Page 367 - From Smart Grid to Internet of Energy

Big data, privacy and security in smart grids Chapter 8 331

the datasets to acquire other users’ data and track the private data. The most
recent AI systems improve the reliability of big data analytics to tackle this
challenge.
In big data applications, encryption mandatory is a convenient method to
provide end-to-end protection for the security of the data centers. The integrity
of data also can be used to prevent data falsification for diverse end-point data.
The underlying attacks may bypass the control algorithms and provide direct
access to databases. Therefore, it is required to decrease visibility of any data
at the source databases. The big data analytics require a fine-grained and scal-
able database access control that is provided by granular access control proto-
cols. It enforces the system by preventing unauthorized access and maintaining
privacy. The privacy preserving methods used in data management are related
with secure data storage, granular audits, and data provenance [5].
The data source and users are separated in big data infrastructure where the
user have not complete control of data and is not able to know the exact location
of stored data. The transaction logs are generated to locate stored data informa-
tion that reach to higher volumes with the increment of stored data size. There-
fore, security of the stored data and transaction logs should be ensured. Another
important privacy challenge is seen in real time security monitoring applica-
tions since it is based on tracking the dynamic analytics. It is required to prevent
intrusions and unauthorized accesses to big data infrastructure. The monitoring
applications are sensitive to denial of service (DoS) attacks, and advanced intru-
sion attacks. Therefore, instant and persistent real-time monitoring applications
are convenient to predict and prevent intrusions [5].

8.4.3 Privacy enhancing applications
The sophisticated features of smart grids such as self-healing, remote control,
self-monitoring, and distributed control properties have increased the attention
to this new kind of power network. The widespread use of services and oppor-
tunities of smart grid have brought concerns on protecting the acquired data
which include personal information of consumers. The collected data could
be used to define personal behaviors, lifestyle, and habits. The demand response
programs which are tailored solutions for each consumer is one of the smart grid
application that privacy and security should be ensured in big data analytics.
Other smart grid applications provide personal data about consumers as well
as demand response or smart meter data. Nowadays, several privacy preserving
and secrecy techniques have been taken into consideration to protect such crit-
ical data. The prominent applications are listed as anonymization, access con-
trol, encryption, differential privacy protecting methods and so on [5, 28].
The anonymization is one of the most widespread protection method that
targets to hide user ID and sensitive personal data. The raw data is anonymized
during processing and before deploying to nodes by using generalization,
decomposition, replacement and interference operations. The generalization