A Survey on Privacy-Preserving Data Aggregation Without Secure Channel

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 03 Issue: 01 | Jan-2016 www.irjet.net p-ISSN: 2395-0072
© 2016, IRJET ISO 9001:2008 Certified Journal Page 133
A SURVEY ON PRIVACY-PRESERVING DATA AGGREGATION WITHOUT
SECURE CHANNEL
Kshitija Nandgaonkar1, Swarupa Kamble2
1 M.E. Student, Computer Engg ., RMD Sinhgad School of Engineering, Pune ,Maharashtra, India
2 Assistant Professor ,Dept. Computer Engg. ,RMD Sinhgad School of Engineering, Pune, Maharashtra ,India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The Privacy-preserving data aggregation
problem becoming an important issue in the field of
applied cryptography. Most research has been carried
out to securely outsource individual's privately owned
data to an untrusted aggregator, or to enable multiple
parties to jointly aggregate their sensitive data while
preserving privacy. However many research require
secure pair-wise channel and it suffers from high
complexity. This paper describes sum and product
calculation protocol that enables an external
aggregator or multiple parties to perform data
aggregation over participants data while preserving
the data privacy.
Key Words: Privacy-preserving, Data aggregation,
Secure channels, SMC, Homomorphic.
1. INTRODUCTION
As people are becoming more concerned about their
privacy these days, the privacy-preservability is very
important. A fundamental problem is that of private data
analysis where a third party has to compute some
aggregate statistics over some sensitive data held by
individuals. This problem finds concrete applications in a
number of situations. When the third party, called
hereafter aggregator, is trusted an easy solution would be
to ask the users to encrypt their data using the
aggregator's public key. Upon receiving the ciphertexts the
aggregator applies its private key to recover the data in
clear and then compute statistics. The problem becomes
much more challenging in the case of an untrusted
aggregator.
In many real life applications such as crowd sourcing or
mobile cloud computing, individuals need to give their
delicate data (location-specific or personal information
related) to get particular services from the entire system
(e.g., location based services or mobile based social
networking services).
The data aggregation problem usually involves two
different models:
 an external aggregator will gather the data and
performs an aggregation function on participants'
data (e.g., crowd sourcing);
 participants will together calculate a specific
aggregation function where input data being
provided by themselves (e.g., social networking
services).
However, the individual's data should be kept secret, and
the aggregator or other participants are not supposed to
learn any useful information about it. Secure Multi-party
Computation (SMC), Homomorphic Encryption (HE) and
other cryp tographic methodologies can be employed to
solve this problem, but these techniques are subject to
some limitations in this problem. Many real-world
applications have benefitted tremendously from the ability
to collect and mine data coming from multiple individuals
and organizations. These applications have also spurred
numerous concerns over the privacy of user data.
In this paper, we study how an untrusted aggregator or
mutiple parties can gather information and learn
aggregate statistics over individual privacy. For example,
consider a smart grid operator who wishes to track the
total electricity consumption of a neighborhood every 15
minutes, for scheduling and optimization purposes. Since
such power consumption data can reveal sensitive
information about individuals presence and activities, we
wish to perform such aggregation in a privacy-preserving
manner.
2. LITERATURE SURVEY
C. Castelluccia, A. Chan, E. Mykletun, and G. Tsudik has
published paper on Efficient and provably secure
aggregation of encrypted data in wireless sensor network
[1], in this authors had designed a symmetric key
homomorphic encryption scheme which is an addition to
homomorphic for conducting the aggregation operations
on the ciphertexts. They uses a modular addition, so the
scheme is good for CPU bounded devices such as sensor

nodes in WSN. Their scheme can also efficiently compute
various statistical values such as mean, variance and
deviation. However, since they used the symmetric
homomorphic encryption, their aggregator could decrypt
each individual sensors data, and they assumed the
trusted aggregator in their model.
R. Sheikh, B. Kumar, and D. Mishra, presented a paper on
Privacy preserving k secure sum protocol [2], in this
paper author proposed a k-secure sum protocol, which is
extended from the work of Clifton et al. [3]. They
significantly reduced the probability of data leakage
occurring in [3] by segmenting the data block of individual
party, and distributing segments to other parties. Here,
sum of each party’s segments is his data, therefore the
final sum of all segments are sum of all parties data. This
scheme can be easily converted to k-secure product
protocol by converting each addition to multiplication.
However, pair-wise unique secure communication
channels should be given between each pair of users such
that only the receiver and the sender know the
transmitted segment. Otherwise, each party’s secret data
can be calculated by performing O(k) computations. In this
paper, we remove the limitation of using secure
communication channels.
T. Jung, X. Mao, X.-y. Li, S.-J. Tang, W. Gong, and L. Zhang,
published a paper on Privacy-preserving data aggregation
without secure channel: multivariate polynomial
evaluation [4], in this paper author assume that all the
communication channels in protocol are insecure. Anyone
can eavesdrop them to intercept the data being
transferred. To address the challenges of insecure
communication channel, they assume that the discrete
logarithm problem is computationally hard if: 1) the
orders of the integer groups are large prime numbers; 2)
the involved integer numbers are large numbers. The
security of their scheme relies on this assumption. They
further assume that there is a secure pseudorandom
function (PRF) which can choose a random element from a
group such that this element is computationally
indistinguishable to uniform random.
M. Joye and B. Libert, has published paper on A scalable
scheme for privacy-preserving aggregation of time-series
data [5] in this authors have proposed a solution for
accommodating large plaintext data from multiple users.
In this paper they address problem of private data analysis
where a third party has to compute some aggregate
statistics over some sensitive data held by individuals. A
scheme supports a large plaintext spaces and number of
users. It also allows the decryption algorithm to operate
in constant time, regardless of the number of users.
Additionally the scheme also provides a on-line/off-line
efficiency using pre-computations, the encryptor is left
with a mere modular multiplication in the on-line phase
(i.e., when the data to be encrypted is known), which is
highly desirable when computations take place on
resource-limited devices. This paper presented a new
scheme allowing an untrusted aggregator to evaluate the
sum of user’s private inputs. In contrast to prior solutions,
there is no restriction on the message space or on the
number of users. This results in always fast decryption
and aggregation, even over large plaintext spaces and/or
population of users
We note that Dong et al. [6] investigated verifiable privacy
preserving dot production of two vectors and Zhang et al.
[7] proposed verifiable multiparty computation, both of
which can be partially or fully exploited later. Designing
privacy preserving data aggregation while providing
verification of the correctness of the provided data is a
future work.
Shi et al. [8] proposed a construction that n participants
periodically upload encrypted values to an aggregator, and
the aggregator computes the sum of those values without
learning anything else. This scheme is close to our
solution, but they assumed a trusted key dealer in their
model. In this paper, the trusted aggregator in [1] is
removed since data privacy against the aggregator is also a
top concern these days. Unlike [2], we assumed insecure
channels, which enabled us to get rid of expensive and
vulnerable key pre-distribution. We did not segment each
individuals data, our protocols only incur constant
communication overhead for each participant. Our scheme
is also based on the hardness of the discrete logarithm
problem like [8], but we do not trivially employ brute-
force manner in decryption, instead, we employ our novel
efficient protocols for sum and product calculation.
3. PROBLEM STATEMENT
3.1 Problem Analysis
Assume that there are n participants { p1,p2,....,pn } and
each participant pi has a privately known data xi from Zp.
The privacy-preserving data aggregation problem is to
compute sum or product of xi jointly or by an aggregator
while preserving the data privacy. That is, the objective of
the aggregator or the participants is to compute the
following polynomial without knowing any individual xi:
Or
Here vector x = (x1; x2,....xn). For simplicity, we assume
that the final result f(x) is positive and bounded from
above by a large prime number P.

3.2 Security Model
Firstly, we assume that all the communication channels in
our protocol are insecure. Anyone can eavesdrop them to
intercept the data being transferred. To address the
challenges of insecure communication channel, we assume
that the following CDH problem is computationally
intractable, i.e., any probabilistic polynomial time
adversary has negligible chance to solve the following
problem:
3.2.1 Definition 1 (CDH Problem in G)
The Computational Diffe-Hellman problem in a
multiplicative group G with generator g is defined as
follows:
given only g, ga, gb ε G where a, b ε Z, compute gab without
knowing a or b.
Additionally, similar Decisional Diffe-Hellman(DDH)
problem is defined as follows:
3.2.2 Definition 2 (DDH Problem in G)
The Decisional Diffe-Hellman problem in a multiplicative
group G with generator g is defined as follows:
given only g, ga, gb , gc ε G where a, b, c ε Z, decide if gab =
gc.
Our protocol is based on the assumption that it is
computational expensive to solve the CDH problem. Then,
we define the security of our privacy-preserving sum and
product calculation as follows.
3.2.3 Definition 3 (CDH-Security in G)
We say our privacy-preserving (sum or product)
calculation is CDH-secure in G if any Probabilistic
Polynomial Time Adversary (PPTA) who cannot solve the
CDH problem with non-negligible chance has negligible
chance to infer any honest participants private value in G.
4. SYSTEM ARCHITECTURE AND DESIGN
There are two models for aggregation of private data as:
One Aggregator Model and Participants Only Model. These
two models are general cases we are faced with in real
applications
Fig. 1 – System Architecture
A figure shows a system architecture of data aggregation.
In this data comes from different user like medical data,
location based data, or personal information etc. has been
aggregated by external aggregator. An external aggregator
or multiple parties will perform encryption on aggregated
data to maintain data privacy. The aggregator or
participants need to learn privacy preserving sum and
product calculation protocol.
4.1 One Aggregator Model
In the first model, we have one aggregator A who wants to
compute the function f(x). We assume the aggregator is
untrustful and curious. That is, he always eavesdrops the
communications between participants and tries to access
their input data, but also follow the protocol specification.
We also assume participants do not trust each other and
that they are curious as well, i.e., they also eavesdrop all
the communications and follow the protocol specification.
In this model, any single participant pi is not allowed to
compute the final result f(x).
4.1.1 Sum and Product Calculation
One Aggregator Model can be used to calculate product
and sum of participants privately owned data. In this
aggregator A acts as n+1th participant and will compute
sum and product for each participants. Here, each
participant will broadcast its ciphertext data to
aggregator.
4.2 Participant Only Model
The second model is similar to the first one except that
there are n participants only and there is no aggregator. In
this model, all the participants are equal and they all will
calculate the final aggregation result f(x).
4.2.1 Sum and Product Calculation
In participant only model, all participant will jointly
compute sum and product of each participants data. In this
data is share among all participants of communication. To
encrypt data a random key is generated and it is share
among all participants.
5. CONCLUSION
This paper presents a privacy preserving sum and product
calculation of privately owned data without need of secure
channel. Two models can be used for aggregation of data:
One aggregator model and Participant only model. A
survey describes that, sum and product of sensitive data
can be computed without the need of pair wise secure
channel. Thus reduces overhead of maintaining each
participants key pair.

REFERENCES
[1] C. Castelluccia, A. Chan, E. Mykletun, and G. Tsudik,
“Efficient and Provably secure aggregation of
encrypted data in wireless sensor networks”,
Transactions on sensor Networks(TOSN), 2009
[2] R. Sheikh, B. Kumar, D. Mishra, “ Privacy preserving k-
secure Sum Protocol”, Arxiv Preprint arXiv:0912.0956,
2009
[3] C. Clifton, M. Kantarcioglu, J. Vaidya, X. Lin, and
M.Zhu, “ Tools for privacy preserving distributed data
mining”, SIGKDD Explorations Newsletter,2002
[4] T. Jung, X. Mao, X.-y. Li, S.-J. Tang, W. Gong, and L.
Zhang, “ Privacy-preserving data aggregation without
secure channel: multivariate polynomial evaluation,”
in INFOCOM. IEEE, 2013.
[5] M. Joye and B. Libert, “A scalable scheme for privacy-
preserving aggregation of time-series data”, in
Financial Cryptography and Data Security, (IFCA)
2013.
[6] W. Dong, V. Dave, L. Qiu, and Y. Zhang, “Secure friend
discovery in mobile social networks,” in IEEE
INFOCOM, 2011, pp. 16471655.
[7] L. Zhang, X. Li, Y. Liu, and T. Jung, “ Verifiable private
multi-party computation: ranging and ranking”, in
INFOCOM Mini-conference, IEEE, 2013
[8] E. Shi, T. Chan, E. Rieffel, R. Chow, and D. Song, “
Privacy-preserving aggregation of time-series data”,
in NDSS, vol.17, 2011.

A Survey on Privacy-Preserving Data Aggregation Without Secure Channel

More Related Content

What's hot (18)

Similar to A Survey on Privacy-Preserving Data Aggregation Without Secure Channel (20)

More from IRJET Journal (20)

Recently uploaded (20)

A Survey on Privacy-Preserving Data Aggregation Without Secure Channel