I am working on Pedestrian dead Reckoning using Particle filtering. I am using mobile accelerometer sensor to detect number of steps and stride length and calculating the total distance traveled.
I want to use particle filter to enhance accuracy in my results. I have read the theory of particle filter but stuck at the point that how to apply the model to my case.
In particular my problem, I have the following questions:
How to determine the initial position
pdf
distribution? I have a Bluetooth beacon to detect the starting position. Particle filter requires a cluster of point at the start, right? I don't think in my case the distribution is Gaussian. Then what should it be?Should I do the particle filtering on the smartphone or pass the data back and do it on the computer using Matlab? Will the computation very intense.