Research On Power Saving Schemes In MAC Layer For Wireless Sensor Networks

As a rapidly developing technology, wireless sensor networks(WSNs, Wireless Sensor Networks) have been used widespread and become one of the research hot spots in the field of information technology. In the designing of WSNs, due to the limits of wireless sensor node's power and hardware device, energy efficiency should be considered except for latency, quality of service, robustness and flexibility. Researchers have proposed all kinds of power saving schemes including many power saving schemes in network layer and MAC sublayer. And the subject of this thesis is just power saving schemes in MAC sublayer. The main works of this thesis are described as follows:First, the energy consumption sources and some power saving schemes in WSNs are investigated. After the investigation, it's clear that the radio transceivers almost consume whole of WSNs'power. Therefore researching how to reduce the energy consumption of transceivers is very important for saving power of WSNs and prolonging lifetimes of them. The thesis has carefully analyzed some protocol in the network layer and MAC sublayer of WSNs. After the research of network protocols, it's found that the main power saving schemes in network layer are shortening the distance of one hop, reducing the number hops, converging the data during transmission. And after the research of MAC protocols, it's found that the main power saving schemes in MAC layer are making the probability of conflicts during data transmission smaller, adopting periodical wakeup/sleep schedules, using preamble sampling technology to reduce idle listening.Then, presented is the research on S-MAC(Sensor MAC, S-MAC) wakeup/sleep schedules of which are synchronous and DPS-MAC(Dual Preamble Sampling MAC,DPS-MAC) wakeup/sleep schedules of which are asynchronous. After analyzing and comparing the energy efficiency and latency of them, it's found that although the synchronous scheme of S-MAC's schedules brings extra power consumption for synchronization, the scheme can effectively avoid the power consumption of short strobed preamble. Moreover, DPS-MAC detects channels and determines whether channels are free using RSSI to sampling channels two times(that is dual preamble sampling), so the energy consumption at the beginning of every schedule in S-MAC is not exist in DPS-MAC. After comparing advantages and drawbacks of them, the thesis presents a schemes with local synchronization, which combines the advantage of both. After analyzing,it's found that this scheme with local synchronization is more energy efficient.At last, based on network simulation platform NS2, the energy efficiency and latency of DPS-MAC with and without local synchronization are evaluated. The energy efficiency is evaluated by the ratio of the time when the transceivers are transmitting or receiving and time of the simulated progress. And the latency is evaluated by packets'average 6 hops delay.The results of the simulation show that local synchronization schemes could make DPS-MAC more energy efficient and keep nearly the same latency.