Enhanced Energy Efficient Air Indexing Technique for Data Broadcasting on

Wireless Channel
Ms. Prachi Goyal Dr. Vikas Goel Dr. Amit Kumar Gupta
Research Scholar, Associate Professor, Associate Professor,
Shri Venketeshwara University, Dept. of CS&E, Dept. of MCA,
Uttar Pradesh Ajay Kumar Garg Engg. College, KIET Group of Institutions,
[email protected] Ghaziabad, U.P. Ghaziabad, U.P.
[email protected] [email protected]

Abstract -Data Broadcasting on wireless channel is the By using key values, Hashing is used to index the
research area where many research has been done and are huge database. The key values are generated with the help
still going on. Earlier data is broadcasted on single channel of some type of formulas that is called hash function.
but now multiple channels are introduced for data A hash function maps data of arbitrary length to fixed
broadcasting. In this paper, we propose a hash based
length data. The value returned by a hash function is
indexing method for multiple wireless data broadcast
channels. Hashing indexing technique on multiple channels called hash value, hash codes, hash sums, checksums or
gives better result than on single channel. In this paper, we hashes. The same hash value should not be produced by a
are proposing an enhanced hashing technique that improves good hash function from two different input values. If the
the result of previous proposed broadcasting technique. In hash function produces the same key value then multiple
this paper, we are taking the advantage of the best features data items are mapped to same location, called a collision.
of hashing technique and proposing enhanced hashing A good hash function offers very low risk of collision [4].
technique for multiple channels. Our proposed technique
also minimizes collision in hashing technique. The proposed II. RELATED WORK
hashing technique aims to address the problem of data
broadcasting on wireless channels. The proposed algorithms
There are some hash based index schemes for
broadcast the data items to significantly reduce access
latency and parameter adjustment time to save time and wireless data broadcasting. Some of these schemes are
reduce power consumption of the mobile devices. based on single channel and some of multiple channels
[4,5].
Keywords: Data broadcasting, Hashing technique, Mhash
technique, collision, Multichannel Hashing techniques

I. INTRODUCTION

In order to disseminate information about a wireless

network Environment, data dissemination has become an
effective means. Example of data dissemination include
the traffic information system, meteorological information
Hashing Simple MHash Multichannel
and news distribution.In all these applications, client hashing MEHash
typically require overlapping data elements on the protocols hashing hashing hashing
channel.As a result, dissemination is an excellent solution (hashing A &
because it is more likely to broadcast a single piece of hashing B)
information(perhaps a large number) of clients.[1, 2].
Hashing is a way to that is used to sorting and Fig. 1. Hashing Technique
retrieving the information about the data base. This
information is associated with the key features and makes All data item are broadcast in buckets. Each
the use of individual character. Hashing is a good bucket consists of two parts: an index part and a given
technique for implementation in keyed table [10, 20]. data part. There is no necessary for another directory to
Hash table is a storage location in memory. It records the be broadcasted with the data. The hash parameter is
hashed value by the hashing algorithm. For creating a included in the data bucket. Each bucket has two parts, a
table certain no of bucket and location is needed [10, 20]. given part and a control part. The control part helps to
Hashing is a scheme of sorting and indexing minimize access time and listening time [11]. Control
when we think about the huge database. In hashing, the part for first N buckets (B) includes:
transformation of character string into a short fixed length - Hash function: h
value. That value shows the original string is done. Thus, - Shift: a pointer to a bucket
it is quite difficult to locate data item with the help of containing a hash key K in this
shorter key value than original key value. Hashing is also manner.
used in optimal search and retrievals because it increases h(k)= address(B)
speed, better ease of transfer, get better retrieval, optimize A data bucket is required to store hashing
searching of data, and reduce overhead [2]. parameters in simple hashing method. A separate index
bucket for the data bucket is not required in the that hash
 method. Each broadcast bucket consists has two parts: and index of that data in the proposed hashing technique.
control part and a given data part. The data part consists The proposed hashing technique’s data access algorithm
of real data item and control part into an index part to shows step by step procedure of construction of data item
guide client to the correct data bucket [12]. and a way to evaluate optimum access time and tuning
In the proposed hashing technique, efficient time. Data retrieving algorithm gives an idea to client how
access broadcast algorithms are proposed to index the a query is answered.
information using hash functions over multiple wireless
broadcast channels. To do this, two hash function, the A. Data structure of broadcasted data
primary hash function (PHF) and the Secondary Hash
Function (SHF) are used. The primary Hash Function All data item are broadcast in buckets. Bucket is
(PHF) determines the channel on which the data items is combination of two parts i.e. index part and data part.
broadcast. The secondary Hash Function (SHF) Index part contains exact location of data. In order to
determines the index of the data broadcasted item i.e. the access the necessary data, first client adjusts the broadcast
location of the data on the channel [21]. channel. It finds the broadcasting time related data. If the
In multiple wireless data broadcasting channel broadcast of the associated data item takes time, the client
system, a base station is continuously broadcasting a enter the active mode and becomes active when the
group of data items on the broadcasting channels. The necessary data is broadcast on the channel.
data set D = {d1; d2; · · ·; dt} is being broadcasted,
where t is the total number of broadcast data items. We Index_part:
have assumed that broadcast data items with their Bucket_ID: identification of this bucket.
primary key are stored in successively increasing order. C_ID: channel id i.e. identification of this channel.
Where pi is the access probability of di and , CL: channel length.
and P is the sum of all probability set of D [11]. N: no of channels.
H1(k): hash function for channel.
III. PROPOSED TECHNIQUE H2(k): hash function for data index.
H3(k,r): hash function for data collision.
This new proposed enhanced energy efficient m_Copy: maximum no of copies in all the data items.
hashing indexing technique for data broadcasting gives Data_part:
optimum average access time and tuning time. Multiple Data_ID: identification of data item.
wireless broadcast channels are used to broadcast the data
Step 2: for each channel access, each data item’s
probability is normalized. With the size of data item and
access probability of data, the number of copies of data
item is calculated. rfi is the number of replication
frequencies of di i.e. each data item can be calculated
with the following formulas:

…………………. (2)

Step 3: Second hash function H2(k) is used to calculate

the location of data items on different channels where
each data items appears once and placing a single bucket
for each data item on the channel to minimize the
conflicts.

...………. (3)
Fig. 2. Structure of bucket Where CL is channel length, a2= 1103515245 and
b2=12345
B. Broadcasting data construction algorithm
Step 4: If more than one data items is hashed in the
Step 1: H1(k) is first hash function that determine the same bucket then there is a collision. To resolve this
channel for broadcasting the data items. collision, a chaining method is used.

………… (1) Step 5: Third hash function is used to generate the

location of the copies of data.
Where N- no of channels, p- prime no > N, a & b -
integers randomly chosen from interval [0, p-1] where
a>0. ...…… (4)
Where r is the rth copy of data item & always r ≠ 1 and, E. Computation of average tuning time:

h = (2r - 2⌈log2r⌉ - 1)CL/2⌈log2r⌉ …………………….(5) Following are the steps to compute the average tuning
time of the proposed hashing technique:
Step 6: We allocate the data chaining data onto the
channel and each copy of data item is assigned to either a 1) Initially, the client checks the next slot on an
bucket or chain. We start with the most popular data item arbitrary channel to see if it contains the desired
of all the chained data items and assigned them to next data, download it. if it contains data. Otherwise,
closest empty bucket of the channel. using this information initially examined, the
client can calculate the target channel ID.
Step 7: In expanding process, define the first copy of 2) In the first analysis, if the client adapts to the
most popular data item into the buckets. correct target channel, it can directly calculate
the Hash values. This save a unit of time. If not,
Step 8: In last step,Assigning a global pointers creates a we must switch to the target channel for further
routing list in the index portion of each bucket.. investigation. This switching process can also
take a unit of time.
C. Data retrieving algorithm 3) Next, the client enters doze mode and adjusts
Hence in this paper, we show the algorithm to retrieve again until the e Hashed slot appears. check if
the data item for client. These steps are: the first location of the target data is included. If
included, the client starts downloading the
Step 1: tune anyone of that channel, read the first necessary data. Otherwise, it gets the routing list
available bucket and get related index information. for the location and a pointer to the target data
item. It will remain doze mode until aslot
Step 2: first hash function H1(k) find the C_ID channel appears in the channel.
id of requested data and then tune to this channel. 4) Finally, the data may be downloaded by client.
For probing, This process may take up to four
Step 3: In initial probing, we get index information i.e. units of time. The tuning time can be estimated
channel length then both H2(k) hash function and H3(k, r) by tprobe (probing time) and data downloading
hash function are calculated. time as following,

Step 4: Mobile client remains in doze mode until the

first hashed bucket of requested data is displayed. Then it
…….. (9)
reads a bucket. If this is the first bucket of the requested
IV. ANALYSIS
data, download the data. Otherwise it reads the routing
list and finds closest data bucket.
In this part, we perform analysis of the proposed
D. Computation of average access latency: algorithm for computing access time and tuning time.
Suppose dk is client requests of data on a random
In proposed system, m_Copy is the maximum broadcast channel and L is the length of the target
number of copies for all data items. So maximum there channel. Suppose channel switching time is always zero.
may be m_Copy number of Hashed values for dk on the
channel. The average length is M = L=m_Copy between Exp 1: Computation of average access time & average
two Hashed values. The physical position of dk is Posi(k) tuning time by varying data items
where 1 ≤ i ≤ ri. We also propose the displacement area In the first experiment, we have analyzed
Disp(k), which is the distance between each Hash value average access time and average tuning time by varying
and the nearest available physical slot. data items of size 10, 15 and 20. The no of channels are
kept constant with value 3. The values of average access
time and average tuning time are depicted in table 1.
………….…….. (6)
TABLE 1: Access time & Tuning time on behalf of no of
dispi(k) - displacement current position to the index data item
where it was assigned
No of No of Average access Average
posj(k) - current position of data item & hi(k)
data channels time(ms) tuning
items time(ms)
..…..………. (7)
10 3 2566 2.34
Total expectation of average access latency is:
15 3 704 3.01
………..……. (8)
 20 3 263 2.16

Figure 5. Graph 3 No of channels vs access time

Figure 3. Tuning time vs no of data items

Figure 6. Graph 4 No of channel vs tuning time

 Summary of exp 1: In table 1 above, there are

10, 15, and 20 data items. Access time and
tuning time are 3 channels. The figure 3 shows
Figure 4. Access time vs no of data items the variation in tuning time & figure 4 shows
the variation in access time and number of data
Exp 2: Computation of access time and tuning time by
items. In figure 3 and 4,increasing the number
varying no of channels
of data items reduces access time and tuning
In this experiment, the data item size is kept time.
cont. with 20 and by varying the no of channels, the
access time and tuning time are shown in the table 2.  Summary of exp 2: This figure 5 shows the
These results are also shown in the form of graph 6. access time. Here number of data item is same
but channel no is varying. This figure 6 shows
TABLE 2: Access time and tuning time on the behalf of no. of channels the tuning time. No of data items are same but
no of channels are same. In figure 5 and 6, the
No. of No. of Access Tuning access time and tuning time decrease.
data channels time(ms) time(ms)
items V. CONCLUSION
20 2 2606 2.79
20 4 2546 2.13 In this paper, we have proposed an enhanced
20 6 2557 2.62 energy efficient hashing technique for data broadcasting
20 8 2587 2.7 on multiple wireless channels. The proposed technique
20 10 2697 2.62 works for flexible data sizes and non-uniform
distribution of requested data items on multi-channel
broadcasting environment. In this paper, we have
provided detail algorithms for accessing the data in
enhanced energy efficient hashing technique. We have
also depicted theoretical and experimental results of the
proposed system performance. Adetailed analysis makes
it possible to conclude that the average access time and
the average turing time given by the proposed technique
are far superior to the technique proposed above.
For future work there can be another hashing scheme that
[11] Mahima singh, Choosing best strategies and Hash functions.
will give better result and will work to evenly distribute
Computer science and engineering dept Thaper University Patiyala,
the copies of data item on broadcasting channels. June, 2009.

REFERENCES [12] Muthuswamy Vijayalakshmi and Arputharaj Kannan: A Hashing

Scheme for Multi-channel Wireless Broadcast,Journal of Computing
and Information Technology - CIT 16, 2008, 3, 197–207
[1] Andre Seifert, Jenjou Hung, Flex Sched., A Flexible data Schedule
doi:10.2498/cit.1001112.
Generator of Multi-channel Broadcast System, University of Konstanz,
Technical Report, 2005.
[13] N. Shivshanker and S. Venkatassubramanian., Energy Efficient
Indexing for Information dissemination in Wireless Systems.
[2] Hameed, S. and Vaidya, N. H., Efficient algorithms for scheduling
ACM/Baltzer Journal of Mobile Networks and Applications (MONET),
data broadcast. ACM/Baltzer Journal of Wireless Network (WINET),
1(4), pp. 433-446, 1996.
pp. 183-193, 1999.
[14] Nitin Vaidya, Sohail Hameed., Scheduling Data Broadcast in
[3] Hu, Q. L., Lee, D. L., and Lee, W.C., Performance evaluation of a
Asymmetric Communication Environment. In the proceedings of the
wireless hierarchical data dissemination system. In Proceedings of the
20th International Conference on Data Engineering, pp. 171-182, 1999.
5th Annual ACM/IEEE International Conference on Mobile Computing
and Networking (MobCom’99), pp. 163-173, Seattle, WA, USA,
[15] Qinglong Hu, Wang-Chien Lee and Dik Lun Lee., A hybrid index
August 1999.
technique for power efficient databroadcast, volume 9 issue 2, March,
Pages 151 – 177, 2001.
[4] Jiaofei Zhong, Zheng Gao, Weili Wu, Weidong Chen and Li Wang:
Multi-channel Energy-efficient Hash Scheme Broadcasting,
[16] Qinglong Hu, Wang-Chien Lee and Dik Lun Lee., 2000. Power
proceedings of The 21st International Conference On Software
Conservative Multi-attribute queries on Data Broadcast. In proceedings
Engineering And Data Engineering SEDE 2012, Los Angeles,
of the 16th International Conference on Data Engineering (ICDE’2000),
California, USA, June 27-29, 2012.
pp. 157-166, San Diego, CA, USA, February, 2000.
[5]Jiaofei Zhong, WeiliWu, Xiaofeng Gao, Yan Shi, and Xiaodong
[17] Su, C. J., Tassiulas, L., and Tsotras, V. J., Broadcast scheduling
Yue: Evaluation and comparison of various indexing schemes in
for information distribution. ACM/Baltzer journal of Wireless
single-channel broadcast communication environment, Springer-Verlag
Networks (WINET), pp. 137-147, 1999.
London 2013.
[18] T.Lmielinski S.Vishwanathan and B.R. Badrinath: Data on Air:
[6] Joe Chun-Hung Yuen, Edward Chan, Kam-Yiu Lam and H.W.
organization and access, IEEE Transactions on Knowledge and data
Leung., Cache Invalidation Scheme for Mobile Computing System
engineering, Vol. 9, No. 3, May/June 1997.
with Real-time Data, SIGMOD Record, 18(3), pp. 392-403, 2006.
[19] T.Lmielinski S.Vishwanathan and B.R. Badrinath: Energy
[7] Konstantinos Stathatos, Nick Rou, and Sopoulos John S.Baras.,
efficient indexing on air ,june 1994.
Adaptive Data Broadcast in Hybrid Networks. In the proceedings of the
23rd International Conference on Very Large Data Bases, pp. 326-335,
[20] Vaidya, N. H. and Hameed, S., Scheduling data broadcast in
1997.
asymmetric communication environments. ACM/Baltzer journal of
Wireless Networks (WINET), pp. 171-182, 1999.
[8] Lee, W.C., Hu, Q.L., and Lee, D. L., A study of channel allocation
methods for data dissemination in Mobile Computing environment.
[21] W.G. Yee and S.B. Navathe, Efficient Data Access to Multi-
ACM/Baltzer Journal of Mobile Networks and Applicatiions
channel Broadcast Program. ACM Conference on Knowledge
(MONET), pp. 117-129, 1999.
Management, pp. 153-160, 2003.
[9] Lin, C. W. and Lee, D. L., 2000. Adapting data delivery in wireless
[22] Xu Yang and Atham Bougettya: Adaptive data access in
environment. In Proceedings of the 20th IEEE International Conference
broadcast-based Wireless Environments, IEEE Transaction on
on Distributed Computing Systems (ICDCS’2000), pp. 444-452,
knowledge and data engineering, Vol. 17, No. 3, March 2005.
Taipei, Taiwan, April, 2000.
[23] Yuxia Yao, Xueyan Tang, Ee-peng Lim, and Aixin Sun: An
[10] M.S. Chen, P.S. Yu, and K.L. Wu., 1997. Indexed sequential Data
energy efficient and access latency optimized indexing scheme for
Broadcasting in Wireless Mobile Computing. In proceedings of the 17th
wireless data broadcast, IEEE Transaction on knowledge and data
International Conference on Distributed Computing Systems
engineering, Vol. 18, No. 8, August 2006.
(ICDCS’97). pp. 124-131, Baltimore, MD, USA.