Roteamento baseado em localização geográfica: Protocolo Multi-Geo

• Autor(a) principal: Silva, Ariana Priscila da
• Data de Publicação: 2008
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/2588
• topic: Redes de sensores sem fio; Protocolos de roteamento; Redes de computadores; Wireless Sensor Network; Routing protocol; Computer network; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

This work presents Multi-Geo, a modified version of Multi-K, a protocol for Wireless Sensor Network (WSN). Multi-K is an adaptive hybrid protocol whose behavior is based on the network trafic in a given time. It is set to a reactive behavior when the frequency of events is low or it is set to pro-active behavior when the frequency of events increases. In the pro-active behavior, the Multi-K generates a partial spanning tree that must be maintained periodically. The diference between Multi-Geo and Multi-K is related to the use of virtual regions and the restriction of communication between the nodes in the network. In the Multi-K protocol, the nodes can communicate with any other node located in its signal range. In Multi-Geo, only nodes that are in the same virtual region are able to communicate. The exception is for a particular node in each region, called clusterhead. A clusterhead can communicate with clusterheads of other regions. It is through them that the messages generated on the network can be carried over diferent regions, in order to reach the sink node. The goal of the Multi-Geo protocol is to reduce the number of control and data messages in the network that are responsible for much of the total energy consumption. In addition, the Multi-Geo preserves the scheme of multi-hop sending that is another effective strategy for the economy in energy consumption of the network. The proposed changes were simulated using the NS-2 (Network Simulator version 2) and the results were compared with those of Multi-K protocol. The simulation scenarios considered two diferent situations: one in which the nodes and the events were uniformly distributed; and another one in which the nodes and the events had non-uniform distribution and the concentration of nodes was located in regions close to the sink node. For the first situation, the simulations showed that the approach proposed in the Multi-Geo protocol, vary considerably with the increase in the probability of failure. In scenarios with the probability of failure equals to zero, the data delivery rate was kept and a reduction in energy consumption was observed when compared to the Multi-K results. For simulation of scenarios with non-uniform distribution of nodes, Multi-Geo protocol showed a very interesting solution, since it provided a reduction of energy consumption quite significant in comparison to the Multi-K protocol and kept the data delivery rate. The latter set of simulations represents real applications in which the events tend to be located around specific areas and the clusterheads can be connected to a power source.