You are on page 1of 4

Fatemeh Najafi* et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 6, Issue No.

2, 261 - 264

Prolonging the Network Lifetime by Artificial Immune System


Fatemeh Najafi
Department of Computer, Science and Research Branch, Islamic Azad University Khouzestan, Iran Najafi_un@yahoo.com

Mashaalah Abbasi Dezfouli


Department of Computer, Science and Research Branch, Islamic Azad University Khouzestan, Iran Abbasi_masha@yahoo.com

Habib Rostami
Electrical Engineering Department, School of Engineering, Persian Gulf University of Bushehr, Bushehr 75168, Iran habib@pgu.ac.ir

Keywords-

Artificial networks; clustering.

Immune

System;

I.

INTRODUCTION

A sensor network is an interconnection of tiny, low cost, lowpowered and multi-functional sensor nodes [1, 2, and 6]. Sensor nodes are usually powered by lightweight batteries, and replacing or recharging these batteries is often not feasible. Therefore, in many cases, the lifetime of a sensor network is over as soon as the battery power in critical node(s) is depleted [2, 6, 7].With the development of wireless communication, electron and sensor technology, wireless sensor networks draw wide attention of people. A wireless sensor network contains a large number of sensors and a base station (BS). Sensors in such a network are equipped with ensign, data processing and radio transmission units. Distinguished from traditional wireless networks, sensor networks are characterized of severe power, computation, and memory constraints. It is usually impossible to supply sensors with energy. Due to the strict energy constraint, the main target of wireless sensor networks protocol is to balance the energy usage and prolong the network lifetime. Hierarchical (clustering) mechanisms divide the network into clusters. In each cluster, cluster head takes charge of members of the cluster and member nodes transmit the sensing data to the corresponding cluster head and the cluster head transmits

IJ
ISSN: 2230-7818

@ 2011 http://www.ijaest.iserp.org. All rights Reserved.

ES
Wireless sensor

Abstract Wireless Sensor Networks (WSNs) present a new Generation of real-time embedded systems with limited Computation, energy and memory resources that are being used in a wide variety of applications where Traditional networking infrastructure is practically Infeasible. Appropriate cluster-head node election can drastically reduce the energy consumption and Enhance the lifetime of the network. In proposed method, by clustering a sensor network into a number of independent clusters using an AIS algorithm, we can greatly minimize the total communication distance, thus prolonging the network lifetime.

the aggregated data to the BS. Because cluster heads are usually far away from BS, multi hop communication between cluster heads and BS is usually more energy efficient than direct transmission because of the characteristics of wireless channel [2]. Our hierarchical cluster-based routing technique is an extension of the LEACH [1] protocol that is a self-organized clusterbased approach for continuous monitoring. In LEACH, the network is randomly divided into several clusters, where each cluster is managed by a cluster head (CH). The sensor nodes transmit data to their cluster heads, which transmit the aggregated data to the base station. In hierarchical cluster based routing, each cluster is managed by a set of associates and the energy efficient clusters are retained for a longer period of time; the energy-efficient clusters are identified using heuristics-based approach. Moreover, in a variation of hierarchical cluster-based routing, the base station determines the cluster formation. An Artificial Immune System algorithm is used to generate energy-efficient hierarchical clusters. The base station broadcasts the (AIS) -based clusters configuration, which is received by the sensor nodes and the network is configured accordingly. For continuous monitoring applications, the simulation results show that hierarchical cluster-based routing is more energy efficient than the traditional cluster-based routing techniques. The remaining of this paper is organized as follow: original Related Work is explained in section 2. Artificial Immune System models are discussed in section 3. Proposed mechanism is explained in section 4. Simulation results are shown in section 5. Section 6 is the conclusion. II. RELATED WORK

Cluster-based approaches are suitable for continuous monitoring applications [1], [2]. For instance, Heinzelman et al. [2] describe the LEACH protocol, which is a hierarchical self-organized cluster-based approach for monitoring applications. The data collection area is randomly divided into

Page 261

Fatemeh Najafi* et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 6, Issue No. 2, 261 - 264

several clusters, where the number of clusters is predetermined. Based on time division multiple accesses (TDMA), the sensor nodes transmit data to the cluster heads, which aggregate and transmit the data to the base station. Bandyopadhyay and Coyle [5] describe a multi-level hierarchical clustering algorithm, where the parameters for minimum energy consumption are obtained using stochastic geometry. Hierarchical clustered based wireless sensor networks can distribute resources in a balanced way and are with special advantages related to scalability, efficient energy consumption, and simple data aggregation. The system model used in this work is based on the model presented in [2], where the energy dissipation is mainly from transmitting and receiving data. The radio module energy dissipation Es. l d d d l d d d (1) Where the Elect (J/b) is the energy dissipated to run transmit or receive electronics and amp (J/b/m2) is the energy dissipated by the transmit power amplifier to achieve an acceptable at the receiver. D0 is the path loss in open space on transmissions between two sensors apart from each other d meters and is a value between 2 and 4. L is the packet length. The main objective of the hierarchical cluster-based routing protocol is to generate energy-efficient clusters for randomly deployed sensor nodes, where each cluster is managed by a set of associates called ahead-set. Using round-robin technique, each associate member acts as a cluster head. CH receives messages from the cluster members and transmits the aggregated messages to a distant base station (BS). As all the transmissions are single-hop, cluster members transmit shortrange broadcast messages and CHs transmit long-range broadcast messages. The head-set approach can be a good solution for clusters where the CH dies during a round. Since the role of a CH is energy consuming, after a specified number of transmissions, a new set of clusters is formed. In other words, the clusters are maintained for a short duration called a round. A round consists of an election phase and a data transfer phase. In an election phase, the sensor nodes self-organize into a new set of clusters, where each cluster contains head-set. In data transfer phase, the head-set members transmit a specified number of long-range transmissions to BS. E lE lE

A. Election Phase During an election phase, clusters are created using a threshold function as given in (2).
p if n G 1 T ( n) = 1 p * ( r mod ) p 0 otherwise (2)

Where p is the desired percentage of cluster head in the sensor population, r is the current round, and G is the set of nodes that

IJ
ISSN: 2230-7818

@ 2011 http://www.ijaest.iserp.org. All rights Reserved.

ES

have not yet become head-set members for the last 1/p rounds. First, each node generates a random number, which is between 0 and 1. If the random number is less than T (n), the node becomes a head-set member and acts as a CH for this election phase. Second, each selected CH broadcasts a short-range advertisement. The sensor nodes may receive Advertisements from one or more CHs. Third; each sensor node chooses its CH on the basis of the signal strengths of the received advertisements. Fourth, the sensor nodes transmit short-range acknowledgments to inform their CHs about their decision. At this stage, the clusters for the current round are determined and each head-set has one associate member. Fifth, the CHs that have relatively large cluster sizes select a pre-defined number of additional head-set members for their clusters; the additional head-set members are chosen based on the signal strength of the acknowledgment messages. The selected head-set members cannot become head-set members until all the remaining nodes have become CHs. Iteration is defined as a duration in which all the nodes have become CHs. Sixth, each CH creates a TDMA schedule for the head-set members, as well as the remaining cluster members. Seventh, CHs broadcast their TDMA schedules to their members. At the end of election phase, each head-set member checks if it has sufficient energy for next round. If the energy of any headset member falls below the given threshold value, it is removed from the head-set; the remaining head-set members update their schedules accordingly. B. Data Transmission Phase During data transmission phase, the member nodes transmit Data to their CHs and the CHs transmit aggregated data to the base station. First, member nodes transmit data according to their TDMA schedule. (Fig1).

Second, CH receives the messages from the member nodes. As head-set members become CH on round-robin basis, only CH receives the messages and the remaining head-set members turn off their radios. Third, CH transmits the aggregated data to the base station. Fourth, CH checks the remaining energy. If the energy level is less than the given threshold value, CH will remove itself from the head-set. Fifth, the outgoing CH informs the incoming CH about its decision to remain the head-set member or become a cluster member. If the outgoing CH withdraws from the headset, the remaining head-set members update their schedules accordingly. At the end of round, all the clusters are not destroyed; however, each cluster is retained for the number of rounds equal to the head-set size. In other words, the nodes of clusters with the head-set size of 1 become candidates in the next round but the nodes of the clusters with the head-set size greater than 1 do not participate in the next election. This approach reduces the number of elections and the burden of long-range transmission

Figure 1. TDMA schedule in the LEACH

Page 262

Fatemeh Najafi* et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 6, Issue No. 2, 261 - 264

is more efficiently distributed among the nodes. Moreover, for the next election, the percentage of headers is decreased according to the number of retained clusters. The retaining of clusters has shown a significant amount of improvement as compared to the LEACH protocol. For performance evaluation, hierarchical cluster-based routing1 represents a protocol that contains a head-set but energyefficient clusters are not retained. Whereas hierarchical clusterbased routing represents a protocol that contains the head-set members and energy efficient clusters are also retained.

which each bit corresponds to one sensor or node. 1means that corresponding sensor is a cluster-head; otherwise, it is a regular node. Artificial Immune System is used to select cluster-heads. B. Affinity parameters The total transmission distance is the main factor we need to minimize. In addition, the number of cluster heads can factor into the function. Given the same distance, fewer cluster heads result in greater energy efficiency. The used energy for conveying the message from cluster to sink and the sending distance are the main factors that we need to minimum them. In addition to these, we can insert the decreasing number of clusters in our function that it can affect the energy function like the decreased sending distance, because the clusters use more energy in spite of other nodes. So the cells affinity of all the above affinity parameters that we define it likes. Cluster distance (regular nodes to cluster head, cluster head to sink distance): The cluster distance, is the sum of the distances from the nodes to the cluster head d x, c and the average of distance from the each node to its cluster head:

III.

ARTIFICIAL IMMUNE SYSTEM MODELS

The base station uses an Artificial Immune System to create energy-efficient clusters for a given number of transmissions. The node is represented as a bit of a cell. The head and member nodes are represented as 1s and 0s, respectively. A cell consists of several ids. The best id is selected as memory cell. A. Problem statement Finding appropriate cluster heads is critically important for minimizing the distance. We use binary representation in

IJ
IV. PROPOSED METHOD
ISSN: 2230-7818

The Artificial Immune System algorithm outcome identifies suitable clusters for the network. The base station broadcasts the complete network details to the sensor nodes. These broadcast messages include: the query execution plan, the number of cluster heads, the members associated with each cluster head, and the number of transmissions for this configuration. All the sensor nodes receive the packets broadcasted by the base station and clusters are created accordingly; thus the cluster formation phase will be completed. This is followed by the data transfer phase.

The base station uses an Artificial Immune System to create energy-efficient clusters for a given number of transmissions. The node is represented as a bit of a chromosome. The head and member nodes are represented as 1s and 0s, respectively. A population consists of several cells.

ES
Affinity=

AIS are distributed adaptive systems for problem solving using models and principles derived from the Human Immune System [3]. The capabilities of the NIS is mainly the inner working and cooperation between the mature T-Cells and BCells that is responsible for the secretion of antibodies as an immune response to antigens. The various theories with regards to the functioning and organizational behavior of the natural immune system (NIS), is discussed. These theories inspired the modeling of the NIS into an artificial immune system (AIS) for application in non-biological environments. Many different AIS algorithm models have been built, including Classical View Models, Clonal Selection Theory Models, Network Theory Models and Danger Theory Models. Artificial Immune Systems have been successfully applied to many problem domains. Some of these domains range from network intrusion and anomaly detection, to data classification models, virus detection, concept learning, data clustering, robotics, pattern recognition and data mining [4].

T
d x, c d c ,s
, ,

cluster

As we explained here Clusteri: means the ith cluster [Clusteri]: number of cluster nodes Ci means cluster head of the ith cluster

C. Affinity function of each cell is defined The used energy for conveying the message from cluster to sink and the sending distance are the main factors that we need to minimum them. In addition to these, we can insert the decreasing number of clusters in our function that it can affect the energy function like the decreased sending distance, because the clusters use more energy in spite of other nodes.

In fact, each cell for nodes clusters in suggests affinity to each cell with energy consumption for data transmission than has ratio. The cellular affinity that is greater than offer the more optimal method of cell. Immune cells that have more affinity are more reproducing the more mutations they found that immune cells that a low affinity are more ratio of this action are. V.
SIMULATION

For simulating of this algorithm, we have selected a rectangle area with 100m in length and 100m in width, and initially positioned 100 nodes in it randomly. The length of each transmission packet is 50 bytes. The radio range for each sensor is 40m. Also the position of the sink is considered in the upper right corner of the area. Sensors were randomly

@ 2011 http://www.ijaest.iserp.org. All rights Reserved.

Page 263

Fatemeh Najafi* et al. / (IJAEST) INTERNATIONAL JOURNAL OF ADVANCED ENGINEERING SCIENCES AND TECHNOLOGIES Vol No. 6, Issue No. 2, 261 - 264

simulated by MATLAB simulator. The Parameters of simulation are listed in Table1.

VI. CONCLUSIONS Energy efficiency in wireless sensor networks is important, so that the energy efficiency increase leads to network lifetime. In this paper, using an artificial immune system into base station as a central runs so we try that has minimum energy consumption. We presented a new clustering algorithm for minimizing the number of cluster heads. The simulation results have demonstrated the proposed clustering algorithm is energy efficient and highly prolonging the network lifetime. This method to minimize communication distance in sensor networks via clustering and increases the lifetime of sensor network.

Table.1 simulation Parameters Parameters


Network size Node Number(N) Initial Energy Eelect fs D0 Eda Packet length(L)

Value
100 *100 100 0.1 J 50 nJ/bit 10 pJ/bit/m2 87 m 5 nJ/bit/signal 50 Bytes

REFERENCES

[1]

A
Leach AIS 100 150 Round Number 200 250

In Fig 2, the presented algorithm and LEACH algorithm have been compared in terms of decrease of rate of energy consumption in 150 periods of transmitting and receiving data. As it is shown, our algorithm makes the rate of decrease in energy consumption of network smoother and more constant comparing to other algorithm. The simulation results have demonstrated the proposed clustering algorithm is energy efficient and highly prolonging the network lifetime. This method to minimize communication distance in sensor networks via clustering and increases the lifetime of sensor network.
Network Life Time 100 80 60 40 20

IJ
Alive Nodes 0 0 50

Figure 2. Compression of proposed method with LEACH protocol

ISSN: 2230-7818

@ 2011 http://www.ijaest.iserp.org. All rights Reserved.

ES
[3] [4] [5] [6] [7] [8]

In a comparison to other algorithms, we found that the life time of the network in our method has been increased giants the normal LEACH [1] algorithm. This relates to nodes positioning based on affinity function in AIS; if nodes have been positioned properly regarding expected energy consumption and network coverage, network's life time would be increased considerably.

[2]

W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Energyefficient communication protocol for wireless microsensor networks, in Proceedings of the Hawaii International Conference on System Sciences, January 2000. W. R. Heinzelman, A.P. Chandrakasan. An Application-Specific Protocol Architecture for Wireless Micro-sensor Network. IEEE Transactions on Wireless Communications, Vol. 1, No. 4, 2002. A. M. Bidgoli, .A. K. Javanmardi, A. M. Rahmani,Application of AIS algorithm for optimization of TORA protocol in ad hoc network, 2010. A. P. Engelbrecht , Computational Intelligence An Introduction ,second edition, wiley 2007. S. Bandyopadhyay and E. J. Coyle, An energy efficient hierarchical clustering algorithm for wireless sensor networks. in Proceedings of the IEEE Conference on Computer Communications (INFOCOM), 2003. C.-Y. Chong, S.P. Kumar, Sensor networks: volution, opportunities, and challenges, Proceedings of the IEEE 91 (8) (2003) 12471256. E. J. Duarte-Melo, M. Liu, Analysis of energy consumption and lifetime of heterogeneous wireless sensor networks, in: Proceedings of the IEEE Global Telecommunications Conference, Taipei, Taiwan, vol. 1, 2002, pp. 2125. K. Akkaya, M. Younis, A survey on routing protocols for wireless sensor networks, IEEE Transactions on Mobile Computing 3 (3) (2005) 325349.

Page 264

You might also like