The Research Of DVS Algorithm Based On Hybrid-task Model For Wireless Sensor Network

As a fire-new technology of information acquisition, information processing andinformation transmission, Wireless sensor network is made up rapidly and flexiblyand not bound by the network. It can be used for emergency search, disaster relief,military fields and medical fields and has a very widely application prospect.However, energy-constrained is a great important characteristic of the wireless sensornetwork. So how to improve the network’s efficiency and extend its lifetime is themost important technology in the design, which directly determines the life of thenetwork. In all parts of the wireless sensor network, the sensor nodes should bedesigned efficiently and energy-saving is the core of the design.The paper focuses on the energy-saving technology in wireless sensor networkand the main research contents are as follows:(1) The background outline of the subject both national and international. Therehave been so many researches about the energy effectiveness. These energy-savingresearches about link-layer mainly concentrated on how to schedule the RF moduleand make the sensor node dormant in order to save energy. But the dynamic voltagescaling technique adjusted the voltage and the frequency to reduce the energyconsumption and maximize the energy efficiency when the CPU was working. Thetraditional DVS algorithm was read and analyzed in detail and its inherent defectswere pointed out.(2) For the application with a high degree of computational complexity, theauthor proposed a more general DVS algorithm based on hybrid-task model, which isabbreviated as H-DVS, in order to reduce the sensor nodes’ energy consumption.Without the limitation of the task’s deadline, the DVS algorithm could be combinedwith the EDF scheduling algorithm and supported the hybrid mission mode, that was,periodic tasks and sporadic tasks exist at the same time. Then the frequencyadjustment factor named α adjusted the CPU’s voltage and frequency in real timeregulation according to the CPU’s work load. Theoretical analysis showed that H-DVSalgorithm could reduce the energy consumption while maintaining the real-timerequirement of the tasks.(3) Using the OMNET++simulated system as the basic platform, an experimentwas successfully carried out to simulate H-DVS algorithm which was also combined with EDF algorithm. The results showed that the method was feasible and effective.The method could reduce the sensor node’s unnecessary energy-consumption and metthe purpose of extending the sensor node’s life.