28-07-2014, 01:59 PM
An Energy Efficient Neighbour Node Discovery Method for Wireless Sensor Networks
An Energy Efficient Neighbour.pdf (Size: 612.9 KB / Downloads: 33)
Abstract
The discovery of neighbouring nodes in multihop wireless networks has become a key challenge. Due to tribulations in communication, synchronization loss between nodes, disparity in transmission power etc, the connectivity of nodes will always experience disruptions. On the other hand, the energy utilization by the nodes also became critical. In this paper, we propose a new method for neighbour discovery in wireless sensor networks (WSNs) which pays an eminent consideration for energy utilization and QoS parameters like latency, throughput, error rate etc. In the proposed method, the network routing is enhanced using AOMDV protocol which can accurately discover the neighbour nodes and power management with HMAC protocol which reduces the energy utilization significantly. A complete analysis is being performed to estimate how the QoS metrics varies in various scenarios of power consumption in wireless networks.
INTRODUCTION
Recent researches in wireless technology focus towards making the devices more sophisticated and portable. The applications of wireless networks became momentous in many areas like military scrutiny, oceanographic studies, mine discovery etc. In the network, the sensors will be deployed on a large area and the data collected by the sensor has to be transmitted from the source to the sink with maximum accuracy and least power consumption. Since recharging of power sources of the nodes is intricate, there should be a proficient energy cutback mechanism. On the other hand, for successful communication of sensor nodes in multihop sensor networks the discovery of neighbour nodes is indispensable. The nodes in the network acts as routers, which transmits data packets from one neighbouring node to another. Most of the sensor networks consist of both static and mobile nodes. Many approaches have been proposed recently for neighbour node discovery. But they are not capable to muddle through the tribulations like frequent addition of new nodes, loss of wireless connectivity, augment in transmission power etc.The most essential prerequisite of a wireless network is efficient routing of information from a source to the desired destination. For this each node should maintain the Neighbourhood information locally. Such information proviso is maintained even in mobile networks also for tracking and other docking applications. As the number of pre-positioned wireless devices become greater than before, the distribution of communication channel became a major concern. In particular for intense networks, the collision of data packets lead to the drop in throughput so that there will not exist any significant network harmonization. In this scenario, the precise assessment of neighbour nodes becomes much pertinent. This paper examines, the most imperative hinder of energy consumption in wireless networks namely neighbour node discovery and formulate an efficient protocol with which the nodes can estimate the exact information of their neighbours even if they hold on to low power mode.
On the other hand, the overall power consumption is being abridged with incorporating much more enhanced power management schemes. With this we can amplify the number of neighbour discoveries per unit time and also allow locating the mobile nodes in the network. Earlier approaches focus on static interconnected networks and experiences bootstrap problem when the nodes desires to begin at the same time. Traditional reactive protocols are not sufficient for managing the error prone behaviour of multichip dynamic networks. They make use of single route strategy for a pair of source and destination node. Due to the dynamic configuration of network the possibility of recurrent node failures are high. This makes the predefined routes unsound. Finding alternative paths in case of failures will augment the delay in packet delivery. The problem of neighbour node discovery in wireless networks can be managed significantly by the method we proposed. The overall power utilization has reduced by introducing momentous variations in wake/sleep modes of every nodes. The nodes in the networks are associated with a default routing table which consists of all route information in the network. On demand routing protocols are found to be more beneficial for dynamic networks. In the proposed method, Adhoc On demand Multipath Distance Vector Routing (AOMDV) is used to enhance default dynamic routing schemes. Whenever there are multiple paths accessible, the route to be pursued preferentiality will depend up on the administrative distance value (AD) assigned to each
Related Work
Researches for competent neighbour node discovery protocols have become lively over many years. Many algorithmic and realistic methods have been proposed recently. Most of them were focusing towards the accuracy of node detection. But for multihop sensor networks, the power utilization problem is more critical. In centralized networks, an extraordinary node will be there, called access point which coordinates the sharing of the medium. All the messages will be transmitted through this node only. Thus the neighbour node discovery is the uncovering of a new node in the network. Conventional method of sending HELLO messages to all nodes is not apposite for networks with restricted power supply or networks for which recharging is a big deal [1]. As the no of nodes in the networks increases the broadcast rate of HELLO messages will rise. Therefore the overall energy consumption cannot be manageable. Dutta et al [2] proposed a sensible solution to the neighbour node discovery problem which is based on the radio wake time scheduling of nodes. It initializes the wake times by multiples of prime numbers in such a way that the sum of their reciprocals is exactly equal to the radio duty cycle of a specific application. The node will wake up for one counter period and increments a local counter if the local counter is divisible by any of the prime numbers. The algorithm was found to be efficient for communication between a single source and destination. When more than one node is desired to communicate with a single destination, the algorithm fails. Figure 2 shows the failure of discovery algorithm. Galluzzi et al [3] surveyed many protocols for low-power operation in sensor networks and formulated a new with randomized schedules, deterministic schedules of wake/sleep and sophisticated methods to ensure discovery of neighbour nodes. Hamida et al [4] analyzed the problem of neighbour node discovery for adhoc networks. Inspired by aloha, they Proposed a HELLO protocol. In this protocol all the nodes will be either in listening mode or in talking mode. Transmission of HELLO messages is initiated by a node randomly.
Problem Statement
typical wireless sensor network consists of both static and mobile nodes. The location of mobile nodes will change continually by time. As a result, the topology of the network cannot be pre-defined. Hence the connectivity between the nodes will get disturbed and nodes moves outside the range of communication. Neighbour node discovery problem is nothing but detecting the mobile nodes within one node’s communication range. The location information of nodes over time has to be updated accordingly. Also multiple numbers of nodes should not be allowed to access the same destination at the same time there by avoiding packet loss and errors. One of the eminent problems of MANETs termed as hidden node problem, which is being solved using RTS/CTS messages, cannot be adopted for solving the same in sensor networks. Because, stations are in sensor networks are either primary or secondary. In sensor networks incorporation of combined stations is very costly. Networks in which the nodes are not synchronized with each other will be detached frequently. Due to dynamic behaviour of nodes the location information also will be changed. The nodes can communicate efficiently only if there is a stable wireless connectivity between them. In most of the networks the power source for the nodes are batteries. Due to many practical limitations, the recharging of these batteries is very difficult. Hence the energy consumed by the nodes should be condensed as much as possible. The power used for communication and processing of messages has to be minimized to the maximum. On the other hand a large amount of energy is exhausted while nodes are idle. During the node listening period, the battery power will be wasted unhelpfully.
Conclusions
We analyzed the neighbour node discovery problems in static and dynamic multihop Sensor Networks. Conventional node discovery techniques were found to be inadequate and they give less significance to the QoS parameters. So as to overcome these issues we proposed a new method for node discovery in which AOMDV protocol is implemented for node discovery and route establishment. The comparative study with other routing protocols, we came to the inference that, our method can produce high discovery rate. On the other hand, for reducing the overall power utilization Hybrid MAC protocol is implemented. Results show significant reduction in power consumption with HMAC. Hence the end to end delay and routing over head and energy consumption are minimized and the throughput is increased remarkably.