| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| Documento principal | 3.07 MB | Adobe PDF |
Advisor(s)
Abstract(s)
A optimização de protocolos com elevada eficiência energética em redes de sensores sem fios constitui um desafio e a sua aplicação em contextos de automação de processos e recolha de dados apresentou-se como uma grande motivação. Simularam-se aplicações capazes de encaminhar os dados pela rede até ao nó Sink para vários protocolos com diferentes parâmetros ajustáveis, tais como o tempo de escuta; dutycycle e o período de contenção. Consideraram-se sensores MICAz da Crossbow, a funcionar a 2,4GHz (IEEE 802.15.4) e suportados pelo sistema operativo TinyOS para instalação prática da aplicação nos nós sensores, ou suportada pelo simulador Castalia/OMNet++. Como cenário para as aplicações sem mobilidade, considera-se uma área para plantação onde os nós sensores a controlar a humidade do solo. Para as aplicações com mobilidade, considera-se que os nós sensores estão a monitorizar a localização de cada animal duma manada. Verificou-se, através de simulação em Castalia, que quanto maior for o período de escuta maior será o consumo de energia. Nas aplicações com mobilidade, o protocolo S-MAC é o que apresenta uma eficiência energética superior. Nas aplicações sem mobilidade, num cenário com 36 nós dispostos numa grelha rectangular, conclui-se que o protocolo T-MAC tem um desempenho energético ligeiramente superior. Caso se esteja a utilizar uma RSSF, numa aplicação prática em que todos os nós têm uma função crítica, deve-se escolher um valor para o período de contenção correspondente ao valor mais reduzido do desvio padrão da energia consumida. Caso a função de um nó sensor não seja crítica para o sistema, o valor óptimo para o período de contenção é dado pelo valor mínimo da média de consumo de energia da rede. No cenário considerado, os valores óptimos para o período de contenção são 40 e 50 ms, para os casos onde missão individual de cada nó é ―crítica‖ ou ―não crítica‖, respectivamente.
The optimization of highly energy efficient protocols for wireless sensors networks is a challenge and its application in contexts of process automation and data collection was faced as an endeavour. Applications able to route data through a network (towards the Sink node) were simulated for different adjustable parameters, such as the listening time; the dutycycle and the contention period. MICAz sensors from Crossbow operating at 2.4 GHz (IEEE 802.15.4) and supported by TinyOS were assumed both for the practical installation of the sensor nodes application and for the Castalia/OMNet++ simulator. As a scenario for applications without mobility, an agriculture planting area was considered, whose soil humidity is monitored by the sensors nodes. In turn, for applications with mobility, one assumes that the sensors nodes are monitoring the location of each animal in a herd. It was found through simulation in Castalia that the longer the listening period is the higher the energy consumption is. In applications with mobility, the S-MAC the protocol is the one with higher energy efficiency. However, for applications without mobility, in a scenario with 36 nodes deployed on a rectangular grid, the simulations showed that T-MAC energy efficiency is slightly better. In a practical application in which all the WSN nodes have a “critical” function, one should choose a value for the contention period corresponding to the lowest value for the standard deviation of the energy spent. If the function of a sensor node is not "critical" for the system, the optimum value for the contention period is achieved for the minimum average energy spent. In the considered scenario, the values of the contention period to be chosen are 40 and 50 ms, for the cases where the individual node mission is "critical" or "non-critical", respectively.
The optimization of highly energy efficient protocols for wireless sensors networks is a challenge and its application in contexts of process automation and data collection was faced as an endeavour. Applications able to route data through a network (towards the Sink node) were simulated for different adjustable parameters, such as the listening time; the dutycycle and the contention period. MICAz sensors from Crossbow operating at 2.4 GHz (IEEE 802.15.4) and supported by TinyOS were assumed both for the practical installation of the sensor nodes application and for the Castalia/OMNet++ simulator. As a scenario for applications without mobility, an agriculture planting area was considered, whose soil humidity is monitored by the sensors nodes. In turn, for applications with mobility, one assumes that the sensors nodes are monitoring the location of each animal in a herd. It was found through simulation in Castalia that the longer the listening period is the higher the energy consumption is. In applications with mobility, the S-MAC the protocol is the one with higher energy efficiency. However, for applications without mobility, in a scenario with 36 nodes deployed on a rectangular grid, the simulations showed that T-MAC energy efficiency is slightly better. In a practical application in which all the WSN nodes have a “critical” function, one should choose a value for the contention period corresponding to the lowest value for the standard deviation of the energy spent. If the function of a sensor node is not "critical" for the system, the optimum value for the contention period is achieved for the minimum average energy spent. In the considered scenario, the values of the contention period to be chosen are 40 and 50 ms, for the cases where the individual node mission is "critical" or "non-critical", respectively.
Description
Keywords
Redes de sensores sem fios Redes de sensores sem fios - Protocolos MAC Protocolos MAC - Validação