PDA

View Full Version : Question Event Triggered or Perdiodic reporting



white_Noise
2011-10-05, 03:46 PM
Hi,

I was asked this question in an interview, and I would like to make sure If I did well,

"which reporting mode is better for the SHO success rate?"

Please advice..

jan74
2011-10-05, 09:31 PM
Hi,

I was asked this question in an interview, and I would like to make sure If I did well,

"which reporting mode is better for the SHO success rate?"

Please advice..


Thats a trick question because it depends on how counters are counted. Generally though, more reports(periodic) lead to better SHO Success Rate but it is a very costly way i terms of signaling load, to get your SHO SR up.

amit3111
2011-10-05, 11:12 PM
This is a question which doesn't have a straight answer. In GSM networks we used to have periodic reporting and handover performance in the networks were great. In WCDMA its more of an event triggered and handover performance is equally good,depending upon the cell plan.

Periodic can have more signalling load in a network as compared to Event triggered.

Amit

white_Noise
2011-10-06, 03:03 AM
This is a question which doesn't have a straight answer. In GSM networks we used to have periodic reporting and handover performance in the networks were great. In WCDMA its more of an event triggered and handover performance is equally good,depending upon the cell plan.

Periodic can have more signalling load in a network as compared to Event triggered.

Amit

Actually what I answered was based on "common sense", since I never read about this anywhere. My answer was is that the more reports you have, the better the judgement, which means the better the SHO success rate, and that lead to the periodic reporting, makes sense?

jan74
2011-10-06, 06:37 AM
Actually what I answered was based on "common sense", since I never read about this anywhere. My answer was is that the more reports you have, the better the judgement, which means the better the SHO success rate, and that lead to the periodic reporting, makes sense?

Not necesarily although the logic seems right. There is also an argument that adding signaling load can lead to more noise at the cell edge, reducing SHO SR. It is not a proper question to ask and sounds like somebody justifying their actions. We use event based for R99 traffic and the SHO SR is over 99.5%. a lot depends on the formula used.

T_ADD
2011-10-06, 08:19 AM
For SHO its better to have the option of "Event Reporting", rather than periodic. The simple reason is, the latter puts extra load on the ACH (in case of cdma) or reverse channel as they frequently need to keep reporting this metric.

So it will all depend on the actual network situation. For instance, if there's good coverage then event reporting would be a better option.

This is just one example, there are other things that could be considered.

white_Noise
2011-10-06, 05:23 PM
For SHO its better to have the option of "Event Reporting", rather than periodic. The simple reason is, the latter puts extra load on the ACH (in case of cdma) or reverse channel as they frequently need to keep reporting this metric.

So it will all depend on the actual network situation. For instance, if there's good coverage then event reporting would be a better option.

This is just one example, there are other things that could be considered.

So basically, there is no absolute answer for my question :)

T_ADD
2011-10-06, 05:51 PM
There can't be. Thats why parameters have ranges and choices. Its there as a provision to suit a particular condition or even a network phenomena.

FrankPintor
2011-10-08, 01:48 AM
From what I've seen periodic reporting is used when the network is new, maybe in an initial tuning phase, to collect good statistics about the soft handover performance. Then when the network is more mature event-driven reporting is used, because of the signaling load issues already mentioned.

Would that be a fair summary?

white_Noise
2011-10-08, 03:57 PM
From what I've seen periodic reporting is used when the network is new, maybe in an initial tuning phase, to collect good statistics about the soft handover performance. Then when the network is more mature event-driven reporting is used, because of the signaling load issues already mentioned.

Would that be a fair summary?

Kind of :) thanks..