MKT,
a) yes, the MS measures only the neighbour BCCHs transmitted in the system information messages SI2 (or SI5 ? i can never remember).
HO Fail can occur if
1- the BCCH(n) is good, then the HO is trigerred,
a) the target TCH is on the BCCH TRX : the RxLev is known to be good, but RxQual can be poor –> interference. The HO ACCESS can’t be heard by the target, or the reply BTS->MS can’t be heard by the MS.
–> interference
–> poor Rx sensitivity of the BCCH TRX
–> hardware problem : TCH can’t be allocated on the timeslot, etc…
b) actually the HO COMMAND message from serving cell -> MS contains the reference of a target “TCH timeslot” that is not on the BCCH TRX. So the MS will jump onto a TCH, which is not on the BCCH TRX.
–> interference on TCH TRX
–> poor tx power of this TRX
–> poor Rx sensitivity of this TRX
–> hardware problem : TCH can’t be allocated on the timeslot, etc…
c) the measured BCCH(n) does not belong to a neighbor defined in OMCR (same BCCH, but different cells ! bad frequency planning, yes !)
Bad luck, it is also using the same BSIC as the real neighbor.
For the BSC point of view, this fake neighbor is the real neighbor. The HO COMMAND is sent to MS, but MS cannot access this “fake” cell : the BSC didn’t prepare the fake neighbour to receive this MS, it prepared the real one only.
–> Poor BSIC/BCCH planning
regards,
pix