Uploaded on

 

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads

Views

Total Views
244
On Slideshare
0
From Embeds
0
Number of Embeds
0

Actions

Shares
Downloads
0
Comments
0
Likes
0

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. Energy-Efficient Cluster-Based Service Discovery in Wireless Sensor NetworksScope: To minimizing the communication costs during discovery of servicesand maintenance of a functional distributed service registryAbstractWireless Sensor Networks (WSNs) is an emerging technology that opens awide perspective for future applications in ubiquitous computing andambient intelligence. A typical application of a WSN consists of gatheringdata from large areas and processing it at a central location. Designing aservice discovery protocol for WSN environments implies a number ofchallenges. Since sensor nodes are likely to be battery powered, the firstobjective is to minimize the energy consumption. As the energy is spent Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 2. mostly during communication, minimizing the energy consumptiontranslates into minimizing the communication cost.Problem Definition: The problem is challenging especially in large scale, dense networks,where significant traffic is generated due to the intrinsic broadcast nature ofthe wireless communication. A second challenge is to react rapidly to thenetwork topology changes, which directly affect the consistency of thedistributed directory.Proposed System: • Energy-efficient service discovery protocol • A lightweight clustering algorithm that builds a distributed directory of service registrations. Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 3. • An energy-efficient service discovery protocol that exploits the clustering structure. • DMAC Clustering AlgorithmModules: Discovery of services Maintenance of the distributed directory Clustering algorithm Construction of clustersDiscovery of services: Our service discovery protocol uses an underlying clusteringstructure, where the cluster heads (or root nodes) form a distributed directoryof service descriptions. During the discovery process, messages are Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 4. exchanged among the cluster head nodes. Therefore, the design issue forminimizing the discovery cost is that the root nodes have to be sparselydistributed on the deployment area. The clustering algorithm shouldconstruct an independent set of cluster heads, i.e. two root nodes are notallowed to be neighbors. Each node keeps a registry of service descriptions of the nodes placedbelow in hierarchy. The root node knows all the service descriptions offeredby the nodes in its cluster. Since the registration process requires unicastmessages to be transmitted from children to parents, it can be easilyintegrated with the transfer of knowledge on adjacent clusters. Thus, themessage Update Info is used for both service registrations and transferringthe knowledge on adjacent clusters. Integrated version of the Update Infomessage, where a node updates the information on both the adjacent clustersand the known services Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 5. Maintenance of the distributed directory: Make decisions based on 1-hop neighborhood information. Clusteringalgorithms that require each node to have complete topology knowledgeover a number of hops are expensive with regard to the maintenance cost.We aim to build a lightweight clustering structure that requires only the 1-hop neighborhood topology information. Avoid chain reactions. Several clustering algorithms suffer from thechain reaction problem, where a single topology change in the network maytrigger significant changes in clustering structure. For a distributed directorycomposed of cluster head nodes, a chain reaction leads to high overhead formaintaining consistent service registries. Therefore, an energy-efficientsolution should avoid chain reactions, such that local topology changesdetermine only local modifications of the directory structure. Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 6. Distribute the knowledge on adjacent clusters among clustermembers. The knowledge on adjacent clusters should be distributed amongthe ordinary nodes. Only the root needs to know all the nearby clusters.Clustering algorithm:Our clustering structure is a forest composed of a set of disjoint trees orclusters. The height of the cluster is the longest path from the root node to aleaf. We say that two trees are adjacent if there are two nodes, one from eachtree, that are connected through a link. • The capability grades are unique, as the node hardware identifier may be used to break ties. • The lower layers (such as MAC) filter out asymmetrical links, so that we can rely on bidirectional communication. • A node is aware of its neighbors and their capability grades. • The lower layers (such as transport) provide a reliable, best-effort message delivery service. Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com
  • 7. Construction of clusters • Nodes that have the highest capability grades among their neighbors declare the cluster heads and broadcast a Set Root message announcing their roles. • The remaining nodes choose as parent the neighbor with the highest capability grade. • When a node receives a Set Root message from its parent, it learns the cluster membership and rebroadcasts the Set Root message. Ambit lick Solutions Mail Id : Ambitlick@gmail.com , Ambitlicksolutions@gmail.Com