Wireless Sensor

(1)

M. Tech (ACS), NIT Warangal Page 1

1. INTRODUCTION

Wireless Sensor Networks, with the characteristics of low energy consumption, low cost, distributed and self organization, have brought a revolution to the information perception.

The wireless sensor network is composed of hundreds of thousands of the sensor nodes that can sense conditions of surrounding environment such as illumination, humidity, and temperature. Each sensor node collects data such as illumination, humidity, and temperature of the area. Each sensor node is deployed and transmits data to base station (BS). The wireless sensor network can be applied to variable fields. For example, the wireless sensor network can be used to monitor at the hostile environments for the use of military applications, to detect forest fires for prevention of disasters, or to study the phenomenon of the typhoon for a variety of academic purposes. These sensor nodes can self organize to form a network and can communicate with each other using their wireless interfaces. Energy efficient self organization and initialization protocols are developed in, [2]. Each node has transmitted power control and an Omni directional antenna, and therefore can adjust the area of coverage with its wireless transmission. Typically, sensor nodes collect audio, seismic, and other types of data and collaborate to perform a high-level task in a sensor web. For example, a sensor network can be used for detecting the presence of potential threats in a military conflict. Most of battery energy is consumed by receiving and transmitting data. If all sensor nodes transmit data directly to the BS, the furthest node from BS will die early. On the other hand, among sensor nodes transmitting data through multiple hops, node closest to the BS tends to die early, leaving some network areas completely unmonitored and causing network partition. In order to maximize the lifetime of WSN, it is necessary for communication protocols to prolong sensor nodes’ lifetime by minimizing transmission energy consumption, sending data via paths that can avoid sensor nodes with low energy and minimizing the total transmission power.

(2)

Figure 1.1 A typical Wireless Sensor Network.

(3)

1.1 Architecture of Wireless Sensor Network:

Figure 1.2 shows a typical schematic of a wireless sensor network (WSN). After the initial deployment (typically ad hoc), sensor nodes are responsible for self-organizing an appropriate network infrastructure, often with multi-hop connections between sensor nodes [30]. The onboard sensors then start collecting acoustic, seismic, infrared or magnetic information about the environment, using either continuous or event driven working modes. Location and positioning information can also be obtained through the global positioning system (GPS) or local positioning algorithms. This information can be gathered from across the network and appropriately processed to construct a global view of the monitoring phenomena or objects. The basic philosophy behind WSNs is that, while the capability of each individual sensor node is limited, the aggregate power of the entire network is sufficient for the required mission.

In general, the wireless sensor networks are deployed for monitoring at a large area so the wireless sensor networks need many sensor nodes. If the sensor node consumes completely energy, it is wasted. We do not consider to recharge and to reuse sensor node. Because of these reasons, the value of the sensor nodes must be inexpensive to practical use. Deployed in harsh and complicated environments, the sensor nodes are difficult to recharge or replace once their energy is drained. Meanwhile the sensor nodes have limited communication capacity and computing power. So how to optimize the communication path, improve the energy-efficiency as well as load balance and prolong the network lifetime has became an important issue of designing routing protocols for WSN. Hierarchical-based routing protocols [6] are widely used for their high energy-efficiency and good expandability. The basic idea of them is to select some nodes in charge of a certain region routing. These selected nodes have greater responsibility relative to other nodes which leads to the incompletely equal relationship between sensor nodes. LEACH (Low Energy Adaptive Clustering Hierarchy) [7], PEGASIS (Power-Efficient Gathering in Sensor Information System) [8] are the typical hierarchical-based routing protocols. As an enhancement algorithm of LEACH, PEGASIS is a classical chain-based routing protocol. It saves significant energy compared with the LEACH protocol by improving the cluster configuration and the delivery method of sensing data.

(4)

1.2 The differences between WSNs and traditional networks

Wireless sensor networks, on the one hand, share the similarity of self-configuration without manual management with Mobile ad-hoc networks; on the other hand, they are different from traditional networks in many aspects due to their strict energy constraints and application specific characteristics.

NO one-size-fits-all solution

:

A WSN is organized as a collection of sensor nodes which co-ordinate with each other to fulfil a certain task. The entire network infrastructure depends directly on the specific application scenario. It is unlikely that a one-size-fits-all solution exists for all these different applications. The old fixed protocol stack which applied successfully to traditional networks is no longer suitable for WSNs. Many new communication algorithms have been developed for different applications. As one example, WSNs are deployed with very different network densities, from sparse to dense deployments. Each case requires unique network configuration.

Environment interaction: The traffic loads relayed in WSNs are generated by the sensors

which interact entirely with the environment. By contrast, the traffic loads of tradition network are mainly driven by human behaviour. Moreover, the environment plays a key role in determining the size of the network, the deployment scheme, and the network topology. The size of the network varies with the monitored environment. For indoor environments, fewer nodes are required

to form a network in a limited space whereas outdoor environments may require more nodes to cover a larger area.

Resource constraints: Resource constraints include a limited amount of energy, short

communication range, low bandwidth, and limited processing and storage in each node. For wireless sensor networks, energy is a scare resource. This is unlike wireless ad-hoc networks which can recharge or replace batteries quite easily. In some cases, the need to prolong the lifetime of a sensor node has a deep impact on the entire WSN system architecture.

(5)

Reliability and QoS: The WSNs exhibit very different concepts of reliability and quality of

Service from traditional networks. They totally depend on the task assigned. In some emergency cases, only occasional delivery of packets can be more than enough; in other cases, very high reliability requirements exist. Packet delivery ratio in WSNs is no longer an sufficient metric, instead, different applications may take their own requirements into consideration

.

1.3Design challenges:

WSNs distinguish themselves from traditional networks due to their application specific and energy constraints. Their structure and characteristics depend on their electronic, mechanical and communication limitations but also on application specific requirements.

One of the major and probably most important challenges in the design of WSNs is their application specific characteristic. A sensor network is set up to fulfil a specific task and the data collected from the network may be of different types due to various application scenarios. Respectively, different types of applications have their own specific requirements. These requirements are turned into specific design properties of a WSN. In other words, a WSN's architecture directly depends on the assigned application scenarios. For the acceptable performance of a given task, the optimal WSN infrastructure should be selected out of the hundreds of network solutions before the practical deployment.

Equally, an issue that has been frequently emphasized in the research literature is the fact that energy resources are significantly limited. Recharging or replacing the battery of sensor nodes may be difficult or impossible. Hence, power efficiency often turns out to be the major performance metric, directly influencing the network lifetime. Power consumption according to the functioning of a sensor node can be divided into three domains: sensing, communication, and data processing. There has been research effort in hardware improvements to optimize the energy consumed by sensing and data processing. Several studies of energy efficiency of WSNs have been discussed and several algorithms that lead to optimal connectivity topologies for power conservation have been proposed [10][18].

(6)

M. Tech (ACS), NIT Warangal Page 6 Another issue in the design of WSNs is that performance assessment of a WSN always happens once deployed. The analysis procedure follows the order that people in this field first put more and more effort into inventing new protocols and new applications; then the solutions are built, tested and evaluated either by simulation or test beds; even sometimes an actual system has to be deployed so that researchers can learn by empirical evidence. A more scientific analysis procedure is ideally required before a WSN is practically deployed. Current WSN designers are mainly experts in wireless sensor networking and hardware who could perceive the communication between each node at the bit level. When a new protocol is developed, they could construct algorithms even if the required simulation tool did not exist. As WSNs immerse deeper into people's work, they must begin to include less specialized users.

1.4 Thesis Contributions

The work reported herein investigates chaining mechanism in PEGASIS using evolutionary algorithms like Ant Colony optimisation and Genetic algorithms and lifetime enhancement by chain leader selection criteria and maintenance of priority queue at each node if the next node fails. Lifetime measurement of WSN using various types of PEGASIS variants for both Homogenous and heterogeneous has been evaluated.

1.5 Thesis Outline

The thesis has been organised in the fallowing manner. Fallowing this chapter, chapter 2 presents extensive literature survey on routing algorithms for WSN. It mainly discusses energy efficient hierarchical routing protocols for WSN. Evolutionary algorithms are also presented in this section. Chain forming mechanism using GREEDY algorithm is presented in chapter 3. It mainly investigates the lifetime of PEGASIS protocol under various scenarios. Chapter 4 deals with Ant Colony Optimisation technique applied to PEGASIS protocol and lifetime Measurement. Chapter 5 deals with Genetic algorithm and its lifetime measurement. Chapter 6 gives the comparative study of all the algorithms proposed.

(7)

2 LITERATURE SURVEY OF ROUTING PROTOCOLS FOR WSN

2.1 Introduction

Wireless sensor networks have their own unique characteristics which create new challenges for the design of routing protocols for these networks. First, sensors are very limited in transmission power, computational capacities, storage capacity and most of all, in energy. Thus, the operating and networking protocol must be kept much simpler as compared to other ad hoc networks. Second, due to the large number of application scenarios for WSN, it is unlikely that there will be a one-thing-fits-all solution for these potentially very different possibilities. The design of a sensor network routing protocol changes with application requirements. For example, the challenging problem of low-latency precision tactical surveillance is different from that required for a periodic weather-monitoring task. Thirdly, data traffic in WSN has significant redundancy since data is probably collected by many sensors based on a common phenomenon. Such redundancy needs to be exploited by the routing protocols to improve energy and bandwidth utilization. Fourth, in many of the initial application scenarios, most nodes in WSN were generally stationary after deployment. However, in recent development, sensor nodes are increasingly allowed to move and change their location to monitor mobile events, which results in unpredictable and frequent topological changes [10].

Due to such different characteristics, many new protocols have been proposed to solve the routing problems in WSN. These routing mechanisms have taken into consideration the inherent features of WSN, along with the application and architecture requirements. To minimize energy consumption, routing techniques proposed in the literature for WSN employ some well-known ad hoc routing tactics, as well as, tactics special to WSN, such as data aggregation and in-network processing, clustering, different node role assignment and data-centric methods. In the following sections, introduction to current research on routing protocols has been presented.

(8)

2.2 Routing Challenges and Design Issues in WSNs:

Despite plethora of applications of WSN, these networks have several restrictions, e.g., limited energy supply, limited computing power, and limited bandwidth of the wireless links connecting sensor nodes. One of the main design goals of WSN is to carry out data communication while trying to prolong the lifetime of the network and prevent connectivity degradation by employing aggressive energy management techniques. In order to design an efficient routing protocol, several challenging factors should be addressed meticulously. The following factors are discussed below:

Node deployment: Node deployment in WSN is application dependent and affects the

performance of the routing protocol. The deployment can be either deterministic or randomized. In deterministic deployment, the sensors are manually placed and data is routed through pre-determined paths; but in random node deployment, the sensor nodes are scattered randomly creating an infrastructure in an ad hoc manner. Hence, random deployment raises several issues as coverage, optimal clustering etc. which need to be addressed.

Energy consumption without losing accuracy: sensor nodes can use up their limited supply

of energy performing computations and transmitting information in a wireless environment. As such, energy conserving forms of communication and computation are essential. Sensor node lifetime shows a strong dependence on the battery lifetime. In a multi hop WSN, each node plays a dual role as data sender and data router. The malfunctioning of some sensor nodes due to power failure can cause significant topological changes and might require rerouting of packets and reorganization of the network.

Node/Link Heterogeneity: Some applications of sensor networks might require a diverse

mixture of sensor nodes with different types and capabilities to be deployed. Data from different sensors, can be generated at different rates, network can follow different data reporting models and can be subjected to different quality of service constraints. Such a heterogeneous environment makes routing more complex.

(9)

Fault Tolerance: Some sensor nodes may fail or be blocked due to lack of power, physical

damage, or environmental interference. The failure of sensor nodes should not affect the overall task of the sensor network. If many nodes fail, MAC and routing protocols must accommodate formation of new links and routes to the data collection base stations. This may require actively adjusting transmit powers and signalling rates on the existing links to reduce energy consumption, or rerouting packets through regions of the network where more energy is available. Therefore, multiple levels of redundancy may be needed in a fault-tolerant sensor network.

Scalability: The number of sensor nodes deployed in the sensing area may be in the order of

hundreds or thousands, or more. Any routing scheme must be able to work with this huge number of sensor nodes. In addition, sensor network routing protocols should be scalable enough to respond to events in the environment. Until an event occurs, most of the sensors can remain in the sleep state, with data from the few remaining sensors providing a coarse quality.

Network Dynamics: Most of the network architectures assume that sensor nodes are

stationary. However, mobility of both BS‘s and sensor nodes is sometimes necessary in many applications. Routing messages from or to moving nodes is more challenging since route stability becomes an important issue, besides energy, bandwidth etc. Moreover, the sensed phenomenon can be either dynamic or static depending on the application, e.g., it is dynamic in a target detection/tracking application, while it is static in forest monitoring for early fire prevention. Monitoring static events allows the network to work in a reactive mode, simply generating traffic when reporting. Dynamic events in most applications require periodic reporting and consequently generate significant traffic to be routed to the BS.

Transmission Media: In a multi-hop sensor network, communicating nodes are linked by a

wireless medium. The traditional problems associated with a wireless channel (e.g., fading, high error rate) may also affect the operation of the sensor network. As the transmission energy varies directly with the square of distance therefore a multi-hop network is suitable

(10)

M. Tech (ACS), NIT Warangal Page 10 for conserving energy. But a multi-hop network raises several issues regarding topology management and media access control. One approach of MAC design for sensor networks is to use CSMA-CA based protocols of IEEE 802.15.4 that conserve more energy compared to contention based protocols like CSMA (e.g. IEEE 802.11). So, Zigbee which is based upon IEEE 802.15.4 LWPAN technology is introduced to meet the challenges.

Connectivity: The connectivity of WSN depends on the radio coverage. If there exists a

multi-hop connection between any two nodes continuously, the network is connected. The connectivity is intermittent if WSN is partitioned occasionally, and sporadic if the nodes are only occasionally in the communication range of other nodes.

Coverage: The coverage of a WSN node means either sensing coverage or communication

coverage. Typically with radio communications, the communication coverage is significantly larger than sensing coverage. For applications, the sensing coverage defines how to reliably guarantee that an event can be detected. The coverage of a network is either sparse, if only parts of the area of interest are covered or dense when the area is almost completely covered. In case of a redundant coverage, multiple sensor nodes are in the same area.

Data Aggregation: Sensor nodes usually generate significant redundant data. So, to reduce

the number of transmission, similar packets from multiple nodes can be aggregated. Data aggregation is the combination of data from different sources according to a certain aggregation function, e.g., duplicate suppression, minima, maxima and average. It is incorporated in routing protocols to reduce the amount of data coming from various sources and thus to achieve energy efficiency. But it adds to the complexity and makes the incorporation of security techniques in the protocol nearly impossible.

Data Reporting Model: Data sensing and reporting in WSNs is dependent on the application

and the time criticality of the data reporting. In wireless sensor networks data reporting can be continuous, query-driven or event-driven. The data-delivery model affects the design of network layer, e.g., continuous data reporting generates a huge amount of data therefore, the routing protocol should be aware of data-aggregation

(11)

Quality of Service: In some applications, data should be delivered within a certain period of

time from the moment it is sensed; otherwise the data will be useless. Therefore bounded latency for data delivery is another condition for time-constrained applications. However, in many applications, conservation of energy, which is directly related to network lifetime, is considered relatively more important than the quality of data sent. As the energy gets depleted, the network may be required to reduce the quality of the results in order to reduce the energy dissipation in the nodes and hence lengthen the total network lifetime. Hence, energy-aware routing protocols are required to capture this requirement.

2.3 Classification of Routing Protocols in WSNs:

In general, routing in WSNs can be divided into flat-based routing, hierarchical-based routing, and location-based routing depending on the network structure. In flat-based routing, all nodes are typically assigned equal roles or functionality. In hierarchical-based routing, however, nodes will play different roles in the network. In location-based routing, sensor nodes' positions are exploited to route data in the network.

A routing protocol is considered adaptive if certain system parameters can be controlled in order to adapt to the current network conditions and available energy levels. Furthermore, these protocols can be classified into multipath-based, query-based, negotiation-based, QoS-based, or routing techniques depending on the protocol operation. In addition to the above, routing protocols can be classified into three categories, namely, proactive, reactive, and hybrid protocols depending on how the source sends a route to the destination. In proactive protocols, all routes are computed before they are really needed, while in reactive protocols, routes are computed on demand. Hybrid protocols use a combination of these two ideas. When sensor nodes are static, it is preferable to have table driven routing protocols rather than using reactive protocols. A significant amount of energy is used in route discovery and setup of reactive protocols. Another class of routing protocols is called the cooperative routing protocols. In cooperative routing, nodes send data to a central node where data can be aggregated and may be subject to further processing, hence reducing route cost in terms of energy usage.

(12)

Figure 2.1 Classification of routing protocols

2.4 Previous Work:

In this section a brief review of the related work on the analysis of PEGASIS protocol is presented. Cosmin Cirstea [10] provides an up to date evaluation of routing protocols as well as a description of state of the art routing techniques for Wireless Sensor Networks (WSNs) that enhance network lifetime through efficient energy consumption methods. The tradeoffs between energy and communication overhead are studied. The advantages and disadvantages of each routing protocol with the purpose of discovering new research directions are highlighted.

Stephanie Lindsey et. al. [11], proposed PEGASIS (Power-Efficient Gathering in Sensor Information Systems) Protocol which is a near optimal chain-based protocol, an improvement over LEACH. In PEGASIS, each node communicates only with a close neighbour and takes turns transmitting to the base station, thus reducing the amount of energy spent per round.

Dali Wei et. al [12], proposes a distributed clustering algorithm that determines suitable cluster sizes depending on the hop distance to the data sink, while achieving approximate equalization of node lifetimes and reduced energy consumption levels. A simple

(13)

energy-M. Tech (ACS), NIT Warangal Page 13 efficient multi hop data collection protocol to evaluate the effectiveness of Energy Efficient Clustering. The end to end energy consumption of this protocol is caluculated. EC is suitable for any data collection protocol that focuses on energy conservation. Performance results demonstrate that EC extends network lifetime and achieves energy equalization more effectively than two well known clustering algorithms, HEED and UCR.

Ossama Younis et. al [13], decribed HEED (Hybrid Energy-Efficient Distributed clustering), that periodically selects cluster heads according to a hybrid of the node residual energy and a secondary parameter, such as node proximity to its neighbours or node degree. HEED terminates in O(1) iterations, incurs low message overhead, and achieves fairly uniform cluster head distribution across the network. With appropriate bounds on node density , intra cluster and inter cluster transmission ranges; HEED can asymptotically almost surely guarantee connectivity of clustered networks.

A three-layered routing protocol for WSN based on LEACH (TL-LEACH) is given by Deng Zhixiang et. al. [14]. The improved LEACH protocol is simulated and the simulation results show that TL-LEACH protocol has greatly improved WSN lifetime than LEACH protocol.

Indu Shukla [15], has discussed PEGASIS protocol. PEGASIS protocol forms a chain of sensor nodes, where each sensor node only communicates with their neighbours. Sensor nodes are deployed in harsh physical environment. Sensor nodes have very limited computation capability because they are limited by the battery power. It has been a challenge to maximize the use of energy of these sensor nodes to extend the network lifetime. The implementation of PEGASIS protocol is also presented.

Jian Wan et. al [16], presented a review of recent routing protocols in WSNs and classified them into three categories based on the network structure in WSNs. A description of the existing routing protocols is presented and discussed their advantages and disadvantages. Finally, paper is concluded with open research issues and challenges.

Tao Liu et. al [17], proposes a new type of routing protocol for WSN called PECRP (Power-efficient Clustering Routing Protocol), which is suitable to long-distance and complex data

(14)

M. Tech (ACS), NIT Warangal Page 14 transmission (e.g. patient-surveillance or chemical detection in agriculture), and for fixed sensor nodes of WSN. PECRP combines the advantages of some excellent cluster-based routing protocols together, such as HEED (Hybrid Energy efficient Distributed Clustering Approach), PEGASIS (Power Efficient Gathering in Sensor Information Systems) and so on. PECRP improves the mechanism in electing CHs (cluster heads) of LEACH, and elects more appropriate nodes to be CHs, which could prolong the lifetime of WSN obviously. In data transmission, PECRP uses multi-hop transmission that is called “circle domino effect based on distance to BS (Base Station)” to balance the energy consumption in nodes. multi-hop transmission can prolong the lifetime of WSN in narrow sense situation is proved based on mathematical proofs

Zheng Gengsheng et. al [18], described a two layer hierarchical routing protocol called Chain Routing Based on Coordinates-oriented Clustering Strategy (CRBCC), which gives a good compromise between energy consumption and delay. First, CRBCC makes balanced clustering according to y coordinates where each cluster has approximately equal number of nodes. Second, CRBCC makes chain routing by simulated annealing algorithm (SA) inside the cluster and elects chain leader in the order of x coordinates. Third, CRBCC makes chain routing again by SA method among chain leaders. Simulation results show that CRBCC performs better than PEGASIS in terms of energy efficiency and network delay.

Hao Wu et. al. [19], proposes a Chain-based Fast Data Aggregation Algorithm Based on Suppositional Cells (CFDASC) to solve this problem. In this algorithm, Author attributed each node to one suppositional cell according to the node location information. The nodes which are in one suppositional cell act as the cluster head of data collection in turn, then the head gathers and transmits data along the cells chain to the sink node. As a result, it accelerates the data aggregation process. Simulation shows that COSEN noticeably give a good compromise between energy efficiency and latency.

Considerable amount of energy may be wasted when nodes which are far away from sink node act as the head. DERP (Distance-based Energy-efficient Routing Protocol) is proposed by Hyunduk Kim et. al. [20] to address the problem of making far away as head. DERP is a chain-based protocol that improves the greedy-algorithm in PEGASIS by taking into account the distance from the

(15)

M. Tech (ACS), NIT Warangal Page 15 HEAD to the sink node. The main idea of DERP is to adopt a pre-HEAD (P-HD) to distribute the energy load evenly among sensor nodes. In addition, to scale DERP to a large network, it can be extended to a multi hop clustering protocol by selecting a “relay node” according to the distance between the P-HD and SINK. Analysis and simulation studies of DERP show that it consumes up to 80% less energy, and has less of a transmission delay compared to PEGASIS.

M. Tabibzadeh et. al.[21], proposed a hybrid protocol, called collectively Chain-based LEACH (CBL) that improves the Low-Energy Adaptive Clustering Hierarchy (LEACH) to significantly reduce energy consumption and increase the lifetime of a sensor network. CBL protocol uses LEACH and the advantages of Power-Efficient Gathering in Sensor Information Systems (PEGASIS) and avoids their disadvantages. LEACH technique improves energy efficiency of a sensor network by selecting a cluster-head, and having it aggregate data from other nodes in its cluster, and PEGASIS is a near optimal chain-based protocol used for communication and extra aggregation between cluster-heads that are neighbours and takes turns transmitting to the sink.

Wenjing Guo et. al. [22], presented a routing protocol for the applications of Wireless Sensor Network (WSN). It is a protocol based on the PEGASIS protocol but using an improved ant colony algorithm rather than the greedy algorithm to construct the chain. Compared with the original PEGASIS, this one, Pegant, can achieve a global optimization. It forms a chain that makes the path more even-distributed and the total square of transmission distance much less. Moreover, in the constructing process, the energy factor has been taken into account, which brings about a balance of energy consumption between nodes. In each round of transmission, according to the current energy of each node, a leader is selected to directly communicate with the base station (BS). Simulation results have shown that the proposed protocol significantly prolongs the network lifetime.

In order to reduce energy consumption, Young-Long Chen et. al. [23] first shows ideal energy mathematical model of PEGASIS topology, since the distance between nodes is the same, this energy mathematical model is the longest network lifetime of WSNs. To achieve this objective, Intra- Grid PEGASIS topology architecture is proposed, which is an architecture based on PEGASIS topology. In this architecture, the sensor area is divided into

(16)

M. Tech (ACS), NIT Warangal Page 16 several network grids, meanwhile, the nodes of each network grid is deployed in random, then the nodes within the network grid are connected, finally, all the network grids are connected.

Yongchang Yu et. al. [24], proposed EECB (Energy-Efficient Chain-Based routing protocol) which is an improvement over PEGASIS. EECB uses distances between nodes and the BS and remaining energy levels of nodes to decide which node will be the leader that takes charge of transmitting data to the BS. Also, EECB adopts distance threshold to avoid formation of LL (Long Link) on the chain.

Feng Sen et. al. [25], propose EEPB (Energy-Efficient PEGASIS-Based protocol). It is a chain-based protocol which has certain deficiencies including the uncertainty of threshold adopted when building a chain, the inevitability of long link (LL) when valuing threshold inappropriately and the non-optimal election of leader node. Aiming at these problems, an improved energy-efficient PEGASIS-based protocol (IEEPB) is proposed in this paper. IEEPB adopts new method to build chain, and uses weighting method when selecting the leader node, by assigning each node a weight so as to represent its appropriate level of being a leader which considers residual energy of nodes and distance between a node and base station (BS) as key parameters.

Young-Long et al. [26], discussed the PEGASIS topology architecture with the PBCA (Phase-Based Coverage Algorithm) to find the redundant nodes which can enter to sleep mode. Therefore, proposed algorithm can reduce the energy consumption of nodes and extend the network lifetime. Simulation results show that the performances of this algorithm outperformance the LEACH topology architecture, the PEGASIS topology architecture, and the LEACH with PBCA topology architecture in terms of energy consumptions, number of nodes alive, and sensing areas.

(17)

3. PEGASIS (Power Efficient Gathering in Sensor Information Systems

)

Wireless sensor nodes sense data and send it directly to the base station or they perform a clustering procedure as in LEACH. LEACH is known for cluster formation which contains cluster members sensing the data and the cluster head which gathers the data collected in a fused manner (all the data is sent as a single packet) to the base station. This procedure has gained in conserving a lot of energy that would otherwise be wasted. PEGASIS is an extension to LEACH; it has better ways of conserving energy which last even more than using cluster mechanism in LEACH [30].

When the nodes in the network which are at some distance from the base station, the easiest and the simplest way of transmitting the sensed data to the base station is to transmit it directly, which may lead to quicker depletion of energy in all the nodes. The nodes at a large distance away from the base station are depleted quicker than the nodes which are closer to the base station as they need some extra energy to reach the farthest base station. Another approach where energy is consumed in low amounts is by forming cluster heads and cluster members using the sensor nodes in the network. Cluster members perform the sensing and computing the data (Data Fusion) and the cluster heads transmit the fused data to the base station. All the nodes in the network take their chance to act as cluster heads to send the fused data to the base station; again the farthest cluster head needs some extra energy to send the data to the base station.

The key idea in using PEGASIS is that it uses all the nodes to transmit or receive with its closest neighbor nodes. This is achieved by the formation of a chain as shown in the Figure below. All the nodes which collect the data fuse it with the data received by the neighbor node and transmit it to the next-nearest neighbor. In this way all the nodes receive and fuse their data, and pass it to the next neighbor in a chain format till they all reach the base station. Every node in the network takes turns as a leader of the chain and the one responsible to transmit the whole fused data collected by the chain of nodes to the base station [31].

(18)

Figure 3.1: Chain formation in PEGASIS

In this way the average amount of energy spent by each node is reduced. Greedy algorithms are used to see that all nodes are used during the chain formation. PEGASIS assumes that all the nodes with varying or low energy levels can be compensated in order to calculate the energy cost of the transmissions with the remaining energy they are left with. It is not necessary that all the nodes need to know its neighboring nodes, the base station can determine the path or form the chain for all nodes, or all the nodes can determine their neighboring nodes by sending a signal. Depending upon the signal strength, the nodes adjust their signal such that they hear only the nearest neighbors in the network.

From Figure 3.2 below, the operation of PEGASIS is clearly understood. A greedy algorithm is applied to form a chain among all the best nodes that are at a one-hop distance from each other and to the base station. If the farthest node is selected, it starts transmitting the data. For example, if node 4 start the chain formation process and it sends the signal to the nodes

(19)

M. Tech (ACS), NIT Warangal Page 19 in the network to find the nearest neighbor, node 3 is the nearest, so it transmits the sensed data to node 4.

Upon receiving the data from node 4 node 3 starts finding the nearest neighbor by sending signals and when it finds that node 2 is the nearest, it fuses its own data with the data received from node 4 and transmits all this data to node 2. Node 2 finds node 1 as the nearest and transmits the sensed data with the fused data (the whole data is formatted a single packet). Now node 1 is the nearest node to the base station, so it acts as a leader and transmits all the data. Only the first node in the chain have nothing to fuse except the data it has during the chain formation, the remaining nodes all have some data to append with the received data from other nodes [31].

This approach will distribute the energy load evenly among the sensor nodes in the network as it uses all the nodes of the network to form the chain and perform simple data forwarding operations. If any node dies in the chain, a new chain is formed, eliminating the dead nodes.

(20)

M. Tech (ACS), NIT Warangal Page 20 From the simulation reported in [13], it is clear that PEGASIS improves on LEACH by saving energy at different stages, such as for example cluster-member forming and cluster heads. Here all the nodes have an equal chance of becoming the leader once and transmit data to the base station in one round. An energy balance is estimated on the nodes in the network which conserves lot of energy. The amount of nodes that die during the chain process is reduced when compared to LEACH for all types of network sizes and topologies. The network lifetime is increased, as all the nodes actively participate and deplete the equal amount of energy on the whole [13]. A simulation analysis of PEGASIS is reported in [13], comparing it with the LEACH protocol using different network topologies. Many experimental results proved that PEGASIS is supporting longer network lifetime, more balanced energy dissipation and higher performance.

PEGASIS uses a greedy algorithm to form a chain using the nodes in the network to transmit Data to the base station; it has no location awareness of the sensor nodes in the network and looks only for the closest neighbour that it can reach. Discovering a new route is difficult if a node fails, as it has a fixed path every time before it starts a new route towards the sink for transmission. Though its approach in conserving energy is better, it lacks in maintaining focus on quality-of-service factors. For instance, it cannot resist uneven traffic distribution for all those nodes which are not in the single-hop range; it has to make a multi-hop structure for adding such nodes

3.1 Greedy Algorithm Chain Formation

:

Greedy chain algorithm begins at a farthest node from the sink, which is the only node in the chain at first. Each terminal node of the chain finds a closest node from the remaining nodes set which are not in the chain. Then the closest node will join the chain and be the new terminal node of the chain. The process repeats till all the nodes join the chain. The greedy chain algorithm in PEGASIS is as follows.

The main advantages of PEGASIS are:

 The transmission distances between nodes are minimized.

(21)

M. Tech (ACS), NIT Warangal Page 21 The main drawbacks of PEGASIS are:

 It has excessive delay introduced by the single chain.

 Greedy algorithm using in PEGASIS is a local search, which cannot provide a global optimal route.

3.2 Data Aggregation in PEGASIS

In cluster-based sensor networks, sensors transmit data to the cluster head where data aggregation is performed. However, if the cluster head is far away from the sensors, they might expend excessive energy in communication. Further improvements in energy efficiency can be obtained if sensors transmit only to close neighbours. The key idea behind chain based data aggregation is that each sensor transmits only to its closest neighbour. In PEGASIS, nodes are organized into a linearchain for data aggregation. The nodes can form a chain by employing a greedy algorithm or the sink can determine the chain in a centralized manner. Greedy chain formation assumes that all nodes have global knowledge of the network. The farthest node from the sink initiates chain formation and at each step, the closest neighbour of a node is selected as its successor in the chain. In each data gathering

Procedure ConstructGreedyChain(N,END) 1. Begin

2. N={all nodes};

3. END = farthest node from SINK; 4. Chain= {END}; 5. N=N-{END}; 6. if (N!=NULL) 7. { 8. END=FindCloseNode(N,END); 9. Append(chain,END); 10. goto 5. 11. } 12. END

(22)

M. Tech (ACS), NIT Warangal Page 22 round, a node receives data from one of its neighbours, fuses the data with its own and transmits the fused data to its other neighbour along the chain. Eventually the leader node which is similar to cluster head transmits the aggregated data to the sink. Figure 3.3 shows the chain based data aggregation procedure in PEGASIS. Nodes take turns in transmitting to the sink. The greedy chain formation approach used in [33] may result in some nodes having relatively distant neighbours along the chain. This problem is alleviated by not allowing such nodes to become leaders.

Figure 3.3 Chain based organization in a sensor network. The ovals indicate sensors and the arrows indicate the direction of data transmission

The PEGASIS protocol has considerable energy savings compared to LEACH. The distances that most of the nodes transmit are much less compared to LEACH in which each node transmits to its cluster head. The leader node receives at most two data packets from its two neighbours. In contrast, a cluster head in LEACH has to perform data fusion of several data packets received from its cluster members. The main disadvantage of PEGASIS is the necessity of global knowledge of all node positions to pick suitable neighbours and minimize the maximum neighbour distance. In addition, PEGASIS assumes that all sensors are equipped with identical battery power and results in excessive delay for nodes at the end of the chain which are farther away from the leader node. In [29], two other protocols viz., a binary chain based scheme and a three-level chain based scheme have been proposed. In the binary chain based protocol, each node transmits data to a close neighbour in a given level of the hierarchy. The nodes that receive data at each level form a chain in the next higher level of the hierarchy. At the highest level, the leader node transmits the aggregated

(23)

M. Tech (ACS), NIT Warangal Page 23 data to the sink. In the three level schemes, the protocol starts with the formation of a linear chain among all nodes and then it divides them into G groups. Each group has N/G successive nodes of the chain where N is the total number of nodes. Only one node from each group participates in the second level of the hierarchy. The G nodes in the second level are further divided into two groups so that only three levels are maintained in the hierarchy.

3.3 Simulation Parameters

One Hundred Wireless Sensor Nodes are deployed randomly in 100m x 100m area each with initial energy of 1 or 2 Joule. Packet length of 1000 bits is assumed. The energy consumed in processing of one bit of data both in transmitting and receiving electronics (Eelec) is taken as

50nJ/bit. The energy consumed in transmitting amplifier (Eamp) for transmitting a bit for unit

distance is taken as 100pJ/bit/m2. The sink or gateway is assumed at the coordinates (25,150) so that a minimum distance of at least 50m from any node is present.

The fallowing figure shows the random deployment of WSN nodes in 100m X 100m area.

Figure 3.4 Random deployment of WSN Nodes

0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

100nodes random deployment

length in meters le n g th i n m e te rs

(24)

M. Tech (ACS), NIT Warangal Page 24 Figure 3.5 shows the chain formed in PEGASIS using greedy algorithm described earlier. The total length of the chain is 903meters.

Figure 3.5 Chain formations in PEGASIS using Greedy algorithm.

0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

100nodes network chain formation using greedy algorithm

lenght in meters le n g th i n m e te rs 903.834

(25)

3.3.1 WIRELESS SENSOR NETWORK RADIO POWER MODEL

A Wireless Sensor Network will comprise of the fallowing  A fixed base station (BS) and N wireless sensor nodes  BS has high-energy

Figure 3.6: WSN forming PEGASIS chain and Base Station

Figure 3.7: Transmitter and Receiver Energy Model diagram

For simulation of different WSN scenarios, we use a radio energy model in [33], in which the energy dissipation ET (k, d) of transmitting k-bit data between two nodes separated by a distance of d meters is given as follows:

2

( , ) ( )

Tx Tx elec Tx amp

(26)

M. Tech (ACS), NIT Warangal Page 26 Where Eelec denotes electronic energy and Eta denotes transmit amplifier parameters

corresponding to the free space model.

The energy cost incurred in the receiver of the destination sensor node is given as follows:

( )

Rx Rx elec

E k  K E _ (2)

For simulation a simple model where the radio dissipates Eelec = 50 nJ/bit to run the transmitter or receiver circuitry and Eamp = 100 pJ/bit/m2 for the transmit amplifier to

achieve an acceptable Eb/N0.

We know the energy resources are mainly consumed by radio and CPU components [8]. We use the above radio power model to compute energy dissipation of radio. However, the energy dissipation of CPU is more difficult to compute in the simulator, because CPU is driven by the software running on it.

(27)

3.4 Homogenous PEGASIS with Greedy Chain

In homogenous network all the nodes Wireless Sensor Network are having same energy of 1Joule each

.

3.4.1 Max Energy Node as Cluster Head

In this case the node having the maximum energy is selected as the cluster head. For this all the nodes while transmitting the data to the chain leader, also indicate their current energy and expected next state energy after the present transmission is also inserted. Thus increasing the packet length and unnecessary, which directly drains the battery of both transmitting and receiving nodes. Also chain leader has to transmit this high data length packet to the base station which consumes a lot of energy. To overcome the disadvantages of this cluster head selection in original Pegasis a new cluster head selection criteria is proposed.

Figure 3.8: Life time of Greedy Homogenous Max Energy Node as cluster head.

0 10 20 30 40 50 60 70 80 90 100 0 1000 2000 3000 4000 5000 6000

100Node Network LifeTime Measurement greedy homo max smooth

number of dead nodes in percentage

n u m b e r o f ro u n d s

(28)

3.4.2 Cluster Head selected sequentially

In this case the present chain leader selects the next chain leader by just passing to the immediate neighbour in the chain. This mechanism greatly eliminates the need for larger packet size and conserves a lot of energy in both transmission and reception electronics. The figure 3.7 shows the lifetime of the Wireless sensor network in this scenario.

Figure 3.9: Life time of Greedy Homogenous WSN, cluster head selected sequentially

3.4.3 Comparison of Max Energy and Sequential cluster Head scenarios

From the above figures it can be concluded that the WSN with cluster head selected sequentially has more lifetime than the one having the max energy node as cluster head. 10% of nodes are dead at around 5000 rounds in former case and around 7000 rounds in the latter case, an improvement of nearly 40% of the network lifetime. Similarly for 50% of dead nodes case it is 6700 for former case and it is 8200 for later case. For 100% dead node case it is 6700 and 9000 round respectively. This can be summarised in a table as fallows.

0 10 20 30 40 50 60 70 80 90 100 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

100Node Network LifeTime Measurement greedy hom seq smooth

(29)

GREEDY HOMO Max Energy Sequential % improvement

10% 5000 7000 40

50% 6700 8200 23

100% 6700 9000 34

Table 3.1 Lifetime comparison of Max Energy and Sequential for Homogenous WSN

Figure 3.10 Lifetime comparison of Greedy max energy and sequential cluster head

Hence selecting the cluster head sequentially greatly enhances the lifetime of the network, although few nodes which are far away from base station die sooner than others in this case. 0 10 20 30 40 50 60 70 80 90 100 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

100comparision of greedy hom max and sequence smooth greedy homo max lifetime greedy homo sequence lifetime

(30)

3.5 Heterogeneous PEGASIS with Greedy Chain

In heterogeneous network 80% of the nodes in Wireless Sensor Network are having energy of 1Joule each and the remaining 20% are having 2Joule of Energy.

3.5.1 Max Energy Node as Cluster Head

In this case since 20% of nodes are having more energy, only these nodes will become cluster head more often in the beginning of the network, soon they deplete their energy in transmission to the base station.

Figure 3.11 Lifetime of Greedy Heterogeneous – Max Energy

Thus from the figure it can be observed that nodes having more energy are alive till the end of the lifetime of the network.

0 10 20 30 40 50 60 70 80 90 100 0 1000 2000 3000 4000 5000 6000 7000 8000

100Node Network LifeTime Measurement greedy hetero max smooth

(31)

3.5.2 Cluster Head selected sequentially

In this case although some nodes are having higher energy, all nodes become cluster head equally. So the nodes having high energy tend to stay alive till the end of network lifetime and also the nodes which are far away from base station die soon. Hence network operation is not feasible after 50% of nodes die.

Figure 3.12 lifetime of Greedy Heterogeneous – Sequential cluster head

3.5.3 Comparison of Hetero Max and Sequential scenarios

It can be observed from the above figures that 10% of node die at around 5800 round for Max Energy where as it is 7300 round for Sequential cluster head selection. 50 % of nodes die in Max Energy case at around 6400 and it is 8400 for sequential cluster head selection. Max energy network Is completely down at 7950 rounds whereas it is 14000 for Sequential.

0 10 20 30 40 50 60 70 80 90 100 0 2000 4000 6000 8000 10000 12000 14000

100Node Network LifeTime Measurement greedy hetero seq smooth

(32)

GREEDY HETERO Max Energy Sequential % improvement

10% 5800 7300 25.9

50% 6400 8400 31.25

100% 7950 14000 76.1

Table 3.2 Lifetime comparison of Max Energy and Sequential for Heterogeneous WSN

Figure 3.13 Lifetime Comparison of Hetero Max and Sequential

3.6 Conclusion:

Hence it can be concluded that for Ant Colony Optimisation for both Homogenous and Heterogeneous sequential cluster head selection maximises the lifetime of Wireless Sensor Network lifetime. 0 10 20 30 40 50 60 70 80 90 100 0 2000 4000 6000 8000 10000 12000 14000

100comparision of greedy hetero max and sequence smooth greedy hetero max lifetime greedy hetero sequence lifetime

(33)

4. PEGASIS USING ANT COLONY OPTIMISATION

The chain formation using greedy algorithm does not give minimum distance path since it only takes the node which is nearer to seed node in formation of chain but it will not consider the total length of the chain formed. To solve this kind of optimization, evolutionary algorithms like Ant Colony Optimization are used which has good performance w.r.t to global optimization of NP hard problems.

Ant Colony Optimization (ACO) is a paradigm for designing metaheuristic algorithms for combinatorial optimization problems. The first algorithm which can be classified within this framework was presented in 1991 [35] and, since then, many diverse variants of the basic principle have been reported in the literature. The essential trait of ACO algorithms is the combination of a priori information about the structure of a promising solution with a posterior information about the structure of previously obtained good solutions.

ACO [34] is a class of algorithms, whose first member, called Ant System, was initially proposed by Colorni, Dorigo and Maniezzo [34]. The main underlying idea, loosely inspired by the behavior of real ants, is that of a parallel search over several constructive computational threads based on local problem data and on a dynamic memory structure containing information on the quality of previously obtained result. The collective behavior emerging from the interaction of the different search threads has proved effective in solving combinatorial optimization (CO) problems.

A combinatorial optimization problem is a problem defined over a set C = c1, ... , cn of basic components. A subset S of components represents a solution of the problem; F ⊆ 2C is the subset of feasible solutions, thus a solution S is feasible if and only if S ∈ F. A cost function z is defined over the solution domain, z : 2C ->R, the objective being to find a minimum cost feasible solution S*, i.e., to find S*: S*∈ F and z(S*) ≤ z(S), ∀ S∈ F.

Given this, the functioning of an ACO algorithm can be summarized as follows [9]. A set of computational concurrent and asynchronous agents (a colony of ants) moves through states of the problem corresponding to partial solutions of the problem to solve. They move by applying a stochastic local decision policy based on two parameters, called trails and

(34)

M. Tech (ACS), NIT Warangal Page 34 attractiveness. By moving, each ant incrementally constructs a solution to the problem. When an ant completes a solution, or during the construction phase, the ant evaluates the solution and modifies the trail value on the components used in its solution. This pheromone information will direct the search of the future ants.

Furthermore, an ACO algorithm includes two more mechanisms: trail evaporation and, optionally, daemon actions. Trail evaporation decreases all trail values over time, in order to avoid unlimited accumulation of trails over some component. Daemon actions can be used to implement centralized actions which cannot be performed by single ants, such as the invocation of a local optimization procedure, or the update of global information to be used to decide whether to bias the search process from a non-local perspective .

4.1 The Ant Colony Optimisation Metaheuristic Framework

A framework can be defined as the skeleton upon which various objects are integrated for a given solution. In other words it is a generic structure which is further specialised for a particular application. ACO is a generic algorithmic structure responsible for the scheduling of three processes:

1. Ants generation & activity 2. Pheromone trail evaporation 3. Daemon actions

This section defines these processes as well as other data structures required for the implementation of an ACO algorithm for a specific optimisation problem. A visual representation of the organisation of these processes is provided as Fig. 4.1

(35)

4.1.1 Pheromone mapping

The pheromone mapping is the means by which solution components are able to be ranked and selected based on past usefulness. The pheromone mapping connects pheromone values from a pheromone map (usually a matrix structure) to specific solution components. The assumption usually being that if a prior solution is good then at least some of its parts (solution components) should also be good and therefore a remixing of these components with other good components may lead to an optimal or near-optimal solution. A first step in defining an ACO algorithm is to define the pheromone mapping.

The problem domain will dictate how the pheromone mapping should be defined. In applying an ACO algorithm to a combinatorial optimisation problem such as the travelling salesman problem (TSP) it is not of interest which specific components are included, as any feasible solution will include every city once (and only once), it is the order of these components which is important in finding an optimal solution. For the TSP the transition points (edges/arcs) between the specific components can be assigned a specific pheromone value in order to reflect which order of cities works the best. That is, that if a solution included an edge connecting city A to city B and the solution is good then this should be reflected in the pheromone level on this specific edge and the other edges included in the solution.

4.1.2 Ants Generation and Activity

This process is responsible for the creation of new candidate solutions to the optimisation problem being addressed by the algorithm. A temporary population of (artificial) ants is used to construct feasible solutions to the problem being addressed. Each ant is evaluated upon the completion of a feasible solution and the solution information encoded into a global pheromone mapping. Each individual ant is discarded after entering their specific solution information into the pheromone mapping and a new ‘empty’ ant is created in its place, until some stopping criterion is met.

(36)

An ant has the following properties :

1. An ant searches for a minimum (or maximum) cost solution to the optimisation problem being addressed.

2. Each ant has a memory used to store all solution components used to date, so that the candidate solution can be evaluated at the completion of solution construction; the memory can be used as a tabu list such as in the case of the TSP so that no component is reused.

3. An ant can be assigned a starting position, for example an initial city in a TSP.

4. An ant can include any feasible solution component (an example of a feasible solution component in a TSP would be a city which has not already been included in the candidate solution) until such time that no feasible components exist or a termination criterion is met (usually correlating to the completion of a candidate solution).

5. Ants include solution components according to a combination of a pheromone value and a heuristic value which are associated with every solution component in the problem, the choice of which solution component is usually a probabilistic one. 6. When including a new solution component in the growing candidate solution the

pheromone value associated with the transition between these components (arc/edge in a TSP), or the solution component itself can be altered (online step-by-step pheromone update).

7. An ant can retrace a candidate solution at the completion of a solution, updating the pheromone values of all transitions and/or solution components used in the solution (online delayed pheromone update).

8.

Once a candidate solution is created, and after completing online delayed pheromone update (if required) an ant dies, freeing all allocated resources

4.1.3 Pheromone Trail Evaporation

Like the biological ant colony, the artificial ant colony employs a pheromone evaporation mechanism. This mechanism serves as a useful way of ‘forgetting’ older search bias [35]. As ACO uses positive reinforcement, if pheromone was allowed to accumulate without decay the system would very quickly converge on a single solution since this solution would continue to be reinforced.

(37)

4.1.4 Daemon Actions

Daemon actions can be used to perform specialised functions which often require more knowledge than an individual ant is allowed [35]. For example, a daemon action could inspect all solutions generated in one search cycle, identify the best solution and increment the pheromone values of its solution components more than the regular pheromone update (offline pheromone update). An alternative daemon action could be the application of a local search procedure.

4.2 ACO Algorithms

4.2.1 Ant System for the Travelling Salesman Problem

In this instance pheromone values correspond to transitions between cities (edges) and are Uniformly initialised to an amount slightly higher than what is expected to be added in one iteration of the algorithm as in Eqn. 4.1. After initialisation the AS algorithm runs a pheromone trail evaporation procedure which is implemented by applying the rule Eqn. 4.5 for every pheromone value. This procedure is followed by ants generation & activity which is implemented in the following steps:

1. A temporary population of m ants are placed at randomised starting cities. 2. Each ant k applies the random proportional rule Eqn. 4.2 to decide which city to

add to its current tour.

3. Step 2 is repeated until every ant k constructs a complete solution.

4. Every individual solution is evaluated and the edges used in this specific solution have their pheromone value adjusted according to Eqn. 4.4. This equation allocates a higher proportion of new pheromone to better solutions in order to ‘reinforce’ good decisions and is an implementation of an online delayed pheromone update

strategy.

(38)

M. Tech (ACS), NIT Warangal Page 38 The pheromone trail evaporation and ants generation & activity procedures are continually repeated until a termination criterion is reached, such as an amount of computation time, or alternatively by implementing a daemon action to observe the similarity of the solutions obtained over several iterations of the algorithm to test the convergence of the algorithm.

(4.1) (4.2)

 

( ) . ( ) 0 ij ij j k k ij ij ij k allowedk t E ifj allowed p t t otherwise           _   _{  }  _   _   _{  }  



(4.3) ij 1 ij d   (4.4) _ij  (1  ) _ij (4.5)

 ij: Pheromone value for edge connecting city i & j

 M : Number of ants  nn

C : Length of path found using a nearest neighbour heuristic  k

ij

p : Probability of ant k selecting the edge connecting city i & j  : Magnitude of pheromone influence on probabilistic decision  ij: Heuristic value for edge connecting city i & j

 : Magnitude of heuristic influence on probabilistic decision  dij: The distance between city i & j

 : Pheromone evaporation rate  Q: Amount of pheromone to deposit  L : Path length ij ij

Q

L









0

( , ),

i j

 

_ij_

m C

/

nn





(39)

4.2.2 Ant Colony Systems

Ant Colony Systems [37](initially introduced as Ant Q]) differs from AS in three areas:

1. Introduction of a local pheromone update. 2. Modification of the global pheromone update.

3. Modification of the random proportional rule to become the pseudo-random proportional rule.

The local pheromone update is applied by all ants during the solution construction phase. Every ant continually applies the update rule to the last solution component used as in Equ. 3.4.6. The aim of this pheromone update rule is to attempt to diversify the search process as much as possible during the solution construction phase. Without it most ants will simply create the same solution which will lead the search into a stagnation behaviour.

0

(1 )

ij ij

       (4.6)

The global pheromone update is modified so that only the best-so-far or iteration-best solution updates the pheromone map at the completion of solution construction. This means that unless a solution component has been included in the best solution it will not receive any modification from the global pheromone update.

(1 ) if ( , ) . otherwise, ij ij ij ij i j belongstothebesttour               (4.7)

The value of IJ reflects the utility of the solution and is dependent on the problem e.g.

for the TSP as in Sec. 4.1 it can simply be the inverse of the path length of the solution. The final and perhaps most important difference between ACS and AS is the modification of the random proportional rule to become the pseudo-random proportional rule. This rule introduces a new parameter q0. When a uniformly random value q in the range [0, 1] is less than q0, the largest transition probability value generated by Equ.3.4.2 is used, rather than using a roulette wheel selection of all generated probabilities.

(40)

4.3 Simulation Parameters

One Hundred Wireless Sensor Nodes are deployed randomly in 100m x 100m area each with initial energy of 1 or 2 Joule. Packet length of 1000 bits is assumed. The energy consumed in processing of one bit of data both in transmitting and receiving electronics (Eelec) is taken as

50nJ/bit. The energy consumed in transmitting amplifier (Eamp) for transmitting a bit for unit

distance is taken as 100pJ/bit/m2. The sink or gateway is assumed at the coordinates (25,150) so that a minimum distance of at least 50m from any node is present.

The fallowing figure shows the random deployment of WSN nodes in 100m X 100m area

Figure 4.2 Random deployment of WSN Nodes

0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

100nodes random deployment

length in meters le n g th i n m e te rs

(41)

M. Tech (ACS), NIT Warangal Page 41 The fallowing figure shows the chain formed in PEGASIS using Ant Colony optimisation described earlier. The total length of the best chain is 834m.

Figure 4.3: Chain formations in PEGASIS using Ant Colony Optimisation.

0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100

aco chain formation

20 834.0646 length in meters le n g th i n m e te rs