RAN 2494 Fast Cell PCH switching – Test results

April 2013 GS PS NPO radio/ MBB WCDMA Five

Kirsi Teräväinen

Contributors: Tommi Mecklin, Tommi Lassila and Sami Järvinen

Confidential

1 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Introduction

Fast Cell PCH switching feature is tested in mcRNC Five Network 11.4.2013
• Fast Cell PCH switch is RU40 new feature.
• Network SW levels during test (RU40)
• mcRNC SW: R_QNCB.13.5.2_12
• BTS SW: WN 8.0 1301 189 00

Confidential

2 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH switching - Introduction

RAN2494 Fast Cell_PCH switching feature attempts to make faster transition from Cell_PCH, URA_PCH
and Cell_FACH to DCH.
• This is done by message size optimisation during Cell_PCH to Cell_DCH
• RAN2494 Fast Cell_PCH switching Feature is enhancement/improvement over the legacy RAN1232
Fast call setup from Cell_PCH state feature.
• Conditions that which triggers fast call setup functionality are required for RAN2494 feature
functionality as well (CUCforPCHtoDCHallowed, CUCEcNoThreshold and TrafVolThrDLLowGTP
threshold)
• It is also recommended that direct Cell_DCH to Cell_PCH functionality should be turned ON as well.
• This is done to ensure that optimisation of UE message is effective in Cell_DCH to/from Cell_PCH
transitions. (if intermediate FACH state is used, then this optimisation is not effective)

Confidential

3 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH switching - Introduction

Optimisation of RRC messages to UE may not be applicable every time. It would be like best attempt from
RNC perspective. Optimisation can be applied mainly in those cases where newly allocated
configurations is same as previous configuration already stored to UE.
Cases which make it difficult to compress RRC messages to UE
• Multi-RAB scenarios / Active RAB changes
• Channel type changes (DCH<->HSPA)
• E-DCH TTI switch leading to E-DCH transmission grant change
• Fixed RLC <-> Flexible RLC switch
• HSDPA peak rate or UL DCH return channel bit rate changes affecting RLC parameters etc.

Confidential

4 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH switching - UE Message Optimisation
Cell_PCH to Cell_DCH Cell_DCH to Cell_FACH
Cell_FACH to Cell_DCH Cell_PCH to Cell_FACH
RRC Connection Re-establishment
Following IE can be skipped
Following IE can be skipped • RB-mapping-Info
• RB-mapping-Info • UL/DL Transport Channel to Delete List
• RLC-Info Messages affected
• UL/DL Transport Channel Add Reconfigure List • RRC Radio Bearer Reconfiguration Request
Messages affected • RRC Cell Update Confirm
• RRC Radio Bearer Reconfiguration Request
• RRC Cell Update Confirm

Cell_DCH Cell_DCH

Cell_FACH Cell_FACH

Cell_PCH Cell_PCH

IDLE_Mode IDLE_Mode

Confidential

5 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

 Test Cases and Parameters

Two different test cases – both stationary and drive test

• TC1 – Reference Fast PCH Switch Disabled
• TC2 – PCH SW Fast PCH Switch Enabled
• Post verification period was activated same time with FastPCHSwich
• Fast call setup from PCH active in both cases (CUCforPCHtoDCHallowed)

Parameter Object TC1 - Reference TC2 – PCH SW

DCHtoPCHEnabled RNFC Enabled Enabled
CUCforPCHtoDCHallowed RNFC Enabled Enabled
CUCEcNoThreshold WCEL 31 (-9) 31(-9)
CUCRSCPThreshold* WCEL -92 -92
FastPCHSwitchEnabled RNFC Disabled Enabled (for all UE’s)
PostVerifPeriodDLSynch RNFC Disabled Enabled

*WBTS -41 – MBLB active and RACHintrafreqmeasquant = RSCP

Confidential

6 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Fast Call Setup from PCH

Fast Call setup from PCH is active during testing

• Enables direct state transition from Cell_PCH to Cell_DCH
• Direct state transition can be done if
• Fast call setup is activated with CUCforPCHtoDCHallowed (RNFC)
• UE must include trafvol indication in cell update confirm message
• CUCEcNoThreshold (default 31 (-9dB)) or CUCRSCPThreshold
(default -92 dBm) must be good enough to trigger fast call setup
otherwise call setup goes via Cell_FACH

Confidential

7 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH switching – Test setup

• Test is done with two Nemo Handy UE’s

Nemo Handy UE Release HSDPA Capability HSUPA Capability
Nokia C7 Rel 6 Cat 9 Cat 5 (10ms)
Samsung Galaxy S3 Rel 8 Cat 24 (DC HSDPA) Cat 6 (2ms)

• Cases and times below (15min counter data collected)

Case TC1 -Reference TC2 – PCH SW

Stationary 10-10.21 11.30-11.51
Drive 10.22-11.00 12-12.45

Confidential

8 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH switching – Test setup

Test is done using one 1400 byte ping with 8 sec interval, new ping start from Cell_PCH
Samples collected: (around)
• Stationary 150 ping with S3 (around 20min)
• Drive1 150 ping with S3 (around 20min)
• Drive2 150 ping with S3 (around 20min)
Total 150 stationary and 300 drive ping samples

C7 run same test period but collect less samples in each case
There is also some failures.

Confidential

9 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Drive route and stationary location

Picture show drive route and RSCP along the Stationary

route. location
• Same route is repeated 4 times to collect samples.
• Same route (4 times) used for both, TC1 and TC2
• Also stationary location is indicated in the picture
• MBLB and CCH layering state transition features
active in the network

Confidential

10 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RF Conditions in Stationary location – TC1 Reference

In stationary location, the RSCP is

around -75..-85 dBm. Stationary – S3
• There is other drive van in the
area, and it is visible in EcNo
when it enters to same cell
• RTT not very stable, especially
with C7
• 1400byte ping start from Cell
PCH

Stationary – C7
EcNo drop when other
(HSDPA) user(s) enters the
cell – visible with both UE’s
(S3 & C7)

problems
Confidential

11 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RF Conditions in Stationary location – TC2 PCH SW

• As in previous case, also here

is period with low EcNo - > Stationary – S3
other HSDPA traffic in the cell
• There is also few very long
RRT samples
• And some problems with C7 in problems
the end – ping stop earlier

Stationary – C7
EcNo drop when other
(HSDPA) user(s) enters the
cell – visible with both UE’s
(S3 & C7)

Confidential

12 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH
switching - Signalling
S3 Stationary cases
• With fast PCH switch enabled – “Cell
update confirm CCCH” is used instead
of “Cell update confirm” messages.
• In attached excel, there is example of
both messages.

Microsoft Office
Excel Worksheet TC1 Ref Stationary S3 TC2 PCHSW Stationary S3

Confidential

13 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

RAN2494 Fast Cell_PCH
switching - Signalling
Areas with low EcNo – setups go via
Cell_FACH, otherwise Cell_PCH to DCH is
used
1
• Example of both cases in right
1. CPICH EcNo -11 dB is reported in cell
update message (default -9 dB). UE goes
DCH via Cell FACH.
2. CPICH EcNo -9 dB is reported in cell
update message. UE goes to DCH directly
from cell_PCH 2

Confidential

14 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Stationary location Results – S3 NEMO

TC2 show worst results for “Setup via cellUpdate” case – but overall number of samples is low (24) and
there are few samples > 5 sec setup time (Number in the bar is number of samples)
• In all other cases – PCH SW enabled case (TC2) is faster (~ 70 ms)

Cases having direct transition from PCH

Stationary S3
2000
to DCH are calculated separately.
1742
1600
1200 1118 Setup via Cell update contains all other
800
952 883 PCH->DCH cases going via cell update.
528 468
400
429 362 Layer change
0 PCH->DCH Separately cases having layer change –
e H e e
dat DC ng ng Average TC1 stationary test done under cell having
a a
ll Up H-
> ch ch Average TC2
Ce PC er ye
r DRRC/CCH layering active. Only
ms

ia l ay la
v te -
tu
p pda te an
d successful layer changes separated
e llU pda H
S
Ce ll
u DC here
vi
a ce H-
>
a
tu
p vi PC
47 24
e p
92 e116 8 te
da 5 18 3 samples
S tu
S p
l lu
ce
a
vi
p
tu
Confidential se

15 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Drive Results – S3 NEMO

Nemo S3 results from drive – cases changing layer are calculated separately
• TC2 – Fast PCH switch seems to improve setup time in all cases.

DRRC and MBLB state transition

Drive S3
1223
features active – setups with Layer
1200 973 931 854
PCH->DCH change are separated
800 PCH->DCH Layer
572 change
436 490
357
400

0
Average TC1
Average TC2
ms

94 130 202 152 16 37 23 19

Confidential

16 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Stationary and Drive Results – S3 ICSU

ICSU script used to collect S3 results also (not working for C7)
• PCH->HSPA and FACH -> HSPA cases calculated

Stationary S3 Drive S3
500 446 600
426 426 492
400 500 440 416
316
400 327
300
300
200
200
100 TC1 Average 100 TC1 Average
TC2 Average TC2 Average
0 0
ms

ms
c c c c
c h_ c h_ c h_ c h_
d d ed ed
_e _e h_ h_
ch ch c c
ds ds ds ds
hs hs hs hs
to to to to
c c _c _c
c h_ c h_ ch ch
p fa p f a
ell_ l l_ ll _ ll_
c c e ce ce

Confidential

17 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Stationary location Results – C7 NEMO

With C7, results not that clear that S3

• Stationary C7 show both average, and 90% case (90% of samples better than..), this filter worst
samples out
• Layering cases having very few samples 1-6 – not statistically valid

Stationary - C7 Average Stationary - C7 90%

1200 1023 1200 1079 1061
988 945
1000
966
PCH->DCH
853
956
1000
757 741
800 698 800 698
596 641 621 641
600 600
400 400
200 200
0 0
Average TC1 90% TC1
Average TC2 90% TC2
ms

ms
Confidential

18 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Drive Results – C7 NEMO

With C7, there is quite variation in samples.

• Drive test average don’t so clear difference between TC1 and TC2
• 90% case shows that each case there is some improvements in setup time
• Again layering cases show quite few samples

Drive C7 Drive C7 - 90%

1200 989 997 11451089
907 924 1200 990 938
772 732
800 648 624 746 734 772 732
800
400 400
0 0
Average TC1 90% TC1
Average TC2 90% TC2

ms
ms

Confidential

19 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

State transition times – counter data – RNC35

FastPCHSwitch Enabled
State transition times FACH->DCH and PCH->DCH (mcRNC35)
11 – 14.30 hours
600 FastPCHSwitchEnabled =enabled
• RNC1289a State
500 transition time from PCH
to DCH decrease.
400 • Slight decrease in FACH
-> DCH times (?)
300 Reference case
PCHSW case

200 Attempts generated by other test

van/lab UE’s under mcRNC35
100

AM A M A M A M A M A M A M A M AM A M AM A M PM P M PM PM P M P M P M P M P M P M P M P M P M P M P M PM
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00: 15: 30: 45: 00: 15: 30: 45: 00: 15: 30: 45: 00: 15: 30: 45: 00: 15: 30: 45: 00: 15: 30: 45: 00: 15: 30: 45:
: : : : : : : : : : : : : : : : : : : : : : : : : : : :
09 09 09 09 10 10 10 10 11 11 11 11 12 12 12 12 01 01 01 01 02 02 02 02 03 03 03 03

Att PCH to DCH M1006C229 Cell-FACH to Cell-DCH RNC_1288A

Cell/URA-PCH to Cell-DCH RNC_1289A

Confidential

20 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

State transition times – counter data – RNC35, BTS40

Stationary test done state transiton times Cell_FACH to DCH and Cell_PCH to DHC - BTS 40
under BTS40 700
FastPCHSwitchEnabled =enabled

600
TC1_Stationary
• 10-10.21 500

TC2_Stationary 400

• 11.30-11.52 300
Att PCH to DCH
Cell-FACH to Cell-DCH
Cell/URA-PCH to Cell-DCH

200

100

AM AM A M AM AM AM A M AM AM PM P M P M PM P M P M P M P M P M P M P M
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
5: 0: 5: 0: 5: 0: 5: 0: 5: 0: 5: 0: 5: 0: 5: 0: 5: 0: 5: 0:
9:4 0:0 0:1 0:3 0:4 1:0 1:1 1:3 1:4 2:0 2:1 2:3 2:4 1:0 1:1 1:3 1:4 2:0 2:1 2:3
0 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0

Confidential

21 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013

Summary

In our test case – the Fast PCH switch impact setup times.
• Impact is more visible with S3 than C7
• Can be seen also in counter data

Confidential

22 © Nokia Siemens Networks 2012 GS PS NPO radio/ April 2013