge” Based on the above threat analysis, this section specifies a
message sent at an off-peak time, but replays it at a peak time, set of security requirements for a smart EV charging service.
the user will have to pay a higher price for the electricity. Figure 3 illustrates some proposed actions or security services
Injecting fake messages into the network could cause to counter the threats/attacks identified in the previous section.
congestion, which may delay successful deliveries of other A. Entity Authentication
legitimate messages causing service delays or disruptions. Entity authentication assures that a communicating entity
C. Eavesdropping is the one that it claims to be. To counter impersonation
REQEV and CSEV may contain sensitive information about attacks and to provide a fair billing with an EV roaming
an EV and its user (e.g. identities, account details, etc). If an support, strong entity identification and mutual authentication
adversary can read these personal details, he may be able to services should be provided. These include authentication
impersonate the user or user’s EV. In addition, REQEV and between:
CSEV may contain information that has commercial values to
  EVi and UA – ensures that only EVs authorized by UA
can connect to the grid via EVSEA.
 UB and SB – ensures that the electricity used by the
EVB is accounted to its user, UB.
 EVB and SB – ensures that UB is accounted only for
the electricity used by the EVB.
 UB and EVAGGB – ensures that any incentives for the
ancillary services offered by EVB are accounted to UB.
B. Message Authenticity
Message authenticity assures that messages (e.g. REQEV,
CSEV, payments, etc.) are exactly what have been sent by the
claimed entities. To counter tampering attacks the authenticity
of these messages should be provided. Keyed hash values,
Message Authentication Code, or digitally signed tokens
(through the use of public key cryptography) [26] are typically
used to counter these attacks and to ensure the messages’
integrity. Usually, in such a cryptographically protected token,
a nonce (a random arbitrary number used only once) or a time
stamp is enclosed to ensure that replay attacks can be detected.
C. Authorisation
Authorization is the process of granting authorized users
legitimate access to resources (systems, data, applications,
Figure 3. A smart EV charging application that supports users’
etc). To minimize the chances of successful impersonation mobility in accessing electricity and in making payments.
attacks, UA should always authorize EVs requesting to use
EVSEA for the first time. Furthermore, all metering or usage  NRR of CSEV - EVAGGB holds evidence that EVB
related data are expected to be managed by the DCC. To has received the CSEV.
prevent misuse of the data, the DCC should impose access
control of data based on the principle of least privilege, e.g. To enhance the level of trust between UB and EVAGGB,
only grant EVAGGB the access to data related to the charging all the actions taken by them should be unforgeable,
of EVB. An access control mechanism, e.g. Role Based Access undeniable and traceable. This is necessary to ensure a fair
Control (RBAC), may be exploited to serve this purpose. resolution of any dispute. A number of cryptographic building
blocks, including digital signatures, recoverable and
D. Confidentiality verifiable tokens [26] and off-line trusted third parties (TTPs)
Confidentiality is a protection of data from any may be used to provide the NR properties.
unauthorized disclosure. To counter eavesdropping attacks the F. Availability
confidentiality of REQEV, CSEV, payments, etc. should be
protected. Confidentiality of these messages can be provided Availability is the property of a system or a system
by using an encryption technique such as symmetric resource being accessible and usable upon demand by
encryption, public key encryption or a combination of both. authorized system entities. DoS attacks impose a threat to this
property. Measures must be in place to ensure that important
E. Non-repudiation service entities (e.g. DCC, EVAGGB) can resist DoS attacks.
Non-repudiation (NR) provides protection against false G. Anonymity and Non-linkability
denial of having participated in a communication/transaction.
There are two aspects of NR: NR of Origin (NRO) and NR of Anonymity is a property of not being identifiable within a
Receipt (NRR). NRO provides protection against a sender’s set. In our problem context, anonymity may be used to prevent
false denial of having sent a message. NRR provides UB’s real identity from being revealed to EVAGGB during a
protection against a receiver’s false denial of having received charging process, and/or prevent UB’s charging location from
a message. In smart EV charging scenarios the following NR being linked to UB’s real identity. Non-linkability protects
requirements should be considered: EVB’s multiple charging sessions from being linked together
and protects UB’s multiple charging sessions from being
 NRO of REQEV - EVAGGB holds evidence that UB linked to UB’s real identity. To preserve UB’s privacy, both
has sent a REQEV. anonymity and non-linkability properties should be provided.
This can be achieved by assigning a dynamic identifier to each
 NRR of REQEV - UB holds evidence that EVAGGB of the charging sessions performed by UB. Of course, a
has received the REQEV. controlled and authorized linkage of UB’s multiple charging
 NRO of CSEV - EVB holds evidence that EVAGGB sessions should be supported, as this is necessary to ensure
has sent a CSEV. accountability and traceability in the event of a dispute or
security incident. As SB has to know the real identities of UB
and EVB (for authentication and billing purposes), it may be a [1] H. Farhangi. The path of the smart grid. Power and Energy Magazine,
candidate TTP and this TTP may be the only entity which IEEE, 8(1):18–28, Jan.-Feb. 2010.
[2] Investigation into the scope for the transport sector to switch to electric
knows the connection between UB’s real and dynamic identity. vehicles and plugin hybrid vehicles. Technical report, BERR & DfT,
In more general cases, EV charging anonymity may be Oct. 2008.
provided via a secure escrow service [27]. [3] J. Driesen et al. The impact of vehicle-to-grid on the distribution grid.
Electric Power Systems Research, 81(1):185– 192, 2011.
VI. FURTHER DISCUSSIONS [4] K.J. Dyke et al. The impact of transport electrification on electrical
A smart charging of the roaming EVB on the premises of networks. Industrial Electronics, IEEE Trans,57(12):3917–3926, 2010.
[5] Wencong Su, H. Eichi, Wente Zeng, and Mo-Yuen Chow. A survey on
UA and payment for using this service involve the the electrification of transportation in a smart grid environment.
participation and interactions of multiple entities such as UA, Industrial Informatics, IEEE Transactions, 8(1):1–10, Feb. 2012.
UB, EVSEA, SMA, DCC, SA, SB, EVAGGB, DNO, TSO and [6] F. Geth et al. Impact-analysis of the charging of plug-in hybrid vehicles
DCC. These also generate different data items intended for on the production park in belgium. In MELECON 2010 - 15th IEEE
different users (entities) and with varying levels of sensitivity. Mediterranean Electrotechnical Conf., pages 425–430, Apr. 2010.
[7] A. Brooks, E. Lu, D. Reicher, C. Spirakis, and B. Weihl. Demand
The different entities should therefore have different access dispatch. Power and Energy Magazine, IEEE, 8(3):20–29, May, 2010.
privileges in accessing these data items. Any unauthorized [8] Willett Kempton and Jasna Tomić. Vehicle-to-grid power
access of the service and/or data generated by the service, or fundamentals: Calculating capacity and net revenue. Journal of Power
any malicious, selfish or dishonest attempt, by any of the Sources, 144(1):268–279, Jun. 2005.
entities could potentially harm the interest of other entities. In [9] Tu Yiyun, Li Can, Cheng Lin, and Le Lin. Research on vehicle-to-grid
technology. In Computer Distributed Control and Intelligent
addition, different households often register with different Environmental Monitoring, 2011 Int. Conf., p 1013– 1016, Feb. 2011.
suppliers (regardless of their physical location), some [10] National grid. Internet: www.nationalgrid.com/uk [24.04.2012].
households may change suppliers from time to time (to hunt [11] L. P. Fernández et al. Assessment of the impact of plug-in electric
for better deals), each household may have multiple EVs vehicles on distribution networks. Power Systems, IEEE Transactions,
registered with different occupants of the household, and some 26(1):206–213, Feb. 2011.
entities (e.g. SMA, EVSEA) may have limited computational [12] Y. Cao et al.An optimized ev charging model considering tou price and
soc curve. Smart Grid, IEEE Transactions, 3(1):388–393, Mar. 2012.
capabilities. All these issues and characteristics dictate that [13] S. Deilami et al. Realtime coordination of plug-in electric vehicle
existing security solutions, such as virtual private networks charging in smart grids to minimize power losses and improve voltage
and secure socket layer (SSL), are not readily applicable to the profile. Smart Grid, IEEE Transactions, 2(3):456–467, Sept. 2011.
smart EV charging application. More work is necessary to [14] O. Sundstrom and C. Binding. Flexible charging optimization for
investigate and design security solutions that at the same time electric vehicles considering distribution grid constraints. Smart Grid,
IEEE Transactions, 3(1):26–37, Mar. 2012.
satisfy the requirements specified in §V and accommodate the [15] Making the connection the plug-in vehicle infrastructure strategy.
aforementioned issues and characteristics of the application. Technical report, Department for Transport, June 2011.
[16] Ev network uk. Internet: www.ev-network.org.uk/ [24.04.2012].
The importance of securing the emerging SG is well [17] U.S. NIST, Guidelines for smart grid cyber security (vol. 1 to 3), NIST
recognized by standardization organizations such as NIST IR-7628, Aug. 2010.
[17], IETF [28] and international communities, e.g. ETSI [29]. [18] F. Marra, et al. Electric vehicle requirements for operation in smart
As EVs will be a major player in the SG, the security of EV grids. In Innovative Smart Grid Technologies (ISGT Europe), 2011 2nd
related applications should also be addressed. Conducting the IEEE PES Int. Conf. and Expo, pages 1–7, Dec. 2011.
[19] J.M. Jorgensen, S.H. Sorensen, K. Behnke, and P.B. Eriksen. Ecogrid
security analysis is the first step towards this direction. The eu - a prototype for european smart grids. In Power and Energy Society
requirements discussed in §V may serve as a benchmark for General Meeting, 2011 IEEE, pages 1–7, Jul. 2011.
designing and evaluating future secure solutions to support [20] Smart metering implementation programme: A consultation on the
smart EV charging. They may also be used for devising a risk detailed policy design of the regulatory and commercial framework for
model for quantitative analyses of risks and costs incurred to dcc. Technical report, DECC, 29 Sept. 2011.
[21] C.C. Chan. The state of the art of electric, hybrid, and fuel cell vehicles.
the various stakeholders in the EV charging system.
Proc. IEEE, 95(4):704–718, Apr. 2007.
VII. CONCLUSION [22] A.M. Foley et al. State-of-the-art in electric vehicle charging
infrastructure. In Vehicle Power and Propulsion Conf. (VPPC), 2010
Although EVs can provide a number of environmental and IEEE, pages 1–6, Sept. 2010.
financial benefits, their uncontrolled charging can destabilize [23] Z.M. Fadlullah, M.M. Fouda, N. Kato, Xuemin Shen, and Y. Nozaki.
the grid. This can be prevented by employing “smart” EV An early warning system against malicious activities for smart grid
communications. Network, IEEE, 25(5):50–55, Sept.-Oct. 2011.
charging. However, this new paradigm can bring a wide range [24] Yilin Mo et al., Cyberphysical security of a smart grid infrastructure.
of security issues. In this paper, we have devised a model of Proc. IEEE, 100(1):195–209, Jan. 2012.
smart EV charging and analyzed security problems and [25] The electricity trading arrangements: A beginners guide. Technical
potential security threats using this model. Based on this report, Elexon, Jul. 2009.
security analysis we have specified a set of security [26] William Stallings. Cryptography and Network Security Principles and
Practices. Prentice Hall, 4th edition, 2005.
requirements, which will be used to guide the next stage of our [27] C. Efthymiou and G. Kalogridis. Smart grid privacy via anonymization
work (i.e. the design of a reliable and secure payment service of smart metering data. In Smart Grid Communications (SmartGrid-
that supports EV roaming on private premises) towards the Comm), 2010 First IEEE Int. Conf., pages 238–243, USA, Oct. 2010.
design of a secure smart EV charging system. [28] IETF RFC 6272: Internet Protocols for the Smart Grid Internet:
www.tools.ietf.org/html/draft-baker-ietf-core [24.04.2012].
REFERENCES [29] ETSI. Machine-to-machine communications (m2m); threat analysis and
counter-measures to m2m service layer. Technical report, ETSI TR 103
167 V1.1.1, Aug. 2011.

View publication stats