You are on page 1of 11

IRAT HO Failure Ratio

Analysis - Huawei
All Rights Reserved Alcatel-Lucent 2008
Introduction
IRAT Fail Ratio for NBH and BBH hours are consistently high in Huawei circles
based on Daily NQI report.
This is one KPI which is usually dragging NQI into negative territory.


NBH IRAT Trend BBH IRAT Trend
0.00
2.00
4.00
6.00
8.00
10.00
12.00
DELHI
CSVIRATFailureRate(
BBH)
MUMBAI
CSVIRATFailureRate(
BBH)
Chandigarh
CSVIRATFailureRate(
BBH)
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
DELHI
CSVIRATFailureRat
e(NBH)
MUMBAI
CSVIRATFailureRat
e(NBH)
Chandigarh
CSVIRATFailureRat
e(NBH)
All Rights Reserved Alcatel-Lucent 2008
IRAT HO Analysis Mumbai Feb 9, Feb 16 (Daily Total Stats)
Feb 9, Feb 16 raw data have been
used for this analysis.
Total IRAT attempts (whole day Feb
9) = 89021. IRAT failure due to
physical channel failure = 2569
(2.88% of total attempts).
Total IRAT attempts (whole day Feb
16)= 86247. IRAT failure due to
physical channel failure= 2197 (2.54%
of total attempts).
IRAT Failure rate during the NBH,
BBH hour only will be higher than
daily total.
Majority of these failures are due to
cause Physical channel failure.
IRAT Failure Causes for Delhi
24-Feb-2011
All Rights Reserved Alcatel-Lucent 2008
Mumbai drive analysis example - 1
e2d (compressed mode) triggered at good
RSCP and Ec/Io
All Rights Reserved Alcatel-Lucent 2008
Mumbai drive example analysis 1 contd.
Failure reason = Physical Channel Failure
Call returned to
3G after the failure
And then successfully
Handed over to 2G
In later stage.
All Rights Reserved Alcatel-Lucent 2008
Mumbai drive analysis example- Layer 3
IRAT Fail due to Physical channel fail. GSM cell levels were good. No
congestion issues found on target GSM cell. As mentioned before, call
handed over to another GSM cell in later stage of the call.

All Rights Reserved Alcatel-Lucent 2008
Analysis of a Cell (12972) with High IRAT failure
RNC1/BSC6900UCell:Label = W_MB_MMBAI_068_10@2, CellID =12972 with 21
IRATHO.AttOutCS (IRAT attempts) and 20 failures, all due to physical channel
failure.
Congestion in target 2G cells checked - No congestion.
Checked for missing 2G neighbor cells 24, 3G-2G neighbors defined and 10
2G-2G neighbors defined on co-located 2G cell. 1 neighbor from 2G/2G
neighbor not defined but HO attempts % from existing 2G/2G cell relation is
only 1.8%.
3G-2G BCCH/BSIC discrepancy: NO discrepancies found.
All Rights Reserved Alcatel-Lucent 2008
IRAT thresholds Mumbai
Inference: Different threshold values are in Mumbai and Delhi
RNCs. In Mumbai Ec/No targets are -15 dBm while in Delhi, same
is set at -14 dBm.
All Rights Reserved Alcatel-Lucent 2008
IRAT HO analysis for Delhi Feb 16 (Daily Total Stats)
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
IRATHO attempts due to RSCP is very high (80,588) as
opposed to IRAT HO attempts due to Ec/No (5381)
Inference: More IRAT HO attempts triggered at RSCP values below
-100 dBm but at good Ec/No values.

These counters are currently not showing any data in Mumbai.
All Rights Reserved Alcatel-Lucent 2008
Key Observations
Attempts in the range of ~ 6000 for BBH and NBH.
Some counters like VS.IRATHO.FailRelocPrepOutCS.Abort ,
IRATHO.FailRelocPrepOutCS.HigherTrafficLod,
IRATHO.FailRelocPrepOutCS.HigherTrafficLodare missing in the Mumbai raw
data but available for Delhi.
The counter that shows a higher percentage of failures is
IRATHO.FailOutCS.PhyChFail .- Physical channel failure.
Very high percentage of IRAT triggers due to RSCP as opposed to Ec/No.
Checked individual cells with high failures and most of the attempts were
triggered by RSCP and failures by physical channel.


All Rights Reserved Alcatel-Lucent 2008
Conclusion & Recommendation
Compressed mode is starting early w.r.t RSCP and needs to be further
optimized (as mentioned below) w.r.t Ec/No.
Our objective is to start compressed mode attempts when quality is poor and
to reduce when quality is good. This could give the UE sufficient time to
initiate and complete IRAT procedure.
Current setting of InterRATCSThd2DEcNo= -15 in Mumbai. This could be
trialed out as -12 in non core areas on a case by case basis. This will start the
compressed mode early before reaching bad quality and decrease possible
failures due to RF issues. InterRATCSThd2fEcNo could also be modified
accordingly and set to -10.
Current setting of InterRATCSThd2DRSCP =-100. This should be set to -104 to
minimize compressed mode attempts which are still of good quality.
InterRATCSThd2fRSCP should be set to -101.

You might also like