Hi all.
I have one question.
In event A5, why we don’t use cell specific and freq specific offset for pcell criteria?
Any specific reason?
The main reason is to avoid ping pong scenario amongst freq layers.
Generally cell specific offset is always set to zero, only hyst takes affect then.
As per spec, for event A5 pcell criteria… cell specific and freq specific offset is not defined.
This is not defined for pcell but defined for nbh cell.
Ideally if pcell and nbh cell are on diff freq then these offset shoul be defined for pcell also.
For source cell it is inherent the other way round where target layer will have similar definition towards source layer so hyst and offset will be there.
Any specifi reason why we defined for pcell in case of event A3?
This is very interesting question, Sir. I never made any deep reflection about this differences… Since it doesn`t have any explanation on 3GPP, I guess the reasoning comes from the nature of the events.
Thinking out loud…
A3 combines both measurements (neighbor and source) in a single metric, single formula, single condition - xx dB of delta between source and target. Providing more flexibility by providing offsets to primary cell may also be helpful in order to control the mobility and avoiding excessive handovers. Also, since A3 is usually used for intra-freq where you see lots of SON/MRO actions, you may also leverage of more parameters to fine tune your relationships.
A5 in the other hand, runs on aboslute values. You define what is the trigger value for the source (lets say, <-110dBm) and the minimum value of the target (let say, >-106dBm). Since it is already a fixed value, not a “delta” as in A3, you do not need to define the source Threshold, then compensate it with offsets, you define only the final value (-110+0 = -112+2). Still, to allow some LMS/SON customizations, you still have neighbor offsets to allow some customization on mobility.
What do you think? Does it make sense? Well, I`ve tried! LOL!