Lte practical questions II
SPECIFIC ISSUES IN REAL PROJECT
… PREVIOUSLY
› LTE practical questions:
1. BW increase vs. Output Power
2. cellRange vs. rachRootSequence
3. Latency vs. PRB Utilization
[Link]
Ericsson Internal | 2016-05-06 | Page 2
PRACTICAL questions II
1. Accessibility degradation vs. new L800 site
2. UL RSSI vs. TMA configuration
3. Low % of MIMO usage
Ericsson Internal | 2016-05-06 | Page 3
PRACTICAL questions II
1. Accessibility degradation vs. new L800 site
2. UL RSSI vs. TMA configuration
3. Low % of MIMO usage
Ericsson Internal | 2016-05-06 | Page 4
Accessibility degradation vs. new L800 sITE
› New LTE 800MHz is deployed in an area mainly covered by LTE 1800MHz
New L800 site
Ericsson Internal | 2016-05-06 | Page 5
Accessibility degradation vs. new L800 sITE
Accessibitliy
› When the LTE800 site is set degradation in
on-air, a clear degradation in nearby L1800
Accessibitliy of the nearby
L1800 sites appears
› Question: What’s the reason
behind this behavior?
L800 on-air
Ericsson Internal | 2016-05-06 | Page 6
Accessibility degradation vs. new L800 SITE
Why did L1800 Accessibility decrease when launching nearby L800?
› Product of inter modulation (PIM) between L800 and L1800 cells
› Increase in the L1800 UL RSSI due to new L800 traffic
› Increase in the number of Tracking Area Updates (TAU)
Ericsson Internal | 2016-05-06 | Page 7
Accessibility degradation vs. new L800 sITE
ANSWERS
› How are we measuring Accessibility? CSSR = 100 * RRC * ERAB
CSSR = 100 *
((pmRrcConnEstabSuccMta + pmRrcConnEstabSuccMod + pmRrcConnEstabSuccMos + pmRrcConnEstabSuccHpa + pmRrcConnEstabSuccEm) /
(pmRrcConnEstabAttEm + pmRrcConnEstabAttHpa + pmRrcConnEstabAttMod + pmRrcConnEstabAttMos + pmRrcConnEstabAttMta
- pmRrcConnEstabAttReattEm – pmRrcConnEstabAttReattHpa – pmRrcConnEstabAttReattMod – pmRrcConnEstabAttReattMos - pmRrcConnEstabAttReattMta))
*
(pmErabEstabSuccInit / pmErabEstabAttInit)
ERAB RRC
(*) Note: S1 part not being considered here
Ericsson Internal | 2016-05-06 | Page 8
Accessibility degradation vs. new L800 sITE
› RRC part, splitted by
type of connection
ATTEMPTS
SUCCESSES
RE-ATTEMPTS
Ericsson Internal | 2016-05-06 | Page 9
Accessibility degradation vs. new L800 sITE
› In this case, the main cause was “Mobile Originating Signaling”
pmRrcConnEstabAttEm: “Emergency”
pmRrcConnEstabAttHpa: “High Priority Access”
pmRrcConnEstabAttMod: “Mobile Originating Data”
pmRrcConnEstabAttMos: “Mobile Originating Signaling”
pmRrcConnEstabAttMta: “Mobile Terminating Access”
Neighbor L800 on-air
Ericsson Internal | 2016-05-06 | Page 10
Accessibility degradation vs. new L800 sITE
› New L800 site was deployed in a
Tracking Area border
› Propagation in L800 is larger, so
coverage of L800 cells
(TAC=10130) is overlapping with
L1800
L1800 coverage (TAC=10230) accessibility
degradation
› Two different TAC covering the
same physical area: high
increase in the Tracking Area
Update (TAU) intensity, when UE
moves between layer L800 and
L1800 New L800
on-air
Ericsson Internal | 2016-05-06 | Page 11
Accessibility degradation vs. new L800 sITE
› After changing the TAC of new L800 site, Trackig Area Update procedure (TAU)
reduces a lot (pmRrcConnEstabAttMos is lower) and Accessibility of neighbor
L1800 cells recovers
RRC signaling (pmRrcConnEstabAttMos) for neighbor L1800 cells before and after TAC change Accessibility
TAC change
TAC change
Ericsson Internal | 2016-05-06 | Page 12
PRACTICAL questions II
1. Accesibility degradation vs. new L800 sector
2. UL RSSI vs. TMA configuration
3. Low % of MIMO usage
Ericsson Internal | 2016-05-06 | Page 13
UL RSSI vs. TMA configuration
› LTE-1800 sector is sharing TMA (Tower Mounted Antenna
Amplifier) and Antenna with WCDMA-2100
› TMA is configured and controlled by 3G NodeB, but
it is not denifed in LTE TMA
Diplexer
WCDMA-2100 LTE-1800
RU RU
› Question: What is the impact on LTE UL RSSI KPI?
Ericsson Internal | 2016-05-06 | Page 14
UL RSSI vs. TMA configuration
What is the impact on LTE UL RSSI KPI due to not having TMA configured in LTE?
› Higher UL RSSI in LTE KPI due to not having TMA configured
› Lower UL RSSI in LTE KPI due to not having TMA configured
› No implact, same UL RSSI regardless TMA configuration
Ericsson Internal | 2016-05-06 | Page 15
UL RSSI vs. TMA configuration
ANSWERS
› UL RSSI KPI in LTE is measured using the counter pmRadioRecInterferencePwr
Details of the measurement
in 3gpp specification 36.214
Ericsson Internal | 2016-05-06 | Page 16
UL RSSI vs. TMA configuration
ANSWERS
› The reference point for the measurement must be the RX antenna connector
3gpp TS 36.214, pag. 11
Ericsson Internal | 2016-05-06 | Page 17
UL RSSI vs. TMA configuration
ANSWERS
› If TMA is not defined, pmRadioRecInterferencePwr is not considering the correct
reference point for the measurement
Antenna Antenna
pmRadioRecInterferencePwr
reference point when TMA is
TMA defined TMA
pmRadioRecInterferencePwr
reference point when TMA is
LTE-1800 LTE-1800
not defined RU RU
Ericsson Internal | 2016-05-06 | Page 18
UL RSSI vs. TMA configuration
ANSWERS
Example:
› This is not affecting signal quality, because TMA
in actually installed in the real world and working
properly (managed by 3G network)
› Common KPIs are OK: CSSR, DCR, Thpt, etc.. Antenna
Ul RSSI = - 110 dBm
› But UL RSSI KPI is showing higher values TMA
UL Gain =
14 dB
compared to other sites, because UL Gain of the
Ul RSSI = - 96 dBm
TMA is not being taken into account
LTE-1800
RU
Value reported when TMA is not defined
Ericsson Internal | 2016-05-06 | Page 19
UL RSSI vs. TMA configuration
ANSWERS
› TMA can only be controlled by one system (3G or LTE)
› Solution: If we want to keep the TMA being controlled by 3G, we can “simulate”
the TMA in LTE using the dlAttenuation and ulAttenuation parameters:
Ericsson Internal | 2016-05-06 | Page 20
UL RSSI vs. TMA configuration
ANSWERS
› For example, if we have 2 dB of feeder losses, and a TMA with 12
dB of UL gain, we can set dlAttenuation and ulAttenuation this way:
– dlAttenuation = 2 dB
– ulAttenuation = -10 dB
i) The unit is 0.1 dB
ii) Each value of the array represent a 5-MHz interval of the operating band
Ericsson Internal | 2016-05-06 | Page 21
UL RSSI vs. TMA configuration
ANSWERS
› After setting dlAttenuation = 20 and ulAttenuation = -100 in our network, UL
RSSI changed from -109 dBm to -119 dBm
Ericsson Internal | 2016-05-06 | Page 22
PRACTICAL questions II
1. Accesibility degradation vs. new L800 sector
2. UL RSSI vs. TMA configuration
3. Low % of MIMO usage
Ericsson Internal | 2016-05-06 | Page 23
Low % of MIMO usage
› Percentage of MIMO usage is a good indicator to check the radio conditions
of an LTE cell
› There should be a correlation
between % of MIMO
utilization, and CQI reported
by UEs
Cells with some kind of
problem: low MIMO
utilization despite good CQI
Scatter plot with daily average % of MIMO utilization and daily average reported CQI. Data from counters (each dot is a cell)
Ericsson Internal | 2016-05-06 | Page 24
Low % of MIMO usage
› Formulas
Percentage_MIMO_usage = 100 * pmRadioUeRepRankDistr[1]/(pmRadioUeRepRankDistr[0] + pmRadioUeRepRankDistr[1])
MO class Counter Counter Type Counter Size Description
The reported rank distribution indicates the rank of UEs that are in an Open
and Closed Loop Spatial Multiplexing mode, where rank is reported.
PDF ranges:
EUtranCellFDD pmRadioUeRepRankDistr ACC PDF[4]
[0]: Rank = 1
[1]: Rank = 2
[2]: Rank = 3
[3]: Rank = 4
MO class Counter Counter Type Counter Size Description
The transmission mode and rank distributions provide more detailed
information on the extent to which each transmission mode and rank
is used.
PDF ranges:
› More detailed distribution can be obtained [0]: Transmit diversity
[1]: Open Loop SM Rank 1
using pmRadioTxRankDistr: [2]: Open Loop SM Rank 2
[3]: Closed Loop SM rank 1
EUtranCellFDD pmRadioTxRankDistr ACC PDF[14] [4]: Closed Loop SM rank 2
[5]: Reserved
[6]: Reserved
[7]: Reserved
[8]: Closed Loop SM rank 3
[9]: Closed Loop SM rank 4
[10]: Multiple Layer Transmission rank 1
[11]: Multiple Layer Transmission rank 2
[12]: Open Loop SM rank 3
[13]: Open Loop SM rank 4
Ericsson Internal | 2016-05-06 | Page 25
Low % of MIMO usage
› Formulas
4G_Average_Reported_CQI = (1*pmRadioUeRepCqiDistr[1] + 2*pmRadioUeRepCqiDistr[2] + 3*pmRadioUeRepCqiDistr[3] +
4*pmRadioUeRepCqiDistr[4] + 5*pmRadioUeRepCqiDistr[5] + 6*pmRadioUeRepCqiDistr[6] + 7*pmRadioUeRepCqiDistr[7] +
8*pmRadioUeRepCqiDistr[8] + 9*pmRadioUeRepCqiDistr[9] + 10*pmRadioUeRepCqiDistr[10] + 11*pmRadioUeRepCqiDistr[11] +
12*pmRadioUeRepCqiDistr[12] + 13*pmRadioUeRepCqiDistr[13] + 14*pmRadioUeRepCqiDistr[14] +
15*pmRadioUeRepCqiDistr[15])
/
(pmRadioUeRepCqiDistr[0] + pmRadioUeRepCqiDistr[1] + pmRadioUeRepCqiDistr[2] + pmRadioUeRepCqiDistr[3] +
pmRadioUeRepCqiDistr[4] + pmRadioUeRepCqiDistr[5] + pmRadioUeRepCqiDistr[6] + pmRadioUeRepCqiDistr[7] +
pmRadioUeRepCqiDistr[8] + pmRadioUeRepCqiDistr[9] + pmRadioUeRepCqiDistr[10] + pmRadioUeRepCqiDistr[11] +
pmRadioUeRepCqiDistr[12] + pmRadioUeRepCqiDistr[13] + pmRadioUeRepCqiDistr[14] + pmRadioUeRepCqiDistr[15])
Ericsson Internal | 2016-05-06 | Page 26
Low % of MIMO usage
EXAMPLE 1. L800 sector, 10MHz BW
› CQI around 10-13, but % MIMO utilization less than 1%.
Ericsson Internal | 2016-05-06 | Page 27
Low % of MIMO usage
EXAMPLE 1. L800 sector, 10MHz BW
› How many antennas are configured?
Two antennas configured
› How many are used?
noOfUsedTxAntennas and noOfUsedRxAntennas
Internal parameter, parse complete MOM
Two antennas used
Ericsson Internal | 2016-05-06 | Page 28
Low % of MIMO usage
EXAMPLE 1. L800 sector, 10MHz BW
› What about VSWR (Voltage Standing Wave Ratio)?
RF A port in SECTOR 1
is not ok (more than 1.5
VSWR is a problem)
RF A port in SECTOR 2
is also bad
Ericsson Internal | 2016-05-06 | Page 29
Low % of MIMO usage
EXAMPLE 1. L800 sector, 10MHz BW
› SECTOR 3 is ok
› SECTOR 1 and SECTOR 2 are not
ok, due to installation problem
Ericsson Internal | 2016-05-06 | Page 30
Low % of MIMO usage
EXAMPLE 2. L2600 sector, 20MHz BW
› CQI around 10-13, but % MIMO utilization less than 1% in general.
Ericsson Internal | 2016-05-06 | Page 31
Low % of MIMO usage
EXAMPLE 2. L2600 sector, 20MHz BW
› Two antennas configured and used
› VSWR is ok
Ericsson Internal | 2016-05-06 | Page 32
Low % of MIMO usage
EXAMPLE 2. L2600 sector, 20MHz BW
› Site covering a football stadium. No matches in this period, only intermittent
traffic captured from outside
Some peaks around 25-30% of MIMO utilization
if we look at ROP (15 min) level
› So no MIMO problem in this case
Ericsson Internal | 2016-05-06 | Page 33
SUMMARY
1. Accessibility degradation vs. new L800 sector
› Overlapping Tracking Areas among different layers can cause accessibility
degradation. Review TAC borders
2. UL RSSI vs. TMA configuration
› If TMA exists but it is not configured in the eNB, UL RSSI values are not
correctly measured. Define it, or simulate it with negative ulAttenuation
3. Low % of MIMO usage
› Low % of MIMO utilization can be used as an indicator of installation problems
Ericsson Internal | 2016-05-06 | Page 34