International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

RESEARCH ARTICLE OPEN ACCESS

Data Caching and Prefetching Using Collaborative Caching

and Prefetching Algorithm in Wireless ADHOC Networks
1
A. Bathsheba Parimala, 2 B. Jefferson, 3 K. Appasamy
Assistant professor, Department of BCA & M.SC(NT&IT) , St . Johns College , Palayamkottai.

Abstract:
Prefetching the information is a prominent strategy that enhances information availability in wired or
remote systems. However, in wireless ad hoc networks, improvement in access latency and cache hit ratio
may diminish because of the mobility and limited cache space of hosts. The proposed scheme uses a
Collaborative Caching with Data Prefetching (CCDP) which is a method that prefetches the image based on
association among data items. The scheme prefetches highly related data items and considers confidence of
association rules. Whenever a node issues a request, the cache request processing module first logs this
request into record and checks whether the desired data item is available in local cache of that node or in
any of the node in the cluster. If it is a cache hit, the cache manager still needs to validate the consistency of
the cached item with the copy at the original server. To validate the cached item, the cache manager checks
the validation of data item from its TTL value. If the data item is verified as being upto date, it is returned to
the node immediately. If it is a cache hit, but the value is obsolete, the cache manager sends an uplink
requests to the server and waits for the data broadcast. When the requested data item appears, the cache
manager returns it to the requester and retains a copy in the cache. The simulations are done using both
MATLAB and Network Simulator 2 tool and the performance metrics like throughput, packet loss and delay
will be compared with previous schemes.

Keywords — Prefetching, Caching, ad-hoc networks

I. INTRODUCTION local cache memory in which prefetched

data is held till it is required. As mobile
A wireless ad hoc network is a nodes in adhoc networks may have similar
continuously auto-configuring, tasks and share common interest,
infrastructure-less network of mobile nodes cooperative caching, which allows sharing
connected wirelessly. Each client in a and coordination of cached data among
WANETs (Wireless Adhoc Networks) is multiple nodes can be used to reduce the
free to move individually in any direction bandwidth and power consumption. Since
and will therefore change its links to other caching and prefetching are both well
nodes frequently. Each must forward traffic recognized for improving client perceived
unrelated to its own use and therefore be response time, the integration of both
a router. Cache prefetching is a technique strategies may be exploited to improve the
used by computer processors to boost system performance. In mobile ad hoc
execution performance by fetching networks cache misses are not isolated
instructions or data from their original events and a cache miss may followed by a
storage in slower memory to a faster local series of cache misses. Therefore data
memory before it is actually needed. Most mining association rules may be used to find
modern computer processors have fast and relationship among data items and hence

ISSN: 2395-1303 http://www.ijetjournal.org Page 35

 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

perform the prefetching. In the prefetching, protocol maintains high packet delivery ratio
access to remote data is anticipated and the and also it is designed to be a middleware
data is fetched before it is required. protocol sitting in between applications and
In this paper a Collaborative Caching network layer routing protocols. Nisar
with Data Prefetching” (CCDP) that Hundewale, Sunsook Jung, , Alex
prefetches the image data based on Zelikovsky [3] has introduced approach to
association among image data items. The constrain route request broadcast which is
proposed scheme prefetches highly related based on node caching. Intuition behind
image data items. To enhance the caching node caching is that the nodes involved in
performance in GCCIP, the generated recent data packet forwarding have more
caching rules are used to prefetch the image reliable information about its neighbors and
data item(s). Simulation is performed using have better locations (e.g., on the
both MATLAB and Network simulator2 for intersection of several data routs) than other
evaluating the performance of this algorithm nodes. Liangzhong Yin, Guohong Cao, and
under several circumstances. Based on Chita R. Das [4] has proposed a cooperative
caching rules, prefetching is performed and cache-based data access framework lets
confidence value along with other caching mobile nodes cache the data or the path to
parameters is used during prefetching. the data to reduce query delays and improve
The remainder of this paper is data accessibility. Dan Hirsch and Sanjay
organized as follows. In Section II we Madria [5] have proposed novel scheme that
briefly review the related studies on cache seeks to distribute the storage, bandwidth
replacement and prefetching in mobile ad- and energy burden through a resource
hoc networks and mobile environment. efficient adaptive caching scheme for
Section III gives description of proposed mobile ad-hoc networks. Our performance
system. Section IV describes the results. results show that our scheme reduces both
Section V concludes the paper. response time and bandwidth utilization by,
36%, through a reduction in hop count, as
II. BACKGROUND WORKS
Dr M.Madheswaran and Mrs. well as both a 79% increase in energy
K.Shanmugavadivu [1] have proposed a new efficiency and a 53% reduction in storage
group data caching scheme for improving utilization. Bhat [6] have proposed An
data access efficiency in MANETs called Efficient Cache Management using
Neighbor Group Data Caching. By Adaptive Buffer Mechanism in MANET. In
collaborating local resources of mobile a mobile environment, as a mobile node
nodes, data availability and access efficiency moves from one point of attachment to
is improved. Cooperative caching has two another during an ongoing application it is
problems namely Cache resolution and subjected to packet loss due to network and
cache management. However, it does not storage capacity [7]. Nayyar [8] have
investigate the integration of broadcasting proposed a Cross-Layer System for Cluster
and cooperative caching. David S. L. Wei, Based Data Access in MANET’S. A cross-
Chao-Chin Chou, Jay Kuo and Kshirasagar layer design approach is utilized to improve
Naik [2] has proposed anonymous the performance of combined cooperative
communication protocol called MANET caching and prefetching schemes [9-12]. For
Anonymous Peer-to-peer Communication future research there is a need to find out an
Protocol for P2P applications over wireless efficient prefetching technique which further
ad-hoc networks. Even under selective improves the data accessibility and reduce
attacks, this anonymous communication query delay to compliment the cooperative

ISSN: 2395-1303 http://www.ijetjournal.org Page 36

 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

caching scheme [7]. Waleed et al., [13] have Figure 1 Caching Framework in WANET
proposed A Survey of Web Caching and There are some problems occurred
Prefetching. Web caching and prefetching during Caching and Prefetching in
are the most popular techniques that play a WANETs. They are as follows. Degradation
key role in improving the Web performance of data access or latency in query processing
by keeping web objects that are likely to be is due to the following reasons with respect
visited in the near future closer to the client. to caching and prefetching problems:
Web caching can work independently or 1. Cache Access: Slowdown in
integrated with the web prefetching [12, 14]. accessing the cache memory due to network
congestion.
III. COLLABORATIVE CACHING 2. Memory constraint to store data:
AND DATA PREFETCHING Memory is fixed, memory to be utilized
Wireless Ad Hoc Networks efficiently to manage the data
(WANETs) consist of autonomous mobile communication
nodes, these nodes cooperate with each 3. Data overflow in prefetch
other to exchange data by multiple-hop memory: Data overflow from prefetch
communication. Although each node has memory due to prefetch memory limit.
limited transmitting range, some nodes 4. Cache Overflow: The contents not
behave as routers and forward data (e.g. stored in the cache memory due to cache
requests) from other nodes. Cache can be memory limit.
deployed either on each node or on some 5. Cache Data Recovery: Data loss
selected nodes to leverage cooperation. from the cache due to vulnerable attacks to
Figure 1 is a typical illustration of caching be recovered.
framework based on WANETs, in which 6. Data consistency in cache and
caching nodes cooperate to cache contents prefetch memory: Due to huge data traffic
and retrieve requested contents in a data consistency to be maintained between
multiple-hop fashion. A collaborative the cache and the prefetch memory
caching with data prefetching framework for 7. Data time stamp record: In
mobile nodes is proposed in which the prefetch memory data will be discarded if
buffer storage of each mobile node is the data is not utilized for certain period of
allocated for prefetching the data from other time.
nodes. Each cache node stores its most The collaborative caching and data
frequently accessed image data items in their prefetching algorithm prefetches the data
buffer. whenever possible to upgrade the
performance of the networks. First the
image is not processed in their format like
.jpeg or .bmp, .png directly in network
simulator. In network simulator tool the
images are processed like hexadecimal value
only. For processing image prefetching in
the simulation, the image must be converted
into hexadecimal value using MATLAB
R2012a. The hexadecimal value will be used
in network simulator tool to perform image
prefetching process. The flowchart of the

ISSN: 2395-1303 http://www.ijetjournal.org Page 37

 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

Start
collaborative caching with image
prefetching is shown in Figure 2.
Whenever a client issues a request Input Image
for data to control server, the control server
processing the request module and search Convert Image into Hexadecimal in MATLAB
the image data into its record and checks
whether the desired image data item is
available in local cache of mobile client or
in any of the mobile client in the cluster. If Mobile Ad-hoc Network Deployment & Parameters
Initialization in NS2
the searched image data is present means,
the cache manager still needs to validate the
consistency of the cached image with the Client and server Initialization
copy at the original server. To validate the
cached image data, the cache manager Upload the Hexadecimal files in Control Server
checks the validation of data item from its with file id and sequence number
time to live value. If the image data is
verified as being upto date, it is returned to
the mobile client immediately. If searched Mobile Client want to download a file from Control
image data is present but the value is server will send a request to control server
obsolete, the cache manager sends an uplink
request to the server and waits for the image
data broadcast. When the requested image
data appears, the cache manager returns it to
the requester and retains a copy in the cache. Control Server Checks
In the case that a searched image data miss the file present in their
occurs, the client cache manager checks the Cache?
caching rule depository to derive the
prefetching rules corresponding to the
requested item. If this request triggers some
N
prefetching rules, the ids of the item implied
by these prefetching rules will also be Control Server will forward the request to its other
piggybacked to the server along with id of mobile clients
missed cache item
File discovery process on other mobile client cache
using GCPT

Y
Control Server Receives
requested File form other
mobile client?

Figure 2 Flowchart of image prefetching

ISSN: 2395-1303 http://www.ijetjournal.org Page 38

 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

Algorithm: mobile nodes forming an ad hoc network,

Collaborative Caching with Data moving about over a 5m/s. In our
Prefetching experiment, we have set the communication
Step 1: Cache processing the request range of mobile node to 300m. The media
from mobile node. Cache will check the access control layer we used in simulation is
requested data is available in local or within 802.11 MAC protocol. We generated all the
cluster. movement scenarios using setdest command
Step 2: If the requested data is in NS2. First the images are converted into
available then check the confidence hexadecimal values using MATLAB,
(originality of the data such as validity) and because in NS2 the images are processed
send it to requested mobile node only in hexadecimal value format. After the
Else if the requested data is available conversion of image to hexadecimal value
but the validity is not satisfying the that will be used to prefetching the data to
threshold level means cache manager sends mobile node. First if a mobile node requests
the uplink request to server. After getting the image data to control server. After receiving
data from server it will return the data to request the control server first check its
requested mobile node and keep that copy of cache and if the requested image data is not
data present in the cache means it will forward
Else the request to nearby mobile nodes. After
Step 3: the data is missing in cache getting the image data it will transfer the
manager checks the caching rule to derive image data to the requested mobile node and
prefetching rules using association rules for keeps the copy of that image data.
requested item. Assume that the client has a Table 1 : Simulation Parameters
cache miss to item in. The proposed scheme Parameter Value
finds out the rules whose antecedent is in.
Channel Type Wireless
Consequent of each prefetch rule along with
the confidence value in a linked list of such Routing Protocol AODV
tuple is stored. The tuples of this list are then Queue Length 50 Packets
sorted in descending order of their
confidence values. If cache miss happens, Number of Nodes in 50,75.100
the prefetch set for the cached miss item is Topography
generated. Node Placement Random
Step4: Instead of only requesting the Simulation End Time 50 sec
cache miss item, the client also requests the MAC Protocol IEEE 802.11
items indicated in the prefetch set. When a
cooperating node or server receives the Packet Size 512 bytes
request, it transfers the requested items over Traffic Type CBR
the wireless channel. The client downloads
the items and stores them in its cache. By Path Loss Model Free Space
prefetching the items, the client can save Energy 500J
future requests and reduce the query latency.
Transmission Power 0.75
IV. RESULTS
The simulations are done using built- Receiving Power 0.5
in random generator in Network Simulator 2 Mobility 5m/s
(version NS-2.34). The protocol evaluations
are based on the simulation of 50 wireless Frequency 912mhz

ISSN: 2395-1303 http://www.ijetjournal.org Page 39

 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

Communication 3300m prefetching increases gradually and it will

Range decrease because of the searching of image
data. From this the throughput will increases
than existing system. Delay is the amount of
time taken for image data transfer between
the client and server node. The proposed
image prefetching technique requires less
time to transfer the data than previous
cooperative caching scheme.
scheme

(a)

(c)
Figure 3 (c) Overhead
Figure 3 (c) shows that snapshot of
overhead. Overhead means excess
consumption of resources that are required
to perform a specific task. The proposed
image prefetching has less overhead than
existing.
V. CONCLUSION
In WANETs Caching and
Prefetching
ching technique can be used to
upgrade the system performance in wireless
ad hoc networks. However caching and
(b) image prefetching also consumes a large
Figure 3 (a) Throughput (b) Delay amount of system resources such as
computation power and energy. Thus it is
Figure 3(a) and (b) shows that the very important to only prefetch the right
screenshot of throughput and delay. data. In this paper, collaborative caching and
Throughput is defined as the amount of the data prefetching algorithm is proposed
bits received at receiver side after the data which upgrades the performance
transmission. It will be measured in bits per improvement
ovement due to caching in wireless ad
second. The throughput of the proposed hoc networks. Thehe proposed techniques will

ISSN: 2395-1303
1303 http://www.ijetjournal.org Page 40
 International Journal of Engineering and Techniques - Volume 3 Issue 6, Nov - Dec 2017

lead to further research work in several areas 9. Artail H, Safa H, Mershad K, Abou-Atme
related to cache management and security Z, Sulieman N (2008) COACS: A
against vulnerable attacks in mobile cooperative and adaptive caching system for
computing environments. MANETs IEEE Trans. Mobile Comput 7:
VI. REFERENCES 961-977.
[1] Mrs. K.Shanmugavadivu and Dr 10. Tian J and Denko MK (2007) Exploiting
M.Madheswaran, “Caching Technique for clustering and cross layer design approaches
Improving Data Retrieval Performance in for data caching in MANETs. Wireless and
Mobile Ad Hoc Networks”, International Mobile Computing, Networking and
Journal of Computer Science and Communications 52.
Information Technologies, Vol. 1 (4), 2010, 11. Marquez J, Domenech J, Gil JA, Pont A
249-255 (2008) A web caching and prefetching
[2] Chao-Chin Chou, David S. L. Wei, Jay simulator. Software, Telecommunications
Kuo and Kshirasagar Naik “An Efficient and Computer Networks 346- 350.
Anonymous Communication 12. Waleed A, Shamsuddin SM, Ismail AS
Protocol for Peer-to-Peer Applications over (2011) A Survey of Web Caching and
Mobile Ad-hoc Networks”, IEEE Prefetching. International Journal of Soft
JOURNAL ON SELECTED AREAS IN Computing and Its Applications 3: 18- 44.
COMMUNICATIONS, VOL. 25, NO. 1, 13. Angrisani L, Bertocco M, Fortin D, Sona
JANUARY 2007 A (2008) Experimental study of coexistence
[3] Sunsook Jung, Nisar Hundewale, Alex issues between IEEE 802.11b and IEEE
Zelikovsky, “Node Caching Enhancement of 802.15.4 wireless networks. IEEE Trans
Reactive Ad Hoc Routing Protocols”, 2009 Instrum Meas 57: 1514-1523.
[4] Guohong Cao, Liangzhong Yin and 14. Chungsoo L and Gregory TB (2008)
Chita R. Das, “Cooperative Cache-Based Exploiting producer patterns and L2 cache
Data Access in Ad Hoc Networks”, for timely dependence-based prefetching.
Published by the IEEE Computer Society, IEEE 1063-6404.
2004
[5] Dan Hirsch and Sanjay Madria “A
Resource-Efficient Adaptive Caching
Scheme for Mobile Ad-Hoc Networks”,
IEEE International Symposium on Reliable
Distributed Systems, 2009
6. Kassem F and Hassan A (2013)
Distributed Cache Invalidation Method For
Maintaining Cache Consistency In Wireless
Mobile Networks. IEEE Transactions on
Mobile Computing 12: 1536-1233.
7. Nayyar A (2012) Cross-Layer System for
Cluster Based Data Access in MANET’S.
International Journal of Computer Science
& Informatics 2: 15-20.
8. Denko MK and Tian J (2008) Cross-layer
design for cooperative caching in mobile ad
hoc networks. Consumer Communications
and Networking Conf 375-380.

ISSN: 2395-1303 http://www.ijetjournal.org Page 41