This paper presents an approach to calibrate GPS position by using the context awareness technique from the Pervasive Computing.
Previous researches on GPS calibration mostly focus on the methods of integrating auxiliary hardware so that the user’s context
information and the basic demand of the user are ignored. From the inspiration of the pervasive computing research, this paper
proposes a novel approach, called PGPS (Perceptive GPS), to directly improve GPS positioning accuracy from the contextual
information of received GPS data. PGPS is started with sampling received GPS data to learning carrier’s behavior and building a
transition probability matrix based upon HMM (Hidden Markov Model) model and Newton’s Laws. After constructing the required
matrix, PGPS then can interactively rectify received GPS data in real time. That is, based on the transition matrix and received online
GPS data, PGPS infers the behavior of GPS carrier to verify the rationality of received GPS data. If the received GPS data deviate
from the inferred position, the received GPS data is then dropped. Finally, an experiment was conducted and its preliminary result
shows that the proposed approach can effectively improve the accuracy of GPS position.
Index Terms—Context awareness, Pervasive Computing, GPS, Newton’s Laws, Markov Model, Maximum Likelihood Function.