A Context Speculation Method Based on Fuzzy Logic Inference in Crowd-Sensing Process

With the development of mobile devices, their abilities of computing power, perception and storage have been strengthened greatly, which makes mobile phone-based crowd-sensing meet an opportunity to develop rapidly. But there are still some challenges in the local analysis method of mobile device acting as client terminal. The paper proposes a context speculating method based on an improved fuzzy logic algorithm, which modifies the membership functions to compute the membership size, and then compute the probability of all the membership size to suit more complex situations, and finally get the context speculating result by the integrating of three axises acceleration information, which can provide a more precise decision to the server and reduce remarkably the burden of the server.