Distributed Particle Flow Filter for Target Tracking in Wireless Sensor Networks

Algorithm 1

Particle flow filter for node at time .

Input:  Output:

1: redraw particles from the Gaussian

 distribution which are the mean and

 covariance, respectively.

2: for do

3:   propagate particles

4: compute

5: for do

6:   Calculate the mean particles value

7:   set

8:   linearize the measurement function at to get

9:   Calculate A and b from (12) and (13), respectively, using

   ;

10:   for do

11:   migrate particles

12: Compute the mean value and error matrix of the particles by (14)