dtu系统的动态电源管理策略的分析及实现word格式论文.docx

ABSTRACTWith the development of ―Internet of Things‖, the data transfer unit (DTU), which is a part of sensory network of IOT, has been widely used. The DTU device is responsible for data collection and transmission. The DTU devices have no access to other power but their own portable batteries generally. Therefore, how to better improve the utilization of battery and prolong system’s effective use of time, has become an important issue.The topic of this paper is about the research and implementation of dynamic power management (DPM) strategy in DTU system. There have been some strategies on DPM, but they do not behave well in some cases, this paper proposes a novel stochastic model, which is based on weighted Markov model. This stochastic model consists of four major components, including service request (SR), power management module (PM), service provider module (SP) and service queue (SQ). The weighted Markov model takes the correlation between historical statuses into account, the autocorrelation coefficients and weights of the sequence of historical statuses for each order must be calculated out firstly, and then calculates out transition matrixes for each order, finally calculates out the state vector by summing weights and transition matrixes, then the next state will be determined from the vector.The DPM strategy based on weighted Markov model is validated in the DTU system without operation system, DTU system has been designed into four layers, including device layer, which is charging for data collection, transmission and implementation of DPM, and the server layer charges for data reception, parsing, and the database layer is responsible for data storage, the upper layer processes the data from database. The DPM strategy also was validated in Android, we have designed an application to simulate the whole DPM strategy, including entering observed information manually and switching the system state as tips received. The experimental results show that the DP