You are on page 1of 8

International Journal of Trend in Scientific

Research and Development (IJTSRD)

International Open Access Journal
ISSN No: 2456 - 6470 | | Volume - 2 | Issue – 3

Efficiently Data Analysis and T

ransmission for
consumer using grid computing

Sindhuja R Vasuinthira N
Research Scholar, A.V.C College of Arts and Science, Asst. Professor, A.V.C College of Arts and Science,
Mayiladuthurai, Tamil Nadu, India Mayiladuthurai,, Tamil Nadu, India


Research focuses on the issue of wireless big data three levels and also enables the intra-communication
computing in smart grid. Smart grid means electrical within each level. Experimental results indicate that
network architecture is purposed for generating, the proposed storage planning scheme significantly
distributing and administering efficiently the power reduces the cost for consumers.
consumption to end-users.users. First investigate the
consistency between the characteristics tics of big data Keywords: smart grid, data sources, data
and smart grid data. A propose a hybrid approach for transmission, data storage, data analysis
storage planning, which consists of an outer
optimization based on genetic and an inner
optimization algorithm for energy scheduling. The rapidly growing wave of wireless data service is
Propose a big data computing architecture for smart pushing against the boundary of the communication
grid, consisting of four main levels: data sources, data network’s processing power. The pervasive and
transmission, data storage, and data analysis. First exponentially increasing data traffic present imminent
level, Data Source Level: data are generated from challenges to all the aspects of the wireless system
different data sources which are distributed, that is, design, such as spectrum efficiency, computing
distribution and transmission data, distr distributed capabilities and front haul/backhaul link capacity. In
generation data, intelligence application related data, this article, they discuss the challenges
c and
and so on. These data are distributed in different opportunities in the design of scalable wireless
places and managed by different systems to embrace such a “big data” era. On one
companies/departments, which belong to different hand, they review the state-of-the-art
state networking
systems. Next, Data Storage Level: data collected architectures and signal processing techniques
from the data ta sources contain a lot of meaningless adaptable for managing the “big data” traffic in i
information, which does not need to occupy a large wireless networks. On the other hand, instead of
amount of storage space. Therefore, the data in a data viewing mobile big data as a unwanted burden, they
center may be stored and processed. This level is introduce methods to capitalize from the vast data
suitable for the planning of smart grid with low real real- traffic, for building a big data--aware wireless network
time demands.
emands. Third level, Data Analysis Level: with better wireless service quality and new mobile
Various analytical methods and tools have been applications. They highlight several promising future
utilized to extract value in different application fields. research directions for wireless communications in the
It categorized by two division one is costumer costumer- mobile big data area. Decades of exponential growth
oriented service another one is utility utility-company- in commercial data services has ushered in the so- so
oriented service.
ice. Final level, Data Transmission called “big data” era, to which the expansive
expa mobile
Level: This level builds a bridge among the other wireless network is a critical data contributor. As of

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr

Apr 2018 Page: 643
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
2014, the global penetration of mobile subscribers has systems, which are mainly designed for content
reached 97%, producing staggeringly of mobile data delivery, instead of analyzing and making use of the
worldwide. The surge of mobile data traffic in recent data traffic.
years is mainly attributed to the popularity of smart
phones, phone cameras, mobile tablets and other With the increasing adoption of cloud computing, a
smart mobile devices that support mobile broadband growing number of users outsource their datasets into
applications, e.g., online music, video and gaming as cloud. The datasets usually are encrypted before
shown. With a compound annual growth rate of over outsourcing to preserve the privacy. However, the
40%, it is expected that the mobile data traffic will common practice of encryption makes the effective
increase by 5 times. utilization difficult, for example, search the given
keywords in the encrypted datasets. Many schemes
In addition to the vast amount of wireless source data, are proposed to make encrypted data searchable based
modern wireless signal processing often amplifies the on keywords. However, keyword-based search
system’s pressure from big data in pursuit of higher schemes ignore the semantic representation
performance gain. For instance, MIMO antenna information of user’s retrieval, and cannot completely
technologies are now extensively used to boost meet with users search intention. Therefore, how to
throughput and reliability at both mobile terminals design a content-based search scheme and make
(MTs and base stations (BSS) of high speed wireless semantic search more effective and context-aware is a
services. This, however, also increases the system difficult challenge. In this thesis, they proposed an
data traffic to be processed in proportion to the innovative semantic search scheme based on the
number of antennas in use. Moreover, the 5G (the concept hierarchy and the semantic relationship
fifth generation) wireless network presently under between concepts in the encrypted datasets. More
development is likely to migrate the currently specifically, they scheme first indexes the documents
hierarchical , BS-centric cellular architecture to a and builds trapdoor based on the concept hierarchy.
cloud-based layered network structure, consisting of a To further improve the search efficiency, utilize a
large number of cooperating wireless access points tree-based index structure to organize all the
(APs) connected by either wire line or wireless front document index vectors. They experiment results
haul links to a big data capable processing central unit based on the real world datasets show the scheme is
(CU). New wireless access structures, such as more efficient than previous scheme. They also study
coordinated multipoint (networked MIMO). the threat model of they approach and prove it does
not introduce any security risk.
On the other hand, timely and cost-efficient
information processing is made possible by the fact Cloud computing is a new but increasingly mature
that the vast-volume mobile data traffics are not model of enterprise IT infrastructure that provides
completely chaotic and hopelessly beyond high quality applications and services. The cloud
management. Rather, they often exhibit strong customers can outsource their local complex data
insightful features, such as user mobility pattern, system into the cloud to avoid the overhead of
spatial, temporal and social correlations of data management and local-storage. However, the security
contents. These special characteristics of mobile of outsourced data cannot be guaranteed, as the Cloud
traffic present us with opportunities to harness and Service Provider (CSP) possesses whole control of the
exploit “big data” for potential performance gains in data. So, it is necessary to encrypt data before
various wireless services. To effectively utilize and outsourcing them into cloud to protect the privacy of
exploit these characteristics, they should be identified, sensitive data. However, encryption for outsourced
extracted, and efficiently stored. For instance, caching data can protect privacy against unauthorized
popular contents at wireless hot spots could behaviors; it also makes effective data utilization,
effectively reduce the real-time traffic in the front such as search over encrypted data, a very difficult
haul links. Additionally, network control decisions, issue. In recent years, many researchers have
such as routing, resource allocation, and status proposed a series of efficient search schemes over
reporting, instead of being rigidly programmed, could encrypted cloud data. The general process of search
be made data-driven to fully capture the interplay scheme is divided into five steps: extracting document
between big data and network structure. Presently, features, constructing a searchable index, generating
however, these advanced data-aware features could search trapdoor, searching the index based on the
not be efficiently implemented in current wireless trapdoor and returning the search results. These search

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 644
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
schemes provide different query capabilities, important and challenging task to implement semantic
including single keyword search, multi-keyword search over encrypted data.
search, fuzzy keyword search, similarity search, and
so on. However, all the existing searchable encryption Distributed devices in smart grid systems are
schemes, which consider keywords as the document decentralized and connect to the power grid through
feature, do not take the semantic relations between different types of equipment transmit, which will
words into consideration, both in the steps of produce numerous energy losses when power flows
extracting document features and generating search from one bus to another. One of the most efficient
trapdoor. As they all know, the semantic relations approaches to reduce energy losses is to integrate
between words are diverse, such as synonymy, distributed generations (DGs, mostly renewable
domain correlation. Considering the potentially huge energy sources). However, the uncertainty of DG may
amount of outsourced data documents in the cloud, cause instability issues. Additionally, due to the
the search accuracy and search efficiency are similar consumption habits of customers, the peak
influenced negatively if the semantic relations load period of power consumption may cause
between words are not handled well. congestion in the power grid and affect the energy
delivery. Energy management with DG regulation is
In this paragraph, they make a detailed description of considered to be one of the most efficient solutions
existing problems of the available searchable for solving these instability issues. In this thesis, they
schemes. Firstly, in the stage of extracting document consider a power system with both distributed
features, the data owner computes the weight of each generators and customers, and propose a Distributed
word in a document and then selects t words with top- Locational Marginal Pricing (DLMP)-based unified
t weights as the feature of the document. In the Energy Management System (UEMS) model which,
process shown above, every two words with different unlike previous works, considers both increasing
spelling are assumed uncorrelated, which is profit benefits for DGs and increasing stability of the
unreasonable. For example, three words “network”, distributed power system (DPS).
“net”, “web” are different in the perspective of
spelling, but they are semantically related, that is they The model contains two parts: a game theory based
are synonym. It is obvious that the weight of word is loss reduction allocation (LRA) and a load feedback
influenced if semantic relations between words are control (LFC) with price elasticity. In the former
ignored and the accuracy of the document features is component, they develop an iterative loss reduction
influenced consequently. Secondly, in the stage of method using DLMP to remunerate DGs for their
generating search trapdoor, the trapdoor is generated participation in energy loss reduction. By using
only based on the search keywords inputted by the iterative LRA to calculate energy loss reduction, the
data user, which is inflexible, because it is impossible model accurately rewards DG contribution and offers
to extend the search keywords when the data user a fair, competitive market. Furthermore, the overall
cannot well express his search intention. In this case, profit of all DGs is maximized by utilizing game
it is likely that a returned document is not needed for theory to calculate an optimal LRA scheme for
the data user or the really needed documents are not calculating the distributed loss of every DG in each
returned. So, it is important to understand the real time slot. In the latter component of the model, they
search intention of the data user to avoid returning propose an LFC sub-model with price elasticity,
unnecessary documents to improve search efficiency, where a DLMP feedback signal is calculated by
as the size of the document set outsourced into the customer demand to regulate peak-load value. In
cloud server is potentially huge. Thirdly, a search UEMS, LFC first determines the DLMP signal of a
request usually focuses on a theme, and some search customer bus by a time-shift load optimization
words can be considered to be the attribute of the algorithm based on the changes of customer demand
theme, for example, “birthday” is an attribute of a which is fed back to the DLMP of the customer bus at
people. In existing search schemes, an attribute value the next slot-time, allowing for peak-load regulation
is usually treated as a keyword which ignores the via price elasticity. Results based on the IEEE 37-bus
relationship with the mean results in larger keyword feeder system show that the proposed EMS model can
dictionary, and then negatively influences the search increase DG benefits and improve system stability.
accuracy and efficiency. Therefore, it is a very
The smart grid is seen as the best approach in
modernization of electric power systems, and the DPS

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 645
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
is one of the most significant parts. The decentralized and creates trustworthiness, leveraging the interaction
devices in the DPS will produce high system loss among things. With the popularity of smart phones
while power flow transmit from one bus to another and other intelligent devices, objects can easily
through the branch or some power devices such as communicate with each other without the restriction
substation, and the similar power-consuming habit of of time and place. There is a significant trend that
customer, which is the main reason to cause peak-load takes full advantage of mobile devices and
period of power consumption, may aggravate the loss applications to support crowd sourcing to solve a
reduction problem. Thus, energy management system diversity of issues involving real-time data
should be designed to dynamically adapt to DPS by acquisition, processing, and collaborations among a
controlling and regulating distributed devices to make large number of participants in mobile environments.
power systems more effective and reliable. As for the
application of energy management, using load Obviously, crowd sourcing can bring not only great
forecasting to regulate energy distribution in micro- social and economic benefits, but also real changes in
grid is very hot in recent research. The authors in the daily lives. In traditional mobile crowd sourcing
presented an electric load forecast architectural model networks, the crowd sourcing tasks are received by
to integrate distributed renewable sources. This is to participants who possess mobile devices with
balance power generation and demand in order to computing power or sensing capacity. To amplify the
offer a good service at a competitive price. Integration function of mobile crowd sourcing networks, they
of DGs in DPS could greatly enhance the introduced the SIOT paradigm into crowd sourcing, in
competitiveness power of distributed companies in a which objects as sensing entities can discover services
competitive environment, and provide benefits for and resources through forming social relationships.
energy loss reduction and improve stability in the This provides a crowd sourcing solution in which
grid. mobile devices transform its role for participants into
sensing entities.
In order to perform some types of complex tasks, the
principles of crowd sourcing have been applied in Efficiently Data Analysis and Transmission for
constructing a mobile crowd sourcing network with a consumer using grid computing:
huge amount of human involvement and powerful
mobile devices. In addition, through the social
Internet of Things (SIOT) paradigm, the capability of The hybrid approach combines the semantic, corpus,
humans and devices to discover and exploit objects ontology and relational based approaches. A novel
with their services is augmented. Thus, SIOT has hybrid approach extracts semantic knowledge from
become a promising network for service discovery the structural representation of Word net and the
and composition. By introducing SIOT, they can statistic information on the Internet. Internet based
expand the traditional mobile crowd sourcing network semantic knowledge estimates the semantic similarity
and achieve broader crowd sourcing application. between the two concepts in Word net. A useful
However, winner selection and payment measure called Normalized Google Distance (NGD)
determination by evaluating the reliability of crowd computes the semantic distance between the adjacent
sourcing participants. Finally, some challenges in concepts, along the shortest path in Word Net using
trustworthy crowd sourcing are discussed. Internet semantic knowledge. It is one of the best
approaches due to the deployment of Internet
With the rapid development in computing and
knowledge in Word Net based semantic relatedness
communication technologies, more and more objects
measure. Between sentences and documents. This
are being equipped with processors, sensors,
work measures the similarity between the documents
memories, and wireless communication modules that
using both ontology-based similarity model and
enable object-to-object communication, forming the
counts based similarity model. Shows the components
Internet of Things (IOT). It is demonstrated that many
used in this methodology. Initially, the proposed
issues need to be considered in order to implement the
system represents the input documents as a bag of
effectiveness and reliability of IOT, such as the
words, and it avoids the repeated terms. The pre-
sensing world, transmitting data, and building
processing step removes stop words from the
applications based on relevant services. Similar to
representation. The important keywords in the
social networks for human beings, the social IOT
documents are identified using the term frequency
(SIOT), introduces social relationships among things

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 646
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
mechanism. Term frequency determines the similarity measure estimates the similarity between
frequently occurring words in a document as the documents by dividing the number of commonly
keywords. The derived keywords are given to the related keywords by the total number of total
ontology to obtain related keywords. The related keywords.
keywords are given to the Jaccard similarity
coefficient to decide the documents are similar or not.
The Jaccard similarity coefficient is the count based
Architecture view:


@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 647
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Problem definition: Demerits:
The cloud services technology as the basis for  It does not facilitate both the large and small-
communication and management mechanism in order scale network.
to provide a strong support for power monitoring and  Implementation cost is high.
early-warning system. With the fast development of  It does not satisfy the consumer requirement.
smart grid, higher storage capacity and stronger  It does not flexible for on demand request.
processing ability of power system platform are
needed .A smart grid can be conceptualized as a Proposed System:
cyber-physical system that connects physical Propose a hybrid approach for storage planning,
electricity systems and cyber-infrastructure, with the which consists of an outer optimization based on
integration of the Internet. This service can Genetic and an inner optimization algorithm for
communicate with the consumer appliances and also energy scheduling. Propose a big data computing
provide the backbone for utilities to assimilate content architecture for smart grid, consisting of four main
and control operations. With the presence of online levels: data sources, data transmission, data storage,
connectivity, smart grids have greater exposure to and data analysis. First level, Data Source Level: data
cyber-attacks that can potentially disrupt the power are generated from different data sources which are
supply. One of the important issues is power theft by distributed, that is, distribution and transmission data,
consumers. This can be done by hacking a smart distributed generation data, intelligence application
meter or changing its communication channel to related data, and so on. These data are distributed in
change the reported electricity usage. Additionally, different places and managed by different
data manipulation is also one of the most security companies/departments, which belong to different
concerns in the smart grid. To overcome these issues systems. Next, Data Storage Level: data collected
in they work. from the data sources contain a lot of meaningless
information, which does not need to occupy a large
Existing System:
amount of storage space. Therefore, the data in a data
Existing approach of Cloud-Based Demand Response center may be stored and processed. This level is
(CDR). It is fast response times in large scale suitable for the planning of smart grid with low real-
deployment. In this architecture, the master/slave time demands. Third level, Data Analysis Level:
demand response model is proposed, in which the Various analytical methods and tools have been
smart meters and the home EMS act as slaves, and the utilized to extract value in different application fields.
utility acts as the master. In such a scenario, the CDR It categorized by two divisions one is costumer-
leverages data-centric communication, oriented service another one is utility-company-
publisher/subscriber and topic-based group oriented service. Final level, Data Transmission
management, instead of IP-centric communication. Level: This level builds a bridge among the other
Two cloud-based demand response models are three levels and also enables the intra-communication
proposed as follows: (a) data-centric communication within each level.
and (b) topic-based group communication. Secure,
scalable and reliable demand response can be Merits:
achieved by using the CDR approach. However, the  Easy to Analysis the data.
demand-response model discussed in and with the  High performance.
implementation of private cloud for a small-sized  Reduce the time.
network. Some of the overhead problems are the  Accurate result.
implementation cost, and the selection of appropriate
strategy. Even for a small-sized network, all the Experimental Result:
features of cloud computing platform should be
supported in order to have reliable, and secure Maximum Throughput:
electricity distribution in a smart grid, and, thus, In this experiment throughput. LDP gives the 45% of
implementation cost of cloud applications is higher the throughput. FTP gives the 55% of the throughput.
for a small-sized network. So, there is a need for EBT gives the 63% of the throughput. GA gives the
developing such a demand-response model using 78% of the throughput.
cloud-based applications.

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 648
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470

Maximum Throughput
40 78
30 63
20 45

In this experiment cost. LDP gives the 73% of the cost. FTP gives the 60% of the cost. EBT gives the 50% of
the cost. GA gives the 35% of the cost.

Less Cost





High Efficiency:
In this experiment efficiency. LDP gives the 60% of the efficiency. FTP gives the 40% of the efficiency.
effi EBT
gives the 55% of the efficiency. GA gives the 83% of the efficiency.

High Efficiency





@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr

Apr 2018 Page: 649
International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
Conclusion: 4) Zhanyu Ma, JiyangXie, “The Role of Data
Analysis in the Development of Intelligent Energy
In this thesis, they provided an overview of Networks”, 2015.
integrating cloud computing in the smart grid
architecture, in order to have reliable, efficient and 5) Kyri Baker, Gabriela Hug, “Optimal Integration of
secure energy distribution. Different aspects of smart Intermittent Energy Sources Using Distributed
grid energy management, information management, Multi-Step Optimization”, 2015.
and security are discussed. In this article, thesis have 6) Werner Vogels, “Big Data: Challenges and
discussed the characteristics of smart grid big data and Opportunities”, 2014.
proposed a big data computing architecture for smart
grid analytics. Then the hierarchical big-data-aware 7) Khaled Shuaib., EzedinBarka, “Cognitive Radio
wireless communication architecture for smart grid for Smart Grid with Security Considerations”,
has been proposed, and enabling technologies are 2014.
introduced. 8) Khaled Shuaib., EzedinBarka, “Cognitive Radio
for Smart Grid with Security Considerations”,
Future Work: 2014.
In future, cloud computing even gives better security 9) Aneri M. Desai, Rutvij H. Jhaveri, “A Review on
capability than the conventional, pure IP-based Applications of Ambient Assisted Living”, 2013.
security mechanisms. Thus, the integration of cloud
10) Diksha M., Raghavendra T. S, “A Survey on Big
computing in smart grid is envisioned to be useful for
Data Energy Based on Smart Grid”, 2014.
evolving the smart grid architecture further in terms of
considerations such as monitoring cost, computing, 11) Suzhi Bi, Rui Zhang, Zhi Ding, “Wireless
data management, power management, and security. Communications in the Era of Big Data”, 2013.
12) IBM Corp. Software Group, “IBM Big Data
Overview forEnergy and Utilities”; http://www-
1) SoumenChakrabarti Ganesh Ramakrishnan,
“Data-based Research at IIT Bombay”, 2017. gy-utilities, 2012.
2) Rong Yu,ShengliXie, “Cognitive-Radio-Based 13) R. Yu et al., “Hybrid Spectrum Access in
Hierarchical Communications Infrastructure for Cognitive-Radio-Based Smart-Grid
Smart Grid”, 2017. Communications Systems,” IEEE Sys. J.,vol. 8,
no. 2, 2014, pp. 577–87.
3) Hervais Simo, “Big Data: Opportunities and
Privacy Challenges”, 2016.

@ IJTSRD | Available Online @ | Volume – 2 | Issue – 3 | Mar-Apr 2018 Page: 650