Transition prob matrix generation:

1. convert timestamp to serial number and then sort
   
2. divide area into grids
   
3. loop: for each taxiid, find current grid & next grid, fill into grid matrix (row is current grid #, column is next grid #)
   
4. get prob matrix by normalizing grid matrix

---

Update generation:

1. divide updates into three different speed groups
   
2. in each speed group, set timestamp(poisson distribution) for each update
   
3. pick GUID (and its corresponding source AS) for each update randomly
   
4. compute destination AS (prob: how to match AS to grid in probability matrix)