- This topic is empty.
10th March 2008 at 16:44 #51389Jamee Codd of ericsson vietnamGuest
Need advice: Vietnam ericsson.
i have a Hi Ping Pong and Hi reversion also. i need to know how to reduce this or possible removed it. because its affecting my hsr badly.
My HO decision event is mostly happening on Lo SS rate.
indoor to outdoor senario.
which parameters should i touch?
pls. advise..10th March 2008 at 18:11 #51390PixGuest
You can increase the “target filtering” threshold. This is the mechanism that occurs after the detection of HO, and which filters out all the cells which do not propose a better RXLEV than the current cell.
For instance :
cell A is serving, threshold for HO due to low rxlev = -93dBm. The “filtering” threshold = +2dB
At a certain point, RXLEV(A) = -95dBm
The only possible targets are the neighbors that give more than -93dBm + 2dB = -91dBm.
If a neighbor is less than -91dBm, it cannot be a candidate. Therefore, the final candidate will have a much better level than cell A. It strongly reduces the risk of ping pong handover, if the pingpong occurs because during a HO “low rxlev”.
If the pingpong occurs during a better cell HO, you should increase the HO margin a little bit, increase the averaging window, and/or increase the ping pong margin.
Pix11th March 2008 at 06:00 #51391BilalGuest
If you are using Ericsson-3 Locating algorithm then there is no difference between a ‘better cell’ and ‘low rexlev’ handovers. My suggestion is that you should increase the value of HYSTSEP. Can you tell me the current value of HYSTSEP you are using and RXLEV of serving cell at which most of the handovers are occurring.19th March 2008 at 10:39 #51392ShaonGuest
What parameter is used to increase pingpong margin?19th March 2008 at 12:13 #51393pixGuest
parameters are vendor dependent, so you must tell us which vendor are you working with…
probably something like ho margin or ho hysteresis (unit = dB)…19th March 2008 at 15:17 #51394JameeGuest
thank you for all this idea.
Bilal, can you please explain on the part of adjusting HYSTSEP. what will happen if i increase the current setting. let say at the moment it is set to 70.
is this possible for IBS to Macro? since both are in a different layer, right?
what about hihyst and lohyst?20th March 2008 at 04:23 #51395BilalGuest
Ericsson 3 has got 2 different values of hysteresis i.e. HIHYST and LOHYST. If RXLEV of serving cell is greater than HYSTSEP HIHYST is used as hysteresis and if lower than HYSTSEP then LOHYST is used. Idea is to have more hysteresis when RXLEV is good and there is no chance that call will get dropped due to delay in handover and if RXLEV is low then to use lower hysteresis to avoid dragging the call.
Now if you are having ping pong handovers at a RXLEV which is lower than HYSTSEP then you may increase HYSTSEP(-ve value) to apply HIHYST to more handovers than before.
But if the cells are on different layers then you must tune LAYERTHR and LAYERHYST. The main target would be to make the IBS dominant everywhere in the building and let not be the handover to macro. If the ping pong handovers are occurring inside the building and you have enough rxlev of IBS there then you may increase LAYERTHR. But be sure that there is no spillage outside the building, in this case increasing LAYERTHR would make it serve outside the building as well.20th March 2008 at 06:57 #51396BilalGuest
You can also go for temporary penalties Jamee. PSSTEMP and PTIMTEMP are the parameters.22nd December 2008 at 10:23 #51397AliGuest
I have hi HO lost rate,from IBS micro site with its three macro neighbours,can any one suggest how can i improve the HSR stats as its current
HSR is 97.6,
HO rev rate 0.63
HO lost rate 1.73
Ping pong rate 21.7
can any one help?31st May 2012 at 10:40 #51398HankGuest
Could you please tell us which formula do you use to calculate the ping pong rate in Ericsson.4th June 2012 at 07:25 #51399BijoyGuest
Ping Pong HSR=
100 * HODUPFT / HOVERSUC
Bijoy28th November 2014 at 20:18 #51400Ashif KhanGuest
What is the ping pong????plz verify berfly??????