Detection and Prevention of Denial-of-Service in
Cloud-based Smart Grid
Abdul Razaq
1
, Muhammad Majid Hussain
2a
, Waqas Javed
3
, Tasmiyah Javed and Zulfiqar A. Memon
4
1
School of Design and Informatics, Abertay University, Dundee, U.K.
2
Department of Electrical & Electronic Engineering, University of South Wales, Cardiff, U.K.
3
Department of Electrical Engineering, University of Engineering & Technology, Lahore, Rachna campus, Pakistan
4
Department of Electrical and Computer Engineering, Ajman University, Ajman, U.A.E.
Keywords: Cloud based Smart Grid, Denial-of-Service, IDS/IPS, Cyber-security.
Abstract: Smart Grid (SG), components with historical set of security challenges, becomes more vulnerable because
Information and Communications Technology (ICT) has its own share of problems while Cloud infrastructure
adds yet another unpredicted layer of threats. Scalability and availability, which are strong aspects of the cloud
platform making it attractive to users, also attracts security threats for the same reasons. The malware installed
on single host offers very limited scope compared to attack magnitude that compromised Cloud platform can
offer. Therefore, the strongest aspect of Cloud itself becomes a nightmare in Cloud-Based SG. A breach in
such a delicate system can cause severe consequences including interruption of electricity, equipment damage,
data breach, complete blackouts, or even life-threatening consequences. We mimic Denial-of-Service (DoS)
attacks to demonstrate interruption of electricity in SG with open-source solution to co-simulate power and
communication systems.
1 INTRODUCTION
SG promises reduced energy consumption, lower
production cost and robust transmission with
intelligent distribution. SGs are exposed to a
distinctive set of security threats because two
different systems of Power Electronics and ICT are
integrated to form a system with single core function:
“Uninterruptable and Cost-Effective Energy Supply”.
At the writing of this draft, not a single country has a
true and complete SG setup in place. Beside the
ordinary cost issues of installations or upgrading
existing systems; critical question of reliability is
unanswered. Cyber security for SG is of immense
concern because of emerging cyber-threats and
security incidents targeting smart grids all over the
world. These threats are severe and obvious in SG
systems if deployed without appropriate
measurements. A cyber-attack at Maroochy Shire
resulted in the release of untreated water and attack
on Davis-Besse nuclear plant, USA, disabled the
safety system (Yang, 2011). Stuxnet worm attack at
Iran’s nuclear power station halted the work and a
a
https://orcid.org/0000-0002-6255-6966
recent bug termed as HeartBleed’ in OpenSSL has
initiated debate in the research community for
dedicated security measures and re-assessment of
user grade software solutions for multibillion national
assets such as power grids.
Existing security measurements are insufficiently
scalable, incompatible, or simply inadequate to
address the challenges posed by highly complex
environments such as smart grid. Below are the
possible security threats originated by DoS attacks
and general aspects of toolsets to emulate such
models in terms of usability, extensibility, and
accuracy
1.1 Security and Privacy in Physical
Grid
The performance of power grid considerably relies on
the actual physical devices and deployed
environment; therefore, devices should be critically
designed to sustain adverse environmental factors
(Stouffer and Falco, 2006) and possible brutal force
attacks. Master terminal units (MTUs) and remote
172
Razaq, A., Hussain, M., Javed, W., Javed, T. and Memon, Z.
Detection and Prevention of Denial-of-Service in Cloud-based Smart Grid.
DOI: 10.5220/0010449901720179
In Proceedings of the 10th International Conference on Smart Cities and Green ICT Systems (SMARTGREENS 2021), pages 172-179
ISBN: 978-989-758-512-8
Copyright
c
2021 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
terminal units (RTUs) controlling programmable
logic controllers (PLCs) from generation to
transmission systems require strict timing to control
the demand and supply of electricity. Automation in
SG is commonly accomplished with Supervisory
Control and Data Acquisition (SCADA) system.
SCADA devices are real-time or non-real-time
(Muhammad and Siddique, 2020) small computer
system that manipulates electrical outputs based on
the condition of electrical input signals and program
logic. In theory, these systems consist of Human
Machine Interface (HMI) situated in operation
facilities which are part of MTU. MTU is used to
monitor RTU which is eventually connected to PLC
for automation (Liu and Xiao, 2012). Data
communication between RTUs and MTU occurs over
wired telephone lines or wireless cellular
technologies.
1.2 Security and Privacy in Information
and Communications Technology
ICT provided necessary communication means and
protocols for control and data exchange in SG.
Security of interconnected devices and subsystems is
important, but it should not result in a degraded and
unreliable system. Efficiency and reliability should be
achieved with only secure and completive open
solutions (Hahn and Member, 2013). The (Bou-Harb,
2013) investigated communication in distributed
networks and smart meters. They suggest that
distributed devices and applications should follow
design objectives including device authentication,
data confidentiality, message integrity, prevention
mechanism for cyber-attacks and facilitating
communication overhead.
Historically (Galloway and Hancke, 2013),
control in industrial automation was done
mechanically with hydraulic controllers or manually.
These mechanical components were upgraded once
electronics such as transducers, relays and hard-wired
control circuits became available. This changed to
new dimensions when small microcontrollers were
introduced which would allow smaller size and ability
to connect over wire or wireless. Electronics based
digital systems are able to remotely control and
monitor various power grid components with various
connectivity means and communication protocols.
These communications protocols are commonly
referred to as fieldbus protocols. Various protocols
and technologies are being used in traditional power
grids for networking purpose such as ModBus,
ModBus+, ProfiBus, ICCP, DNP3, PROFINET,
INTERBUS, WorldFIP, etc. It is important to note
that all these were designed without considering
cyber security. Existing deployed communication
protocols were developed under the umbrella of
IEEE, IEC and DNP3. IEC 60870-5 and DNP3 are
considered most widely used protocols in the
automation industry. IEC is typically used in
European countries and recognized by IEEE 1379
standard which is used in Asia and North America
(Liu and Xiao, 2012).
1.3 Security and Privacy in
Cloud-based Smart Grid
Cloud virtualization presents a logical view of
resources while encapsulating heterogeneity,
complexity, and distributed characteristics with
unified interface. Cloud computing provides virtually
aggregation of dispersed resources beside remote
accessibility. The emerging technologies in Cloud
based platforms present feasible solution for
processing and storage thriving domains such as SG.
Cloud-Based capability will allow ordinary resource
limited devices in SG with low power CPU and
limited memory to execute resource intensive
security protocols. It is also important to note that
Cloud platforms are prone to open access and
anonymous attacks
.
Data and software in cloud resides on network
which poses a unique set of security and privacy
challenges to the system. A survey conducted by
(Bera, 2015) provides comprehensive discussion on
Cloud applications for SG in contrast to conventional
solutions that exist without Cloud. Traditional
solutions need point to point physical connection
whereas cloud can provide virtual control of any node
without any real physical connection or media.
Similar (Genge and Beres, 2014) survey on existing
cloud–based solutions for SG without any details on
security criteria concludes that security in cloud
storage, communication paths between Cloud with
Grid and security within SG are basic system
requirements.
The (Markovic, 2013) provided comprehensive
survey on cloud in SG. An interesting comparison is
provided between conventional and smart grids and
how cloud solutions can effectively apply in
integration, reduced control latency, virtualized
energy sources and load management. Study
conducted on Los Angeles SG to address S&P in this
development investigates software layers on various
levels in cloud settings. The similar work is
investigated by (Simmhan, 2011)and emphasis on
data mining and aggregation of 1.4 million power
customers in municipality of United States of
America. Reporting, transporting, and storing huge
Detection and Prevention of Denial-of-Service in Cloud-based Smart Grid
173
data (Bigdata) without native support of encryption
can result in poor performance on Cloud platform. SG
user applications should address privacy threats while
utilizing datasets obtained continuously which
represent user behaviour.
A comprehensive report published by (Akyol,
2012) under the umbrella of Department of Defence,
United States concluded that power grids will
eventually deploy cloud computing and this shift will
be gradual only if the cyber security challenges are
addressed. Report further recommends that utility
providers must comply with NIST 2010 and NIST
2011 cyber security requirements analysis and design
methodologies. A mechanism implemented on
Eucalyptus open-source platform by (Baek, 2014)
based on ID, encryption and proxy validation
provides an ideal solution. However, cascading of
trust layers can prove to be dangerous if cooperating
layer is compromised. Proposed Smart-Framework is
interesting as far as it fulfils the trust criteria.
The (Alcaraz, 2011) investigated this in very
details of SCADA and how a virtual secure blanket
can efficiently execute console to hardware
commands with cryptic and efficiently. This virtual
control not only provides access to resources but also
guarantees scalability and availability with re-routing
solutions for communication infrastructure.
However, introducing virtual control layer over
critical physical instalments would introduce latency
which can result in fatal consequences. Recently,
(Armando, 2014) investigated smart metering data in
cloud with attribute-based control model. They have
used open-source Energy Home platform for
proposed security framework. Solution partially
addresses the security access problem assuming that
data transportation and storage is secured.
1.4 Denial-of-Service Threats in Smart
Grid
(He and Chan, 2014 ) proposed a mechanism based
on MicaZ and TelosB motes to resist DoS attacks
against adversaries and legitimate insiders. They
suggest that public key infrastructure (PKI) is valid
solution for uninterruptible service; however,
deployment of PKI is directly proportional with cost
for large scale networks such as SG. Different
security protocols have been suggested depending
upon the applications’ scalability and resources on
board. (Lu and Wang, 2010) reviewed the security
threats of DoS in DNP3 and classified these into three
types: network availability, data integrity and
information privacy, and evaluated their feasibility
and impact on the smart grid.
(Liu, 2013) investigated DoS attacks on load
frequency control (LFC) in SG while analysing
dynamic performance of communication channels
connected to RTUs in power systems. Case-studies of
simulated DoS attacks were modelled as a switched
(on/off) power system and two-area LFC theoretical
model was built for different attack-launching
instants. It has been concluded that adversaries can
make power system unstable via DoS attacks if
communication channels of RTUs are jammed.
Similar work by (Liu and Chen, 2013) also
investigated jamming threat but for wireless networks
in the power systems. They suggested that traditional
anti-jamming techniques can serve the purpose with
additional measurements. A local controller with
channel hoping is proposed to reduce possible DoS
threats.
In (Manandhar, 2014) investigated theory of false
data injection (DoS) attacks in power systems and
proposed Kalman Filter (KF). They suggest that DoS
attacks can be averted with linear quadratic
estimation (KF) detectors for sensors in SG such as
PMUs which measure current phase and amplitude in
power systems. The projected values by KF and
incoming instant values can be compared to detect an
anomaly in the system.
Recently, (Sgouras, 2014) presented qualitative
assessment of DoS attacks with simulation in
OMNeT++ and INET framework. They examined
performance of AMIs, routers and utility servers
under such situation. An attack on AMI would result
in minor consequences connected to single entity
whereas similar scenarios for utility server would
cause drastic effects during peak hours. Similar work
was conducted by (Yi, Zhu, 2014) to demonstrate the
impact of DoS in ICT without involving power grid
simulation. They termed DoS attack as puppet attack
which would penetrate in the system like warm and
continue to congest the communication channels with
false data until the network is exhausted.
1.5 Cloud based Intrusion Detection
System and Intrusion Prevention
System in Smart Grid
Intrusion Detection System (IDS) with Intrusion
Prevention System (IPS) or combined Intrusion
Detection and Prevention System (IDPS) is a system
which is meant to detect and prevent the unwanted
activity. In a cloud environment where infrastructure
and resources are in the form of services, a system of
IDPS should also adopt service-oriented design.
(Patel, 2013) presented a systematic review of
existing IDPS techniques and how they are
SMARTGREENS 2021 - 10th International Conference on Smart Cities and Green ICT Systems
174
insufficient for Cloud platform. Classic IDPSs are
unable to address security challenges that are
presented in mix-network-topology, multi-user, and
mixture of software layers in Cloud platforms.
(Mohamed and Adil, 2013) adopted to monitor
system calls generated by virtual machine's programs
to the hypervisor. This approach is limited to and
false alarm would be huge problem if legitimate, but
not-document application is executed. This technique
also suffers with typical system overloading and
lower throughput problem as monitoring all the
systems calls and their call sequence.
In Cloud computing IDS has its applicability on
Virtual machine level. The (Alina, 2013) proposed a
multilevel protection mechanism , the system
consists of front-end control and IDS on virtual
machines. Authors claim to reduce number of false
alarms with efficient detection of real threats. This
approach short falls if trust system is compromised, a
similar attack reported but with different weigh will
be undetected. Recently, (Li and Sun, 2012) proposed
anomaly-based solution with neural settings for
artificially sensing the activity. Neural configuration
helps to distribute the workload of IDS to spread
algorithm across Cloud rather chocking a single
machine. This technique of load balancing is natural
as far as all entities share the same interest and
participate with resources and time. The (Vaid and
Verma, 2014) proposed IDPS based on user
behaviour to detect malicious activities, the proposed
system is meant to be implemented as SaaS layer of
cloud. In (Maiti and Sivanesan, 2012) presented
simple but practical implementation of Cloud based
intrusion detection system. The presented idea
provides a simple yet robust example how cloud-
based security service can leverage on vast resources
for resource limited system.
A researcher (Tupakula, 2011) elaborated design
suggestions for virtual machine based IDPS targeted
for IaaS. Proposed design will require continuous
monitoring of operating systems calls and application
logging , this method is not ideal for systems with
strict timing needs and limited resources untilled data
mining techniques are adopted. Researchers
(Mehmood and Habiba, 2011) presented a
comprehensive discussion on different solutions for
Cloud security. Authors have suggested that IDS
should utilize multiple or hybrid solutions for
effective detection in Cloud security challenges.
Recently, (Prajapati, 2014) critically investigated
classical solutions and concluded that databases of
incident used to detect anomaly have low reliability.
This approach drastically increases the database size
and can introduce latency in monitor and detector
functionality. (Sathya and Vasanthraj, 2013) also
investigated Cloud based pattern matching IDPS and
have introduced multilevel detection system. Their
solution can suffer from overloading as auditing and
logging will require ample resources and monitoring
each activity will reduce overall system throughput.
In section 1 evaluate existing work in literature
review that encircles challenges presented in SG
system from power systems to communication
technologies and cloud. Section 1 also describes
possible security threats originated by DoS attacks
and general aspects of toolsets to emulate such
models in terms of usability, extensibility, and
accuracy. Section 2 represents a proposed technique.
Finally, work in progress is presented in section 3
with simulation of DoS in SG system for protected
and insecure settings before concluding with
discussion and future directions.
2 PROPOSED TECHNIQUES
In this section, we propose techniques to simulate
DoS attack and methodologies adopted to address this
attack (Abdul and Huaglory , 2016). Power and
network architecture is presented with specifications
of physical components and link layouts with
functional configurations. Several solutions were
evaluated as described in the literature review section.
Most of these solutions simulate SG with a
combination of network and power simulators (PS)
based on commercial and open-source tools.
Our proposed simulation uses NeSSi2 due to its
open-source license and ability to simulate power and
IP networks as a single application. NeSSi2 is a
scenario and profile-based simulation tool. Each
network in NeSSi2 consists of at least one scenario
which is eventually profiled depending upon required
simulation. A scenario defines type of profiles that
can be deployed on each node of the network.
Multiple profiles can be deployed on a single node
within a single scenario. A profile is a component to
provide a specific set of functionalities incorporating
single or various features relevant to power and IP
network simulation, which can be deployed onto SG
nodes. Finally, profiled scenario requires a simulation
component which allows the mapping of power and
network domains while linking the corresponding
entities. NeSSi2 is capable to generate various attack
scenarios and traffic analysis.
Figure 1 presents the high-level topology of
proposed SG simulation for both networks. Power
network is represented on the left side with the
corresponding IP network presented on the right side.
Detection and Prevention of Denial-of-Service in Cloud-based Smart Grid
175
Power network consists of one generator and two
consumption subnets representing insecure and
secure gird configurations. Parallel IP network also
consists of similar topology with server as a main
subnet connected to insecure and secure subnets.
Mapping between power and IP network is
configurable on node level within the simulation.
Figure 1: Proposed SG network topology.
Detailed configuration of proposed power
network based on is shown in figure 2. The power
grid consists of green generation based on solar panel
with output of 5147W. Two step-down transformers
with varying current of 380kV to 220kV are in place
between generation and consumption which depict
transmission and distribution. These swing bus
profiled transformers are eventually connected to two
consumption local grids: “Secure” and “Protected.”
As solar generation is considered unreliable, depends
on surrounding environment, a swing bus is used
between the links from solar panel to transmission
transformer. Swing bus accommodates system losses
by emitting or absorbing active/reactive power
to/from the system. The transformer connected to
insecure subnet is profiled with line failure profile to
simulate the unviability of load during power
interruption. Line failure profile of NeSSi2 allows
simulating the load unavailability in target power line
between required time intervals. This line failure
profile only accepts a single power link and can be
mapped to one instance of application. Consumption
subnets further consists of one transformer and two
smart houses. These smart houses are capable to
simulate load consumption and are mapped to
corresponding IP nodes in IP network for
communication simulation. Smart houses are capable
to emulate load usage depending upon time, weather,
and number of persons.
Configuration of proposed 100Mbps IP network
is presented in Figure 3. Main subnet or Server subnet
is connected to two subnets labelled as Insecure
Subnet and Secure Subnet. Server node with Server
subnet is profiled as Echo server application serving
both connected subnets.
Server subnet is connected to insecure subnet
without any protection mechanism in place, whereas,
secured subnet is connected via front-end firewall.
Moreover, secured subnet deploys additional firewall
at client's interface level in case router has been
compromised by malicious activity. DoS attacks
simulated via BOT component are presented in both
insecure and secure subnets. The firewall alone is not
a sufficient solution; an Intrusion Prevention System
(IPS) along with Intrusion Detection System (IDS)
must be carefully designed and deployed side-by-side
for complete protection. NeSSi2's firewall and packet
sniffer profiles are very limited which results in
restricted functionality.
Figure 2: Power network layout.
Figure 3: IP network layout.
SMARTGREENS 2021 - 10th International Conference on Smart Cities and Green ICT Systems
176
3 SIMULATIONS OF
DENIAL-OF-SERVICE
ATTACKS AND DETECTIONS
IN SMART GRID
All nodes of IP network are configured with default
load (echo client/server) at the beginning of
simulation, but the packet flow is increased to
demonstrate the packet loss which is turn depicts the
DoS. DoS or inability to serve the legitimate
incoming requests is a phenomenon where systems
capacity has reached its maximum throughput and
prompts to unavailability of given service. Ability to
handle DoS attacks is crucial due to the power electric
grid’s strict availability requirements. Botnet attack is
emulated in both secure and insecure networks for
DoS attack. The BOT profile of NeSSi2 is limited to
only specify the attack start time and can only target
single IP node. All these IP nodes are mapped to
smart houses in a power network with a one-to-one
relationship, which means that a house in a power
grid has a counterpart IP client in an IP network.
This simulation is executed for 1000 ticks and
failure or interruption of electricity and
communication is simulated between 105-350, 500-
600 and 800-900 ticks (time intervals). A tick is the
smallest possible time interval (event) in NeSSi2. The
actual time length depends on the simulation,
simulation mode and underlying hardware platform.
The solar panel model is set to produce 5147W peak
production whereas smart houses are simulated for 5
Figure 4: Server communication - secure and insecure
subnets.
Figure 5: Secure IP client.
persons each with 0.90% consumption of received
load.
Figure 4 provides simulation statistics of 1000
tick of server’s echo response to secure and insecure
subnet nodes. The communication packets are
presented along vertical axis whereas horizontal axis
present numbers of ticks or simulation time itself.
Interruption of insecure subnet is visible between
failed intervals with less density of packets compared
to tick intervals for secure subnet. The successful or
uninterrupted communication between server and
secure subnet can be verified in figure 5 which
demonstrated the continuous request sent by a secure
client.
Communication of insecure client is presented in
Figure 6. Packet drop statistics marked with cyan
colour simulates the failure scenario when client was
under attack from BOT and failed to process the echo
packets. Successful echo request is marked in yellow
colour whereas packets in magenta colour represent
forwarded packets. The successful and drop packets
can be compared with server’s statistics which
illustrates the fewer number of packets during failed
communication of insecure client. These failed
intervals can also be cross-checked in figure 7 which
presents the load statistic of insecure smart house.
The mapping between insecure client and smart house
is done prior to simulation and smart house is profiled
with smart house consumption profile along line
failure profile on transformer link level. The
corresponding IP client is profiled with echo client
and device failure.
Figure 6: Insecure IP client.
Figure 7: Power interruption of insecure house.
Detection and Prevention of Denial-of-Service in Cloud-based Smart Grid
177
4 DISCUSSION AND
CONCLUSIONS
Virtualization technologies in SG focus on
availability, aggregation and accessibility of
computing, storage, and network resources; without
consideration of security. A shift to cloud-based SG
is focused on the deliverability of services with
respect to cost and throughput. However, security
threats within these architectures not only remain the
same but it is even more severe when deployed in a
critical system of SG. Security in Cloud-based SG is
complex because of its multi-layered nature. Which
entity should be responsible for threat handling? Is it
applications, devices, services, or infrastructure
providers?
DoS is an instinctive security threat which include
hardware failure to service interruption. Efficiency to
revert DoS without considerable change in QoS is
important both during under-attack and afterwards.
Detection and prevention are a set of a single security
function divided into two conceptual sub-functions:
IDS and IPS. These two distinguishable methods
offer first line of defence in the event of DoS attack.
DoS and IDPS are also directly proportional in terms
of resources. The more intense the DoS attack, the
greater the resources of processing and data mining
techniques are required for IDPS. Cloud based
security-as-a-service technique for IDPS overcomes
this classic limitation.
We examined the importance of SG for future
energy requirements followed by detailed security
challenges that are presented in Cloud Based SG
systems. We demonstrated DoS attack and how this
can cause the interruption of electricity while
simulating the power and IP networks in parallel. We
also simulated ICT configuration for SG security that
can withstand DoS attack while ensuring the
uninterruptible communication and power. Finally,
we suggested that simulators for critical systems such
as SG should be able to simulate defence techniques,
including firewalls and IPS/IDS practices to avert
possible attacks.
REFERENCES
Akyol, B. A., 2012. Cyber security challenges in using
cloud computing in the electric utility industry. Pacific
Northwest National Laboratory.
Alcaraz, C., Agudo, I., Nunez, D., Lopez, J., .2011.
Managing incidents in smart grids a la cloud. IEEE 3rd
International Conference on Cloud Computing
Technology and Science. pp. 527-531.
Armando, A., Carbone, R., Chekole, E. G., Petrazzuolo, C.,
Ranalli, A., Ranise, S., 2014. Selective Release of Smart
Metering Data in Multi-domain Smart Grids. Intl.
Workshop on Smart Grid Security. pp. 48-62.
Baek, J., Vu, Q., Liu, J., Huang, X., Xiang, Y., 2014. A
secure cloud computing-based framework for big data
information management of smart grid. IEEE
Transactions on Cloud Computing. pp.233-344.
Bera, S., Misra, S., Rodrigues, J. J., 2015. Cloud Computing
Applications for Smart Grid: A Survey. IEEE
Transactions on Parallel and Distributed Systems. pp.
1477-1494.
Bou-Harb, E., Fachkha, C., Pourzandi, M., Debbabi, M.,
Assi, C., 2013. Communication security for smart grid
distribution networks. IEEE Communications
Magazine. pp. 42-49.
Galloway, B., Hancke, G. P., 2013. Introduction to
Industrial Control Networks. IEEE Communications
Surveys & Tutorials. pp.860-880.
Genge, B., Beres, A., Haller, P., 2014. A survey on cloud-
based software platforms to implement secure smart
grids. 49
th
International Universities Power
Engineering Conference. pp. 1-6.
Hahn, A., Member, S., Ashok, A., 2013. Cyber-Physical
Security Testbeds: Architecture, Application and
Evaluation for Smart Grid. IEEE Transactions on
Smart Grid. pp. 847.855.
He, D., Chan, S., Zhang, Y., Guizani, M., Chen, C., Bu, J.,
2014. An enhanced public key infrastructure to secure
smart grid wireless communication networks. IEEE
Network. pp. 10-16.
Hussain, M. M., Siddique, M., Raees, A., Nouman, M.,
Javed, W., Razaq, A., 2020. Power management through
smart grids and advance metering infrastructure. 6
th
IEEE
International Energy Conf. pp.767-772.
Liu, J., Xiao, Y., Li, S., Liang, W., Chen, C. L. P., 2012.
Cyber Security and Privacy Issues in Smart Grids. IEEE
Communications Surveys & Tutorials. pp.981-997.
Liu, S., Liu, X. P., Saddik, A. E., 2013. Denial-of-Service
(dos) attacks on load frequency control in smart grids.
IEEE PES Innovative Smart Grid Technologies
Conference. pp. 1-6.
Liu, H., Chen, Y., Chuah, M. C., Yang, J., 2013. Towards
self-healing smart grid via intelligent local controller
switching under jamming. IEEE Conference on
Communications and Network Security. pp. 127-225.
Lonea, A. M., Popescu, D. E., Tianfield, H., 2012.
Detecting DDos attacks in cloud computing
environment. International Journal of Computers,
Communications & Control. pp. 70-78.
Li, Z., Sun, W., Wang, L., 2012. A neural network based
distributed intrusion detection system on cloud
platform. IEEE 2nd International Conference on Cloud
Computing and Intelligence Systems. pp. 75-79.
Lu, Z., Lu, X., Wang, W., Wang, C., 2010. Review and
evaluation of security threats on the communication
networks in the smart grid. Milcom Military
Communication Conference. pp. 1830-1835.
Maiti, A., Sivanesan, S., 2012. Cloud controlled intrusion
detection and burglary prevention stratagems in home
SMARTGREENS 2021 - 10th International Conference on Smart Cities and Green ICT Systems
178
automation systems. 2
nd
Baltic Congress on Future
Internet Communications. pp. 182-186.
Manandhar, K., Cao, X., Hu, F., Liu, Y., 2014. Combating
false data injection attacks in smart grid using Kalman
filter. International Conference on Computing,
Networking and Communications. pp. 16-20.
Markovic, D .S., Zivkovic, D., Branovic, I., Popovic, R.,
Cvetkovic, D., 2013. Smart power grid and cloud
computing. Renewable and Sustainable Energy
Reviews. pp. 566-577.
Mehmood, Y., Habiba, U., Shibli, M. A., Masood, R., 2013.
Intrusion detection system in cloud computing:
challenges and opportunities. 2nd National Conference
on Information Assurance. pp. 59-66.
Mohamed, H., Adil, L., Saida, T., Hicham, M., 2013. A
collaborative intrusion detection and Prevention
System in Cloud Computing. AFRICON Conf. pp. 1-5.
Patel, A., Taghavi, M., Bakhtiyari, K., Júnior, J. C., 2013.
An intrusion detection and prevention system in cloud
computing: A systematic review. Journal of Network
and Computer Applications. pp. 25-41.
Prajapati, N. M., Mishra, A., Bhanodia, P., 2014. Literature
survey - IDS for DDoS attacks. Conference on IT in
Business, Industry and Government. pp. 1-3.
Sathya, G., and Vasanthraj, K., 2013. Network activity
classification schema in IDS and log audit for cloud
computing. International Conference on Information
Communication and Embedded Systems. pp. 502-506.
Simmhan, Y., Kumbhare, A. G., Cao, B., Prasanna, V.,
2011. An analysis of security and privacy issues in
smart grid software architectures on clouds. IEEE
International Conf. on Cloud Computing. pp.582-589.
Stouffer, K. K. K., Falco, J., 2006. Guide to Supervisory
Control and Data Acquisition (SCADA) and Other
Industrial Control System Security. NIST Guide to
Industrial Control Systems (ICS) Security.
Sgouras, K. I., Birda, A. D., Labridis, D. P., 2014. Cyber-
attack impact on critical Smart Grid infrastructures.
Innovative Smart Grid Technologies. pp. 1-5.
Tupakula, U., Varadharajan, V., Akku, N., 2011. Intrusion
detection techniques for infrastructure as a service
cloud. IEEE 9
th
International Conference on
Dependable, Autonomic and Secure. pp. 740-751.
Vaid, C., Verma, H. K., 2014. Anomaly-based IDS
implementation in cloud environment using BOAT
algorithm. Proceedings of 3rd International Conf. on
Reliability, Infocom Technologies and Optimization.
pp. 1-6.
Yang, Y., Littler, T., 2011. Impact of cyber-security issues
on smart grid. IEEE PES Innovative Smart Grid
Technologies Conference Europe. pp. 1-7.
Yi, P., Zhu, T., Zhang, Q., Wu, Y., Li, J., 2014. A denial-
of-service attack in advanced metering infrastructure
network. IEEE International Conference on
Communications. pp. 1029-1034.
Abdul, R., Huaglory, T., Bernardi P., Hong, Y., 2016.
Simulating smart grid cyber security. Taylor and
Francis group: Smart Grid Networking, Data
Management, and Business Models. pp. 97-116.
Detection and Prevention of Denial-of-Service in Cloud-based Smart Grid
179