Looking for the suggestion to Improve CQI in High Speed Track.
Does we have any set of parameter or feature that can give some improvement in CQI KPI?
There is some high speed settings, but I think it’s for Mobility Handovers.
Not sure it will improve CQI or not.
To improve CQI is better to work on RF plan for high speed trains: improve dominance, improve best server area, improve SINR, reduce to zero areas with no dominance (more than 3 LTE cells within an RSRP window of 4 dB).
Also work on time to trigger for event A3, range and hysteresis on high speed cells.
There is no substitute to improve CQI by parameter/feature other than improving RF plan.
What are your settings for A3 event (A3offset, hysteresis and time to trigger) on high speed cells?
More than 3 LTE cells within an RSRP window of 4 dB?
What does it mean?
Second, regarding A3offset = 3 dB.
Yes, this is called no dominance.
So when you have area with 3 cells with their difference in RSRP less than 4 dB.
For high speed cells, do you serve it by towers having 3 sectors or by towers having 2 sectors dedicated just for high speed trains?
And hyst for A3?
Regular site with 3 Sector + 4CC (80 Mhz).
1 Sector we dedicate to High Speed track.
Now imagine what 320 msec mean for a high speed train of, let’s say, 200 km/h.
200km/h means 55 meters per second so about 19 meters for your 320 msec.
Does your train runs faster than 200 km/h?
Between 200 to 400 Km/h.
At 400 km/h means 320 msec corresponds to about 40 meters.
So everytime your change the serving cell by A3 event you will have the CQI and SINR degraded seriously for about 40 meters before A3 events is triggered.
So what would you suggest to setting for A3?
You can reduce time to trigger to 160 msec and increase A3 offset to 4 dB.
So you mean to delay A3, in High Speed scenario?
I mean to make it faster.
This way your sinr will still be degraded but only for 20 meters not for 40 meters.
So you will notice an improvement on CQI samples reported.
Yes, that’s our goal, to improve the sample of CQI reported from CEM.
I suggest: make a trial on 10 cells with this new setting and then conclude.
I was also thinking to do some PDCCH target BLER adjustment…
PDCCH target BLER to what value?
Overshooting cannot be combated by PCI blacklisting, because it will still degrade SINR.
PDCCH target BLER is 1.5%.
It’s a good value.