- This topic has 9 replies, 1 voice, and was last updated 13 years, 5 months ago by akefa.
4th June 2007 at 08:27 #47557adeelGuest
Hi, i have a confusion. in my network, our optimizers have defined a lot of neighbors for a single cell and some of them dont even make sense. i want to know the effect of defining un-necessary neighbours on the network.4th June 2007 at 13:42 #47558MaheshGuest
Some times it causes to have low Hand Over(HO) KPIs. Because unnessary neighbours always creates HO revisions and HO losts. As a results HO success get reduced.
With Ericsson experience, there are counters for total no of neighbour relations (NRs) in BSC and total no of measurement frequencies can be defined in BSC. With unwanted NRs these counters can be reached.
I think frequently you should remove unwanted NRs (Check stastics and remove NRs having low HO attempts)4th June 2007 at 13:51 #47559JorgeGuest
Number of (inactive) neighboring cells should not be a Hand Over issue since signalling received is not within useful range by the mobile. It may delay the mobile from attaching to the network when first powered on but this depends on what type of signal or coverage the subscriber is getting when mobile power-up happens.20th June 2007 at 15:52 #47560akefaGuest
Mahesh, the unnessary neighbours never decrease the KPI since the MS will ignore them. As Jorge said, it will delay the MS from attaching the network. Furthermore, the MS battery consume much power in a network containing unessary neighbours which will decrease the life time.20th June 2007 at 17:29 #47561PixGuest
And the mobile will loose time by trying to listen to all the neighbours, which means it will not update the rxlev of a certain neighbor as often as it could if it has less neighbours to listen to.
Please anyone correct me if i’m wrong…
regards,21st June 2007 at 21:09 #47562akefaGuest
Onc an turned-off MS is switched on, will search for the best bcch and there is no limit in time !
In active mode, an MS listens to the rxlev and quality for all the registered neigbours and send his measurements to the RBS on SACCH at regular intervals. PIx, you are right only if the time of listenning is longer than the regular intervals of sending the measurements. From practical view, does that happen? if so, could we play with the measurements rate?22nd June 2007 at 12:38 #47563PixGuest
Akefa, we can’t play with measyurements rates…22nd June 2007 at 13:13 #47564akefaGuest
pix, that means we can not be certain that unessary neigbours will affect the KPI22nd June 2007 at 14:22 #47565PixGuest
Yes, t’hat’s an open question. Actually, as far as i know, it can delay the reactivity of the handovers, because the n-cell measurements are a bit slow.
Will it degrade KPI or not…?22nd June 2007 at 16:20 #47566akefaGuest
Theoritically speaking, Yes
I think it is a practical issue since we have no clear figure about how many n sec will be needed to read m rxlev. Maybe the extra time to read extra rxlev is ignored compared to the HO message measurement rate.