Research And Improvement Of Scheduling Algorithm For Context Reasoning Based On Correlation

The main aiming of pervasive computing is to realize the data fusion of information space and the physical world. Computing will not be limited to traditional servers and desktops. Intelligent devices collect information, and share each other by network communication. In this way, devices can work together and provide users the services they want.The important characteristic of pervasive computing is intelligence and self-adaption of devices. Intelligent devices must have the ability to response the environment change and user behaviors in time and provide appropriate services. Context-aware computing is the important way to realize this. Different from traditional man-machine interaction, context-aware computing can make auto information collection and execution by intelligent devices. From the analysis of the information, intelligent devices can "perceive" the user demand, and then provide services to them.Context reasoning is the essential part of context-aware computing. And task scheduling is the important constituent part of any system. Also, context reasoning needs an effective strategy to manage reasoning tasks. This thesis discusses the characteristics of context reasoning scheduling and analyses some scheduling algorithm already proposed. On that basis, this thesis proposes a scheduling algorithm improvement based on correlation; Simulates the important application of pervasive computing—smart home as the lab table and conducts experiments; Verifies the applicability of the scheduling algorithm this thesis proposed in pervasive environment by comparative analysis among different algorithms.