rf,
t3212 forces the MS to do a LU periodically (every t3212).
the point is that the VLR will clean its database from all MS which are “dead” (that did not perform a periodic LU).
If MS doesn’t do the LU, then the VLR will assume the MS is “off” and not pageable. The VLR actually waits longer than t3212 (it is t_detach, i believe, a VLR parameter usually = 2*t3212)
So by increasing t3212, more MS will be “dead” but not yet detected as such by the VLR. So paging will go on for those, and will not be answered. So paging response rate decreases. Not very important IMO, but some operators are monitoring this indicator.
Interference do not cause fake LU specifically… In your case I would assume that the area is
1/ rather empty, the number of calls is very low, the MS are mostly doing their LU once in a while. It looks like a lot of traffic, but it’s not that much
2/ faulty MS is firing channel requests “LU” to your BTS
3/ the area is in between two LACs and MS are jumping from one LAC to the other. Fix this by increasing your Cell Reselect Hysteresis parameter.
To ensure it is not an external interferer, just assign a new BCCH frequency to this BTS. Choose an ARFCN far away from the one you were using. If LU are still going on, then it is not because of interferer.
regards
pix