Font Size: a A A

Energy Research On Clustered Protocols Of Ad Hoc Network

Posted on:2010-10-10Degree:MasterType:Thesis
Country:ChinaCandidate:W H ChenFull Text:PDF
GTID:2178360278472624Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
Ad Hoc Network is a wireless multi-hop network without pre-established infrastructure. The nodes work as both wireless terminal nodes with sending and receiving functions and router which forward information, All of the nodes are equal. There is no central controller and the MANET is completely distributed structure. Because of its non-infrastructure and network topology changing in dynamic characteristics, Ad Hoc network is faced with much new problems and challenges which are different from traditional networks, including the routing problem, the energy problem, QOS and security issues.In Ad hoc network, energy consumption is a very important standard of system performance. The energy issue not only has an impact on network performance, and even is related to the survival of the entire network. In order to reduce energy consumption, network routing algorithms use constraint to the selection of the most low-power routing, but it would affect the system throughput ability and the energy balance between the nodes. Though the energy consumption of whole network reduces, if the energy of individual nodes consumes too fast, it would impact the network connectivity and then shorten the lifetime of the network. Therefore, the node energy consumption balance is also a very key issue compare with energy consumption of network, especially in clustering routings.This paper mainly researches the way of saving energy in clustering Routings. The main research and innovative work are as follows:First, this thesis briefly discusses and analyzes the current clustering algorithms , and summarizes its advantages and disadvantages. Then the algorithm HD is improved. The betterments are as follows:Puts forward the idea of using selectable number of the clusterhead, when the degree of clusterhead is a bit larger, the algorithm adds deputy clusterhead to reduce the load of main clusterhead, and also expand the scope of cluster. So it improves the stability of the group. Generates the boundary nodes in clustering algorithm to optimize exterior-cluster routing.Introduces energy threshold in the process of cluster maintenance, so the clusterhead can automatically initiate a electoral process of new clusterhead when its energy consumption lowers a certain threshold. The selection process of the clusterhead would consider on the actual load and energy consumption of the nodes, and select the node with longest time, which it can work as clusterhead. The time can be calculated by each node, based on it own recent data stream and the load of the clusterhead.This paper analyses current routing protocol, and improves the clustering routing protocol CGSR. The new routing uses the improved clustering algorithms. In order to balance the energy consumption and prolong the life time of the network, the new protocol improves the following aspects on routing strategy,In intra routings, uses the method of energy-hop conversion so that avoid key nodes and nodes with low energy.In exterior routings, uses the border node to make the route as short as possible, and changes the original rules that exterior routing method which must be forward by the clusterhead. The clusterhead only provides routing information between neighboring clusters at setting up exterior routing, rather than forwards message, so the energy consumption of clusterhead is greatly reduced.At last, in the NS2 simulation software, we make the simulation of the clustering algorithm and the improved routing protocol respectively in the same scenes. Comparing the network run-time, delay between nodes and the failure time of nodes, the results prove that energy consumption is lower, and the life time of network is longer than the original protocol CGSR when the new protocol is used.
Keywords/Search Tags:Ad Hoc Network, Energy Balance, Lifetime, Clustering Algorithm, Routing Protocol
PDF Full Text Request
Related items