IJRIT International Journal of Research in Information Technology, Volume 1, Issue 1, January 2013, Pg. 87-96

International Journal of Research in Information Technology (IJRIT) www.ijrit.com

ISSN 2001-5569

Entropy Based QoS Routing Algorithm in MANET Using Hierarchical Clustering Himanshu Gupta1, Pratibha Chauhan2, Suryambika Singh3, Pankaj Kumar Sharma4 1

Department of Information Technology, ABES Engineering College, Ghaziabad, Uttar Pradesh, India Department of Information Technology, ABES Engineering College, Ghaziabad, Uttar Pradesh, India 3 Department of Information Technology, ABES Engineering College, Ghaziabad, Uttar Pradesh, India 4 Sr. Associate Professor, Department of Information Technology, ABES Engineering College, Ghaziabad, Uttar Pradesh, India 1 [email protected], 2 [email protected], 3 [email protected], 4 [email protected] 2

Abstract A Mobile Ad Hoc Network (MANET) is a dynamic wireless network that can be formed without the need of any pre-existing infrastructure in which each node can act as a router. Because mobile nodes have limited battery power, it is therefore very important to use energy in MANET efficiently. Due to bandwidth constraint and dynamic topology of mobile ad hoc networks, multipath supported routing is very important research issue. This paper proposes an ad hoc on demand distance vector routing algorithm on the basis of entropy. Two entropy metrics are used for this purpose, relative motion entropy and energy entropy. Hierarchical clustering algorithm is used to optimize the set of possible routes for communication. The proposed algorithm assigns the construction of multiple paths to the destination node, resulting in the improved performance of packet delivery, average end-to-end delay and control packets ratio incurred at intermediate nodes.

Keywords: AODV, Energy Entropy, Hierarchical Clustering, Euclidean Distance, Multipath routing.

1. Introduction A Mobile Ad hoc NETwork (MANET) is an autonomous system of mobile nodes connected by wireless links. There is no static infrastructure such as base station as it is in cell mobile communication. In MANET, if two nodes are not within radio range, all message communication between them must pass through one or more intermediate nodes. All the nodes are free to move around randomly, which changes the dynamic changes of the network topology [1-6]. Routing protocols also maintain connectivity when linking on these paths break due to effects such as node movement, battery drainage, radio propagation, and wireless interference. The design of efficient and reliable routing protocols in such a network is a challenging issue [1-6]. In MANETs, each mobile node has limited energy resource (battery), and each node operates in an unattended manner. In order to maximize the lifetime of MANETs, traffic should be sent via routes that can avoid nodes with low power while minimizing the total transmission power [7-9]. Multipath routing protocols that improve energy efficiency of a network fall into this category. A mobile node is usually equipped with a battery of limited capacity. An energy efficient multipath on-demand routing (MDR) has been introduced in Dulman et al. [7]. In [8], Liang and Pratibha Chauhan, IJRIT

87

Ren have proposed an energy and mobility aware geographical multipath routing for wireless sensor networks. Wang et al [9] propose an energy efficient and collision aware (EECA) node-disjoint multipath routing algorithm. This power is then used both node to transmit a packet over the link, and as the link weight in a minimum- eight path search algorithm. In this way, transmit power can be tuned in order to build the desired connectivity diagram. AODVM (AODV Multipath) [10] is an extension to AODV [5]. In the AODVM protocol, a destination node selects paths that pass through more reliable nodes. The objective of energy conservation is to help extend the lifetime of a network with energy-limited nodes. It is desirable to balance energy dissipation of these nodes so that they would not run out of energy early in some area, resulting in a disconnection of the entire network. Entropy [11-14] presents the uncertainty and a measure of the disorder in a system. There are some common characteristics among self-organization, entropy, and the location uncertainty in MANETs. Robinson et al [13] proposes a distributed clustering algorithm which takes into consideration the local information available to all the nodes. This local information is measured in terms of entropy. The energy entropy was defined to optimize cluster head power use when electing cluster members. Kawahigashi [14] proposes a new measure, urgency index and urgency entropy, to evaluate the effectiveness of coding in urgent communication and discuss its usefulness for designs of bandwidth constrained, control-based MANETs from a viewpoint of utility functions and elasticity. In this paper, the authors propose an energy Entropy based QoS Routing algorithm in MANET based on Hierarchical Clustering. Our motivation is to provide the improvement of on-demand multipath routing method over on-demand unipath routing in terms of packet delivery ratio, average end-to-end delay through simulation using Network Simulator (NS2).

2. Hierarchical Clustering 2.1 Clustering Method In data mining, hierarchical clustering is a method of cluster analysis which seeks to build a hierarchy of clusters. Strategies for hierarchical clustering generally fall into two types: 2.1.1 Agglomerative: This is a "bottom up" approach: each observation starts in its own cluster, and pairs of clusters are merged as one moves up the hierarchy. 2.1.2 Divisive: This is a "top down" approach: all observations start in one cluster, and splits are performed recursively as one moves down the hierarchy. In general, the merges and splits are determined in a greedy manner. The results of hierarchical clustering are usually presented in a dendrogram. In the general case, the complexity of agglomerative clustering is O(n3), which makes them too slow for large data sets. Divisive clustering with an exhaustive search is O(2n), which is even worse. However, for some special cases, optimal efficient agglomerative methods (of complexity O(n2) ) are known: SLINK for single-linkage and CLINK for complete-linkage clustering. By taking into reference the complexity of both types of clustering schemes we opted for agglomerative clustering for our research. Hierarchical clustering has the distinct advantage that any valid measure of distance can be used. In fact, the observations themselves are not required: all that is used is a matrix of distances. In order to decide which clusters should be combined (for agglomerative), or where a cluster should be split (for divisive), a measure of dissimilarity between sets of observations is required. In most methods of hierarchical clustering, this is achieved by use of an appropriate metric (a measure of distance between pairs of observations), and a linkage criterion which specifies the dissimilarity of sets as a function of the pairwise distances of observations in the sets.

2.2 Metric The choice of an appropriate metric will influence the shape of the clusters, as some elements may be close to one another according to one distance and farther away according to another.

Pratibha Chauhan, IJRIT

88

Metric used for hierarchical clustering in this proposal is: 2.2.1 Euclidean distance: the Euclidean distance or Euclidean metric is the "ordinary" distance between two points that one would measure with a ruler, and is given by the Pythagorean formula. By using this formula as distance, Euclidean space (or even any inner product space) becomes a metric space. The associated norm is called the Euclidean norm. Older literature refers to the metric as Pythagorean metric. The Euclidean distance between point’s p and q is the length of the line segment connecting them (

)

Euclidean distance can be calculated between two nodes having link between them using the following formula

(1)

3. Energy Entropy Information theory developed by Shannon [11] is a fundamental field in mathematical sciences to deal with transmission of information through communication systems. In information theory, the standard and basic quantity to deal with information is entropy. There are some common characteristics among self-organization, entropy, and the location uncertainty in MANETs [12, 13]. Though, the remaining battery is easy to measure, the rate at which it will deplete is still uncertain. It is known that more power is required to communicate to a larger distance. Thus, transmit power depends on the relative distance between the transmitter and the receiver nodes. Note, that the maximum range (Rmax) attainable by a node is limited by the maximum allowable transmit power, Rmax. Now calculate the uncertainty in the relative distance between a transmitter and a receiver. Since, the nodes are randomly scattered, the receiver will lie anywhere in the circle with radius Rmax with equal probability, with the transmitter node being at the center of the circle. In polar co-ordinates, the radial distance is assumed to be uniformly distributed between o and Rmax, and the angle uniformly distributed direction between 0 and 2∏. The position of the receiver is characterized by fR (r) and fθ (θ), denoting respectively the distance probability density function (pdf) and the directional pdf The two pdfs are defined as follows[13]:

, 0  r  R  f r   0,  

 

, 0    2! fӨ    0,  

Pratibha Chauhan, IJRIT

(2)

(3)

89

The joint pdf is given by[13]

, )*+*  "#Ө ,   $ 0,  %

#&'(

)*%*

(4)

Given this pdf of the distance of the receiver from the transmitter, the transmission power distribution, and hence the energy dissipation can be obtained. For the joint pdf of distance as fRӨ (r, θ), calculate the pdf for the transmission power. It assume that the attenuation in the signal strength is inversely proportional to the square of the distance i.e., if Pt and Pr are the transmit and receive powers respectively, Pr=Pt*d-α

(5)

Where α is the path loss exponent and usually lie between 2 and 6. A generic expression to calculate the energy required to transmit a packet p is: Pt=i*v*tp Joules, where i is the current consumption, v is the voltage used, and tp is the time required to transmit the packet. It is supposed that all the mobile devices are equipped with the IEEE 802.11b network interface cards (NICs). The energy consumption values were obtained by comparing commercial products with the experimental data.

" - /0 , )*+*  ", -.    #Ө . 0,  )*%*

Therefore, the pdf for the transmission power, fp (pt), is given by[13]

(6)

Since according to the authors’ assumption transmission power is directly proportional to the energy consumed, the transmission power pdf to calculate the energy entropy.

Shannon’s entropy for a random variable with Y with pdf fy (y) is[13]

12 3  4/; "5 678"5 696 :;

(7)

Thus the energy entropy is given by[13]

1",   4) &'("#Ө -./0 78"#Ө -./0 9-. #

(8)

We used clustering to remove non-participating or less efficient nodes out of the data transmission. After applying clustering algorithm the data is sent to the node with minimum entropy. Entropy is obtained within the route reply Pratibha Chauhan, IJRIT

90

(RREP) message. Minimum entropy accounts for the minimum uncertainty in the depletion rate of the battery power of the node which enhances the chance of successful data transmission.

4. Simulation Experiments 4.1. Simulation model To conduct the simulation studies, we have used randomly generated networks on which the algorithms were executed. To effectively evaluate performance of proposed algorithm, we compare it with other famous multicast routing protocols AODVM for cost to control information, average link-connect time, the success rate to find the path and the feature of data transmission. Table I lists the simulation parameters which are used as default values unless otherwise specified.

TABLE-1 SIMULATION PARAMETERS Channel type

Wireless channel

Radio propagation model

Two ray ground

Network interface type

wirelessPhy

MAC Type

802.11

Interface queue type

Drop Trail/PriQueue

Link layer type

LL

Antenna model

Omni antenna

Max packet in ifq

50

Number of mobile nodes

100

Routing protocol

AODV

X dimension of topography

500

Y dimension of topography

400

Time of simulation end

150

Pratibha Chauhan, IJRIT

91

We will compare the performance of AODV protocol under the same movement models. We evaluate the performance according to the following metrics: 4.1.1

Average end to end delay of data packets: It represents the average value of the time that the received data packets to reach the destination from their origin. Figure1 shows the average end to end delay of AODV.

Figure 1: End-to-End delay of AODV

4.1.2

Packet delivery ratio: The packet delivery ratio is the ratio of the correctly delivered data packets, and is obtained as follows.

Packet delivery ratio =

<=.=? @ABCD.E FDGHID%DF <=.=? @ABCD.E EDJ.

The number of delivered data packets is the summation of total numbers of delivered data packets received by each node. The number of sent data packets is the summation of total number of sent data packets of each node. The packet delivery ratio shows the transmission efficiency of the network with the given protocol

4.1.3

Jitter: Jitter is often used as a measure of the variability over time of the packet latency across a network. A network with constant latency has no variation (or jitter). Packet jitter is expressed as an average of the deviation from the network mean latency. Figure2 shows the jitter of AGT packets of AODV protocol.

Pratibha Chauhan, IJRIT

92

Figure 2: jitter of AGT packets of AODV

4.1.4

Throughput: Throughput or network throughput is the average rate of successful message delivery over a communication channel. This data may be delivered over a physical or logical link, or pass through a certain network node. The throughput is usually measured in bits per second (bit/s or bps), and sometimes in data packets per second or data packets per time slot.

4.2 Simulation Results The results of the simulation are positive with respect to performance. We use the NS-2 simulator to evaluate the EAODV (Entropy based AODV) protocol. NS-2 is a discrete event simulator targeted at networking research. NS-2 provides substantial support for simulation of TCP, routing, and multicast protocols over wired and wireless networks. Overhead comparison with respect to the number of nodes is shown in figure 3. We can observe that the overhead is more in AODV in comparison of EAODV as in AODV large number of control messages are sent for route discovery and route maintenance.

Overhead comparison overhead

4000 3000 2000 EAODV 1000 AODV

0 200

400

600

800

network size(nodes)

Figure 3: overhead comparision between AODV and EAODV Pratibha Chauhan, IJRIT

93

The packet delivery ratios as a function of mobility speed and throughput are shown in Fig. 4. We can observe that as speed increases because of links break the packet delivery ratios decrease in AODV and EAODV. As the nodes maximum speed increase, a packet delivery rate of methods decreases. This because, in higher speeds, more frequent link breakage may occur and therefore a packet loss fraction is increased. However, at high speed like 20 m/s of EAODV does much better while the performance of AODV. It is mainly due to the fact that not only the higher node mobility induces more frequent link breakage but also the larger number of connections increases the probability of link breakage.

packet delivery ratio packet delivery ratio ()%

150 100 EAODV

50

AODV 0 0

4

8

12

16

20

node's mobility speed (m/s)

Figure 4: packet delivery ratio

Fig. 5 shows the average end-to-end packet delay of the three protocols under evaluation. EAODV represents shorter end-to-end delay than AODV by up to 20 percent. The increase of movement speed induces more frequent topology change and therefore the probability of broken links grows. Broken links may cause additional route recovery process and route discovery process. Because of this reason, the average end-to-end delay of packet increases as node speed increases. From the Fig. 5 we can see that when the node’s mobility speed increases, EAODV algorithm has the lowest average end-to-end delay of all two methods. Since as speed increases, more route requests are needed thus, delay increase with speed in all methods. This is mainly due to more robust routes and less route discoveries, which minimize the potential possibility of link breakage.

average end to end delay average end to end delay (s)

0.3 0.2 0.1

EAODV

0

AODV 0

4

8

12 16 20

node's mobilty speed (m/s)

Figure 5: average end to end delay comparison

Pratibha Chauhan, IJRIT

94

5

Conclusion In this paper, we propose an Ad hoc On-demand Distance Vector Multipath Routing Protocol with energy entropy. Our motivation is to provide the improvement of on-demand multipath routing method in terms of packet delivery ratio, average end-to end delay, and control packets ratio through simulation using Network Simulator (NS2). The simulation experiments showed that the considered EAODV algorithm is able to cope with this type of dynamic networks, in particular its ability to improve the system performance which has been reflected in the model. The key idea of EAODV algorithm is to construct the new metric-entropy with the help of entropy metric to reduce the number of route reconstruction so as to provide route packets in the ad hoc network.

6. REFERENCES [1] L. Sun, S. C. Pi, C. Gui, et aI, "Multiple Constraints QoS Multicast Routing Optimization Algorithm in MANET based on GA," Progress in Natural Science, Vol. 18,No. 3,2 008, pp. 331-336. [2] L. Sun, C . Gui,Q . F. Zhang, H. Chcn," Fuzzy Controller Bascd QoS Routing Algorithm with a Multiclass Scheme for MANET," International Journal of Computers, Communications & Control, Vol. IV, No.4,2 009, pp. 427-438. [3] L. Sun, S. C. Pi, C. Gui, J. Lian, "An Entropy-Based Stability Multipath Routing Algorithm in MANET," Journal of Computational Information Systcms,Vol. 5,No. 1,2009,p p. 229-234. [4] B. L. Sun, C. Gui, Q. F. Zhang, et aI, "A Multipath on-Demand Routing with Path Selection Entropy for Ad Hoc Networks," Proceedings of The 9th International Conference for Young Computer Scientists (rCYCS 2008), Zhang Jiajie, Hunan, China, 18-21 November, 2008, pp. 558-563. [5] C. E. Perkins, E. M. Belding-Royerand, I. D. Chakeres, "Ad Hoc On- Demand Distance Vector (AODV) Routing," IETF internet draft (July 2004),h ttp://www. i etf. orglinternet- drafts/draft -ietf-manet -aodv -13 .txt. [6] Z. Ye, S. V. Krishnamurthy, S. K. Tripathi, "A Framework for Reliable Routing in Mobile Ad Hoc Networks," Proceedings of IEEE INFOCOM 2003,30 March 30-- 3 April 2003, Vol. I,p p. 270 - 280. [7] S. Dulman, J. Wu, P. Havinga, "An energy efficient multipath routing algorithm for wireless sensor networks," Proceedings of the 6th international symposium on autonomous decentralized systems with an emphasis on advanced distributed transportation systems, Pisa, Italy, April 2003. [8] Q. L. Liang, Q. C. Ren, "Energy and mobility aware geographical multipath routing for wireless sensor networks," Proceedings of the IEEE Wireless Communications and Networking Conference, March 2005,p p.1 867-1 871 [9] Z. J. Wang, E. Bulut, B. K. Szymanski, "Energy Efficient Collision Aware Multipath Routing for Wireless Sensor

Networks," Proceedings of the 2009 IEEE international conference on Communications (rCC

2009), Dresden,G ermany, 14-18 June 2009,p p. 91-95.

Pratibha Chauhan, IJRIT

95

[10] Z. Ye, S. V. Krishnamurthy,S . K. Tripathi," A Framework for Reliable Routing in Mobile Ad Hoc Networks," Proceedings of the IEEE INFOCOM 2003,30 March -3 April 2003, Vol. I,p p. 270--280 [11] C. E. Shannon," The mathematical theory of communication," The Bell System Technical Journal,Vol . 27,1948,p p. 379-423,6 23-656 [12] B. An,S . Papavassiliou, "An Entropy-Based Model for Supporting and Evaluating Route Stability in Mobile Ad hoc Wireless Networks," IEEE Communications Letters, Vol. 6,No. 8,2 002, pp. 328-330. [13] K. Robinson, D. Turgut, M. ChattCljee, "An Entropy-based Clustering in Mobile Ad hoc Networks," Proceedings of the IEEE Conference on Nctworking , Sensing and Control (ICNSC06), Florida, USA, 2325 April, 2006,p p. 1-5. [14] H. Kawahigashi, Y. Terashima, N. Miyauchi, "A Proposal for a New Mcasure Analogous to Entropy for Bandwidth Constraincd, Control- Bascd Ad-Hoc Network Design," Military Communications Conference, 2006. MrLCOM 2006, Washington, DC, USA, 23-25 Oct. 2006,7 Pages [15] A. Shiozaki, “Edge extraction using entropy operator,” Computer, Vision, Graphics, and Image Processing, Vol. 36, No. 1, 1986, pp. 1–9. [16] C. E. Shannon, “A mathematical theory of communication,” Bell System Technical Journal, Vol. 27, No. 3, 1948, pp. 379–423. [17] B. Waxman, “Routing of Multipoint Connections,” IEEE Journal on Selected Areas in Communications, Vol. 6, No. 9, 1988, pp. 1617–1622. [18] The Network Simulator – NS–2: http://www.isi.edu/ nsnam/ns/.

Pratibha Chauhan, IJRIT

96