Transcript Clustered Content Replication for Hierarchical Content Delivery

An Effective
Communication
Framework for Data
Management using Query
Caching in Mobile Location
Based Services (LBS)
R. Gobi
Faculty in Computer Applications, National Institute of
Technology, Tiruchirappalli, India
E. Kirubakaran
Bharat Heavy Electricals Limited, Tiruchirappalli, India
2016 International Conference on Data Mining and Advanced
Computing (SAPIENCE)
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Introduction
Location-based services (LBS) provide service providers the
means to deliver personalized services to its subscribers.
LBS reflect the convergence of multiple technologies: Internet,
Geographic Information System and Mobile devices.
Introduction
LBS are defined as service provided to a subscriber based on
the current geographic location of the Mobile Station.
The services include a wide range right from information to
entertainment.
Examples of such applications include emergency services,
vehicle navigation systems, tourist tour planning etc...
Introduction
Managing large volume of data and increase in data access in
telecommunication opened an option to define data models
and query management techniques for location data.
Caching and Replication techniques store & make sure the
availability of data all the time and which will reduce the query
cost and network bandwidth.
Due to the increasing complexity of mobile information
systems, Database Management System techniques on mobile
devices plays vital role.
Introduction
The process of LBS mainly comprises of getting the location of
the user and to provide a service to the user based on utilizing
the information.
Apart from positioning through GPS based or network based
mechanisms, data management also plays a vital role in LBS.
The challenge of data access for mobile object is that the same
query may need to be answered with entirely different results.
Introduction
As the number of location based mobile services increases
rapidly every day, traditional information access techniques
may not suit LBS due to its highly dynamic nature with respect
to user movement.
From the service provider's perception it is a double pay on
location and data management.
Introduction
Increase in number of mobile users, data usage, and frequent
access of same information specific to location cost
performance issues, network unavailability.
Providing location based service on dynamic nature when user
moves from one location to another are challenging.
Developing proper architecture would be solution to make LBS
fruitful.
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
ComFrame: a Framework for Data
Management
The framework was designed with several principles to
replicate the existing and predictable future condition of the
wireless communication technologies.
The communication framework design for data management in
mobile LBS consists of Mobile Device, Communication Local
Server and Communication Central Server.
This work is based on the conceptual research to construct the
framework for the data management using query caching and
the query caching concept is stored in local communication
server for retrieval of data by the mobile device.
Mobile Device
Mobile Device is a user who acts as a client to support for
information transfer and to support user interface services.
The client can change its position at any time, so it is the
challenging issue to face both the dynamic data and queries.
It is composed of LBS application, Navigation/Browsing and Log
manager.
Communication Local Server
Communication Local Server helps in connecting the mobile
device to the Communication Central Server.
All local servers must be connected together to provide
efficient data management.
Basically, the Communication Local Server is viewed as a single
cell.
Here, each cell has its own Communication Local Server in the
wireless communication.
Communication Central Server
Communication Central Server is used to handle the data such
as the Banking Database.
Usually, the request data which is not available in the
Communication Local Server will be forwarded to the
Communication Central Server.
The Communication Central Server has the response of
handling many local servers.
ComFrame: a Framework for Data
Management
The main scope of the study is to enhance the data availability
in the local server using query cache model and also improves
the data retrieval in the mobile computing.
Every time when we access the data, the query is forwarded to
the remote server and the server response is low.
So, the concept of local server is used in between the mobile
device and the central server.
ComFrame: a Framework for Data
Management
The cache manager and Query cache are implemented by using
local server.
The Cache manager will manage all the data and the Query
Cache will list and update the number of caches in the system.
Based on this framework, a detailed analysis can be made to
improve response time and reduce service cost with respect to
dynamic access.
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Query Caching in Mobile Location
Based Services
In mobile computing, the data handling approach is based on
both data and queries.
In general, the user query is based on object, range and map.
Caching is a key technique to improve data retrieval
performance in mobile environments.
The Mobile Clients connects to the base station and the Base
Station will communicate with the Communication Server
where the cache techniques are stored.
Query Caching in Mobile Location
Based Services
The Location information and the Geographical information are
stored in the Communication Server.
The fundamental concept behind this is to leverage the cached
results from prior spatial queries for answering future queries
at the communication central server.
The user can access the data by using their mobile device and
therefore the data can be supplied to the user by using the
communication local server or by using Communication central
Server.
Query Caching in Mobile Location
Based Services
There are several steps involved to deliver a data to the user
from server side:
• Prepare to receive or respond to the user by server
• Replace the old data
• Update the frequently access data
• Add or Register the new data
Query Caching in Mobile Location
Based Services
The proposed work is a cache management model for
enhancing the availability of the location based information
with respect to server side caching.
Based on this model one can get access from cache if the query
is available with local server model in a reduced response time
compared with traditional information access time.
Query Caching in Mobile Location
Based Services
• It provides solution to the instability of mobile network in a
distributed mobile system. The cache management model is a
customizable approach, meaning that host specific and
application specific constraints can be enforced like
emergency applications
• Cache management maintains the current location of MD
there by solving the problem of identifying location of MD. It
acts as a server for handling data dissemination to provide
mobile data access, in both server-push and client-pull
models.
Query Caching in Mobile Location
Based Services
• The cache management can also cache Mobile Device
specific data and reduce the response times for many client
queries. It also supports disconnected operations of the MD
by buffering client requests or using the cached data to
handle them
• It provides optimal utilization of wireless bandwidth, as the
cache management knows the current network connectivity
and other constraints of its corresponding host.
The workflow of Q-caching
Q-Cache Process flow
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Implementation of Prototype
The proposed architecture is designed based on two ways.
The client side designed using Sun Java J2ME wireless Toolkit
2.5 simulator.
Java RMI is used for simulating distributed environment for
location based web application done by Tomcat 5.0 and Oracle
GIS Database is used.
Several technologies like RMI, web component and J2ME are
used by the cache management model.
Implementation of Prototype
The client side is implemented by using the Android emulator
and the server side is implemented by using the Java language.
The central server has all the information that is to be supplied
to the user.
The local server gets the user data and stores in the local
database.
The Query Directory maintains the list of query raised by the
user and Query Manager controls what information is to be
given to user based on query.
Transmission of messages
between MDs
Implementation of Prototype
The major factor that analysis the performance is response time (The
time taken for moving an object from one MSS to another MSS).
The response time for data access with cache management model is
less when compared with the customized cache management model.
The migration starting time and the migrated time of the object is
taken for analysis during implementation.
The study shows that the response time is quicker for the customized
cache management model than the cache management model.
Hence, the mobile customers are able to get the required
information in a shorter time.
Response time of cache in
seconds
Implementation of Prototype
The table shows that the data access with cache management
need only 39.3 seconds but the data access without cache
management needs 76.7 seconds to perform the operation.
The figure shows the graphical representation of the
comparison of the data access with cache and data access
without cache.
Performance analysis
Implementation of Prototype
The table shows that the proposed Cache model with Cache hit
rate gives the improved performances when it is compared
with traditional access with Cache miss rate.
The client requests are given in the column 1.
The Column 2 defines the cache model with cache hit rate.
The Column 3 defines the traditional access with cache miss
rate.
Comparison of Cache model with
traditional information access
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Types of cost analysis
Cost Analysis
The main focus of cost analysis is to reduce the operation and
reporting of each issue one by one.
It is used to evaluate the desirability of a particular model.
It helps to forecast whether the benefits of a framework or a
model be more important than its cost.
This study involves in costs and related economic implications
that comprise a generic Cost-benefit Analysis in ComFrame.
Cost Analysis
The Cost-Based Analysis attempts to figure out the pros and
cons of framework which includes the following:
• Special outcome on mobile users
• Effects on non-mobile users
• Externality causes on clients and servers
• Option value or any social issues
Scenario 1: BCA Measures
Several variations on the basic benefit-cost rule can be used to
compare the benefits and costs of investments, projects, or
decisions.
After collecting all the factors that relates to analysis, the data's
are grouped together to calculate accurate Cost-Benefit
analysis.
Scenario 1: BCA Measures
Scenario 1: BCA Measures
The net present value (NPV) is the current value of all project
net benefits.
Net benefits are simply the sum of benefits minus costs.
The sum is discounted at the discount rate.
Based on the above formulae, the achieved NPV = 16,220.09.
Using this method, if the project has a NPV greater than zero
then it appears to be a good candidate for implementation.
Scenario 1: BCA Measures
The benefit-cost ratio (BCR) is calculated as the NPV of benefits
divided by the NPV of costs:
Here, Bt is the benefit in time t and Ct is the cost in time t.
If the BCR exceeds one, then the project might be a good
candidate for acceptance.
The discount rate applied here is 10%.
Scenario 1: BCA Measures
The internal rate of return (IRR) is the maximum interest that
could be paid for the project resources, leaving enough money
to cover investment and operating costs, which would still allow
the investor to break even.
In other words, the IRR is the discount rate for which the
present value of total benefits equals the present value of total
costs.
In general, the IRR should be greater than the discount rate for
a project to be accepted.
Based on the above table the IRR is 46%.
Scenario 2-Calculating the Discount
Rate for Any Particular Service Used
The discount rate calculation is used to implement any model
or any framework is carried out by compiling the discounted
stream of costs (or benefits) over time.
Here, p is the present value, f is the future cost (or benefits) at
year k and r is the annual discount rate given to the customer
when they try to access a service based on the ComFrame for
mobile Location Based Services.
The main advantage of using this formulae is to find out the
discount rate given to customers for any particular service used
like application cost, system utilization charge, etc.,
Scenario 3-Calculating a Getback
Time for ComFrame
Considering the cost and benefit list into account, the amount
of time is taken which will recover the projected costs.
Example: It is needed to implement a local server which will
interact with mobile user at a cost.
Cost Analysis
The Cost-Benefit based analysis is carried out to support for
clients and also for service providers and site administrators.
The company can use the model to predict their benefits
before they implement the framework like ComFrame.
The discount rate is applied and it supports for the customer to
get benefits.
Cost Analysis
The Cost-Benefit based analysis is the de facto standard for
representing financial possibility to implement any model.
It also does the comparison and to perform selection among
the investment in that particular field.
The main necessity of the Cost-Benefit based analysis is to
perform some analysis and to present some data related to risk
management, costs and benefits of a given framework like
ComFrame so that it can be compared to any other investment
opportunities.
Outline
I.
Introduction
II. ComFrame: a Framework for Data Management
III. Query Caching in Mobile Location Based Services
IV. Implementation of Prototype
V. Cost Analysis
VI. Conclusion
Conclusion
A framework is designed for delivering of data using query
caching from central server to local server.
The developed framework has the several success factors.
As per this research, query caching using local server results in
improvements of quality, productivity, performance, cost
benefit and interoperability.