UMTS Power Control

i
Contents
1 Uplink Open Loop Power Control of R99................................................................................................ 1
1.1 Uplink open loop power control of R99 common channel ............................................................... 1
1.2 Uplink open loop power control of R99 dedicated channel .............................................................. 2
1.3 Uplink Open Loop Power Control of R99 in Compressed Mode ..................................................... 8
2 Downlink Open Loop Power Control of R99 ........................................................................................... 9
2.1 Configuration of R99 downlink common channel initial power ....................................................... 9
2.2 Downlink open loop power control of R99 dedicated channel ......................................................... 9
2.3 Downlink open loop power control of R99 in compressed mode ................................................... 13
3 Uplink inner loop power control of R99 ................................................................................................. 15
3.1 Uplink inner loop power control of R99 in normal mode ............................................................... 15
3.2 Uplink inner loop power control of R99 in compressed mode ....................................................... 17
4 Downlink Inner Loop Power Control Of R99 ....................................................................................... 21
4.1 Downlink inner loop power control of R99 in normal mode .......................................................... 21
4.2 Downlink inner loop power control in compressed mode............................................................... 22
5 Uplink Outer Loop Power Control of R99 ............................................................................................. 25
6 Downlink Outer Loop Power Control of R99 ........................................................................................ 27
7 Downlink Power Balancing ..................................................................................................................... 29
7.1 AlgorithmDescription..................................................................................................................... 29
7.2 Related measurement ...................................................................................................................... 31



1
1 Upl ink Open Loop Power Control of R99
1.1 Uplink open loop power control of R99 common channel
Algorithm
The uplink open loop power control of common channel mainly refers to determining
the PRACH transmit power.
In the FDD mode, UE performs the following operations before it transmits signals to
PRACH.
1. UE obtains Primary CPICH DL TX Power and Constant Value fromSystem
information Block type 6 (or Systeminformation Block type 5 if type 6 is not
broadcast). UE obtains UL Interference fromSysteminformation Block type 7.
2. UE measures and obtains CPICH_RSCP, the channel code power of CPICH.
3. UE calculates the transmit power of the first prefix using the following formula:
Preamble_Initial_Power = Primary CPICH DL TX power CPICH_RSCP +UL
interference +Constant Value
Where,
Primary CPICH DL TX power (PcpichPwr) is the transmit power of the main
pilot channel.
UL interference is the uplink interference, which is measured and obtained by
NodeB and updated in real time in SIB7.
Constant Value (ConstVal) is a value related with the cell environment. It is a
value depends on the service rate and quality carried by PRACH.
If parameters in the systems broadcast information changes, UE calculates the initial
transmit power again and submit the result to the physical layer.
When the physical randomaccess process gets started, UE sets the preamble transmit
power as Preamble_Initial_Power. If the value of Preamble_Initial_Power exceeds the
allowed maximumpower MaxRACHTxPwr, UE sets the preamble transmit power as
the allowed maximumpower. If no response (+1 or -1) of AICH is received after the
preamble composed of selected signature and scramble is sent out, PRACH selects a
UMTS Power Control

2
new signature in the next timeslot, uses it to forma preamble together with the
scramble and sends the preamble again. Next, PRACH increases the preamble transmit
power by Power Ramp Step[dB]. If the transmission counter is 0, the access process is
existed. If the positive response is received fromAICH, the randomaccess message is
transmitted. The power of the control part of the randomaccess message is the last
transmit power of the preamble plus the offset P p-m[dB].
Fromthe preceding description we get to the formula for calculating the transmit
power of the control part of the PRACH message:
m - p P Step Ramp Power er nitial_Pow Preamble_I wer PRACH_C_Po + + =


Where,
Power Ramp Step (PRStep) is the power offset between two continuous
preambles.
Pp_m (POPpm[MAX_ PRACH _TFC]) is the power offset between the control
channel and the last preamble of the message part.
In addition, RACH is similar to the uplink DPCH. That is, its data domain and control
domain are sent out after being multiplexed with I and Q channels and then added by
scramble on the physical layer. Therefore, parameters
c
and
d
(gain factor of the
control channel and data channel of the message part) also need to be determined.
Either
c
(BetaC[MAX_ PRACH _TFC]) or
d
(BetaD[MAX_ PRACH _TFC]) is
15.
UL interference can be updated in SIB 7 in two ways, which can be selected with the
parameter SIB7Originator.
If SIB7Originator takes the value of RNC, NodeB reports the common
measurement report of RTWP to RNC. When RNC detects that the change of
RTWP is no less than the uplink interference update threshod (UlInterUpdtTh),
it broadcasts it to UE through the broadcast channel.
If SIB7Originator takes NodeB, NodeB updates the UL interference directly in
the system message based on the change of RTWP.
1.2 Uplink open loop power control of R99 dedicated channel
Uplink open loop power control of the dedicated channel refers to determining the
initial transmit power of DPCCH, and determining the gain factor
c

and
d

of the
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
3
uplink control physical channel and uplink data physical channel.
1. Power configuration of DPCCH:
As required by related standard, UE should start uplink inner loop power control
according to the following power level when the first DPCCH is being set up:
DPCCH_Initial_power =DPCCH_Power_offset - CPICH_RSCP
Where,
The value of DPCCH_Power_offset is determined by DPCCH open loop
power control method.
The value of CPICH_RSCP is the CPICH channel code power obtained by
UE through measurement.
DPCCH_Power_offset is calculated using the following formula:
Power TX CPICH dB PG dBm I N dB N E Offset Power DPCCH
T T b
_ _ ) ( ) )( ( ) ( / _ _
0
+ + + =

Where,
Eb/No is the quality factor of the DPCCH PILOT domain (DpcchPilotEbN0).
NT+IT is the uplink interference, which is obtained by NodeB through
measurement and updated in real time in SIB7.
PG is the spectrum spread gain, 256.
CPICH_TX_Power(PcpichPwr) is the transmit power of the P-CPICH.
Description: The quality factor (DpcchPilotEbN0) of the DPCCH PILOT domain
depends on the diversity mode and sub-types of service. To obtain DpcchPilotEbN0,
first obtain the Diversity PC Index (DivPcIndex (Utran Cell)) fromthe configuration
items of Utran Cell. Next, in Power Control Related to Service and Diversity Mode,
query DivPcIndex (Power Control Related to Service and Diversity Mode), diversity
mode (TxDivMod) and sub-service type (SrvType). All parameters related to power
control, if depending on the diversity mode and sub-service, can be obtained in this
way.
2. How |c and |d (gain factor of uplink control / data physical channel) are
determined and configured:
UMTS Power Control

4
Different strategies are adopted depending on the features of a single service and mixed
services:
For a single service, |
c
and |
d
are configured directly according to different service
rates and different TFCs. It is usually required that either |
c
or |
d
must be 15. 0~0 list
the |
c
and |
d
values configured in RNC for several common services in the case of
different TFC formats. In the tables, 1144 is the format of signaling transmission.
Table 1-1 |
c
and |
d
Values for the UL WAMR6.60k~23.85k Service
TFC Format
c

d

040, 0405, 00, 0144 15 1
140, 0405, 00, 0144 15 8
154, 178, 00, 0144 15 11
164, 1113, 00, 0144 15 13
172, 1181, 00, 0144 15 15
172, 1213, 00, 0144 14 15
172, 1245, 00, 0144 14 15
172, 1293, 00, 0144 13 15
172, 1325, 00, 0144 12 15
172, 1389, 00, 0144 12 15
172, 1405, 00, 0144 11 15
040, 0405, 00, 1144 15 8
140, 0405, 00, 1144 15 11
154, 178, 00, 1144 15 14
164, 1113, 00, 1144 15 15
172, 1181, 00, 1144 13 15
172, 1213, 00, 1144 13 15
172, 1245, 00, 1144 12 15
172, 1293, 00, 1144 12 15
172, 1325, 00, 1144 11 15
172, 1389, 00, 1144 11 15
172, 1405, 00, 1144 11 15
Table 1-2 |
c
and |
d
Values for the UL NAMR4.75k~12.2k Service
TFC Format c

d

039, 0103, 060, 0144 15 1
139, 0103, 060, 0144 15 9
142, 153, 060, 0144 15 12
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
5
TFC Format
c

d

149, 154, 060, 0144 15 12
155, 163, 060, 0144 15 13
158, 176, 060, 0144 15 14
161, 187, 060, 0144 15 14
175, 184, 060, 0144 15 15
165, 199, 140, 0144 14 15
181, 1103, 160, 0144 13 15
039, 0103, 060, 1144 15 12
139, 0103, 060, 1144 15 14
142, 153, 060, 1144 14 15
149, 154, 060, 1144 13 15
155, 163, 060, 1144 13 15
158, 176, 060, 1144 12 15
161, 187, 060, 1144 12 15
175, 184, 060, 1144 12 15
165, 199, 140, 1144 11 15
181, 1103, 160, 1144 11 15
Table 1-3 |
c
and |
d
Values for the UL PS64k streaming/interactive/background Service
TFC Format c

d

0336, 0144 15 1
1336, 0144 15 14
2336, 0144 11 15
4336, 0144 8 15
0336, 1144 15 8
1336, 1144 14 15
2336, 1144 10 15
4336, 1144 8 15
Table 1-4 |
c
and |
d
Values for the UL PS128k streaming/interactive/background Service
TFC Format
c

d

0336, 0144 15 1
1336, 0144 15 14
2336, 0144 11 15
4336, 0144 8 15
8336, 0144 6 15
UMTS Power Control

6
TFC Format
c

d

0336, 1144 15 9
1336, 1144 14 15
2336, 1144 10 15
4336, 1144 8 15
8336, 1144 6 15
Table 1-5 |
c
and |
d
Values for the UL PS384k and services with higher rates streaming/
interactive/background Service
TFC Format
c

d

0336, 0144 15 1
1336, 0144 11 15
2336, 0144 8 15
4336, 0144 8 15
8336, 0144 6 15
12336, 0144 5 15
0336, 1144 15 8
1336, 1144 10 15
2336, 1144 8 15
4336, 1144 8 15
8336, 1144 6 15
12336, 1144 5 15
For mixed service, the |c and |d values are calculated by RNC and configured for UE.
3. Calculation of |c and |d for mixed services
Suppose to mix service A and service B (including signaling).
(1) For the transmission combination (TFC
multi
) when service A and B are being
combined, calculate the number of bits per frame mapped to the transport channel by
each service according to the transmission format indication (TFI_i) of each service
corresponding to TFC
multi
.
(2) Select the service with the most bits transmitted by a frame as the reference service,
and the service corresponding to the TFI that is corresponding to TFC
multi
as the
reference service.
(3) Take |
c
and |
d
corresponding to TFC
single
(formed by TFI_i of the selected reference
service) as the reference |
c
and |
d.
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
7
(4) Use the following formula to calculate and obtain the |
c
and |
d
corresponding to
this TFC
multi.
ref
j
j
ref
ref c
ref d
j
K
K
L
L
A =
,
,

If A
j
>1, make 0 . 1
,
=
j d
,
j c,
is of the maximum quantified value that
satisfies the condition of
j c,
s 1 / A
j
. Note: If
j c,
=0 is obtained, then make
j c,
=1/15.
If A
j
s 1, then
j d ,
is of the minimum quantified value that satisfies the
condition of
j d ,
> A
j
, while 0 . 1
,
=
j c
.
Where,

c,ref
and
d,ref
are the gain factors corresponding to the above mentioned
TFC
single
;
c,j
and
d,j
are the gain factor corresponding to the j
th
type of
TFC
multi
.
L
ref
is the number of dedicated physical channels needed by the number of
bits to be sent out in the case of TFC
single
.
L
,j
is the number of dedicated physical channels needed by the number of bits
to be sent out in the case of TFC
multi
.

=
i
i i ref
N RM K
Where: RM
i
is the semi-static rate matching factor of transport channel i in the TFC
single

combination; N
i
is the number of bits mapped fromtransport channel i to a radio frame
before rate matching is performed; refers to sum up all transport channels in the
TFC.

=
i
i i j
N RM K
Parameters in this formula take the same meanings as those in the previous formula.
But refers to sum up all transport channels in the TFC
j
(TFC
multi
).
4. To ensure that the power of the data channel reaches the required value before
data is transmitted, the power control preamble is sent before data transmission on
the uplink dedicated channel. In addition, closed loop power control is already
being performed while the power control preamble is sent out. Length of the
UMTS Power Control

8
preamble depends on DpcchPcpLen. At the same time, while starts to send uplink
DPDCH data, no signaling ranging RB0~RB4 is included in the first several
frames. The number of delayed signaling frames depends on SrbDelay.
1.3 Uplink Open Loop Power Control of R99 in Compressed Mode
The gain factors
c,C,j
and
d,C,j
corresponding to a certain TFC used by the compressed
frame in the compressed mode are obtained from|
c
and |
d
used by radio frames in
normal mode. The formula for calculation is as follows:
N pilot C slots
C pilot
j j C
N N
N
A A
, ,
,
,
15

=
Where,
A
j
is the ratio of |
d
and |
c
in normal mode.
A
C,j
is the ratio of
d,C
and
c,C
in compressed mode.
N
pilot,C
is the number of pilot bits per timeslot in the compressed frame in
compressed mode
N
pilot,N
is the number of pilot bits per timeslot in normal mode.
N
slots,C
is the number of timeslots used for data sending in the compressed
frame in compressed mode.
A
C,j
is obtained with the previous formula when the current frame is compressed. The
following rules is then used to obtain the values of
d,c,j
and
c,C,j
.
If A
C,j
>1, then 0 . 1
, ,
=
j C d
,
j C c , ,
is of the maximum quantified value
that satisfies the condition
j C c , ,
s 1 / A
j
. Note: If
j C c , ,
=0 is obtained,
then make
j C c , ,
=1/15.
If A
C,j
s 1, then
j d ,
is of the minimum quantified value that satisfies the
condition
j d ,
> A
j
, while 0 . 1
,
=
j c
.


9
2 Downli nk Open Loop Power Control of
R99
2.1 Configuration of R99 downlink common channel initial power
In the downlink direction, the initial transmit power of P-CPICH, S-CPICH, P-CCPCH,
SCH, AICH, PICH and S-CCPCH should be configured. These channels are downlink
common physical channels.
The transmit power (PcpichPwr) of P-CPICH depends on the proportion of maximum
transmit power of a cell. The values of P-CCPCH, P-SCH (PschPwr), S-SCH
(SschPwr), AICH (AichPwr), and PICH (PichPwr) depend on the offset to P-CPICH
(PcpichPwr). BCH is mapped one-to-one to P-CCPCH physical channel. P-CCPCH power is same
with the power of BCH (BchPwr). Presently, S-CPICH power (ScpichPwr) S-CPICH
power (ScpichPwr) is based on the coverage of MIMO cell.
As the physical channel S-CCPCH bears the transport channel of PCH and FACH, and
the number of FACH channels beared is variable, the transmit power (PchPwr) of each
PCH and the maximumtransmit power (MaxFachPwr) allowed for each FACH beared
by a certain S-CCPCH is specified in related protocol. As the data rate of PCH is
invariable, the transmit power of PCH is determined by the fixed rate. The transmit
power of FACH is determined by the maximumdata rate beared by this FACH. For
different rates, the transmit power can be measured in actual environment. The transmit
power of the data domain of S-CCPCH depends on the PCH transmit power and the
maximum value of the maximumtransmit power of FACH beared on S-CCPCH. The
transmit power of the TFCI domain and Pilot domain of S-CCPCH are indicated
respectively by the offsets (PO1 and PO3) as opposed to the transmit power of the data
domain.
2.2 Downlink open loop power control of R99 dedicated channel
The transmit power of downlink dedicated physical channel is related with the load of
cell, interference, path losses, and rate of bearer service. In related protocol, the initial
transmit power of a specific dedicated channel is for physical channel. Therefore, the
initial transmit power should be calculated separately for single services and mixed
UMTS Power Control

10
services. At present, the estimation power algorithmbased on CPICH Ec/N0 is adopted
for calculating the initial transmit power.
1. Initial power of downlink dedicated channel
When a subscriber is accessing or being handover, and a downlink dedicated physical
channel should be set up for this subscriber. RNC should configure the downlink initial
transmit power for NodeB. The strategy for configuring the initial transmit power of
downlink dedicated channel affects the performance of links and capacity of the
system.
The following formula is used to calculate the initial transmit power:
t PowerOffse
10 1
,
k2
1 k
min max
min
0
,
,
+
(
(
(

|
|
|
.
|

\
|
+

+ =

total tx
L
cpich c
CPICH tx
init tx
P
k
N
E
P
PG
P

Where,
PG is the service processing gain, that is, W/R, W being 3.84M while R being
the bit rate of the service.

CPICH tx
P
,
is transmission power of the CPICH (dBm).

0
N
E
cpich c
is CPICH Ec/N0(dB) reported by the UE.

min is the lower limit of the downlink orthogonal factor (MinOrthogFactor).

max is the upper limit of the downlink orthogonal factor (MaxOrthogFactor).

k is the coefficient factor. Its fixed value is 0.01.
L represents path loss. L

is obtained from the measurement result reported by
the UE. If L

cannot be obtained from the measurement result, its value is
130dB.
The following table describes the rule for obtaining L fromthe measurement result
reported fromthe UE.
If the measurement quantity of the UE is Pathloss, L=Value
pathloss
.
If the measurement quantity of the UE is RSCP, L= P
PCPICH
- Value
RSCP
. The P
PCPICH

Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
11
refers to the transmit power of the PCPICH.

k1 and k2 are scenario parameters. The values of parameters k1 and k2 vary
with the specific scenarios, including densely-populated urban area, suburban
area, rural area.
Dense Urban Urban Suburb Rural
K1=-32.9116 K1=-53.5116 K1=-51.1716 K1=-48.8116
K2=-33.5849 K2=-25.8549 K2=-22.8249 K2=-21.5249
P
tx,total
is the total transmit power of a cell before a subscriber accesses the
cell. It is obtained from the common measurement report: TCP- Transmitted
Carrier Power. Note: for HS cell, P
tx,total
is the valid load of TCP, and obtained
through Node B common measurement report of HS-DSCH Required Power
and Transmitted carrier power of all codes not used for HS
(
MaxSpi
Spi
Spi 0
NOHSDSCHPower HSDSCHRequiredPower
=
+

).
=10^((Eb/N0)/10), where Eb/No is the Eb/No of the sub-service configured
corresponding to the current rate of the access service. Typical values of
Eb/No are:
Traffic Class Data Rate Downlink Traffic Eb/N0 (dB)
Conversational DL WAMR6.60k~23.85k 5.1
Conversational DL 64K(PS Conversational Video) 5.2
Conversational DL NAMR4.75k~12.2k 5.1
Streaming PS64k 1.7
Streaming PS384k 0.9
Streaming PS128k 0.9
Interactive PS64k 4.8
Interactive PS384k 0.9
Interactive PS128k 0.9
Background PS64k 1.7
Background PS384k 0.9
Background PS128k 0.9
Streaming CS64k 1.7
Interactive PS8k 6.9
Background PS8k 6.9
PowerOffset is different for different situation as following:
i. Add RL in SRNC.
Situation of Adding RL in SRNC PowerOffset
UMTS Power Control

12
Call Setup POSetup
Soft or Softer Handover POSoftHO
RAB Hard Handover (RAB on DCH)
PORabHardHO
Incoming Relocation (RAB on DCH)
CELL_FACH or CELL_PCH Transfer to CELL_DCH state (RAB on
DCH)
Hard Handover for only SRB on DPDCH
POSrbHardHO
Hard Handover for F-DPCH
CELL_FACH or CELL_PCH Transfer to CELL_DCH state (RAB on
HS-DSCH)
Call Re-Establishment POReEstablish
ii. Add RL in DRNC.
Situation of Adding RL in DRNC PowerOffset
Add RL with RAB on DPDCH
POSoftHO
Add RL for F-DPCH transmission and without service on this RL
Add RL with SRB on DPDCH
POSrbHardHO
Add RL for F-DPCH transmission and with service on this RL
iii. For radio link reconfiguration, if need to calculate the initial power again, the
PowerOffset is 0.
To avoid too large power occupation of the dedicated channel, the maximum
(MaxDlDpchPwr) and minimum(MinDlDpchPwr) values of DPCH are specified in
3GPP protocol. In order to show the differentiation of different basic priority users, the
actual maximum allowed downlink DPCH transmission power is configured
distinctively for different users with different basic priority.
To calculate the transmit power of downlink dedicated physical channel for mixed
services, first obtain the transmit power (data part)
j
DPCH_POWER needed to transmit
each service with the calculation method used for single service. Next calculate the
initial transmit power of DPCH for mixed services based on the transmit power needed
for each service. The formula is given as follows:

=
=
N
j
j
Power DPCH Power Initial DPCH
1
_ _ _
On the DPCH, the bits of TFCI, TPC and PILOT are also multiplexed besides the data
bits because the information carried by these bits is important. Therefore, the needed
power is also a little higher than that of the data domain. The power value depends on
the offset as opposed to the power of the data domain and is indicated with PO1
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
13
(DpchPO1), PO2 (DpchPO2) and PO3(DpchPO3) respectively.
In the condition that the dynamic update PO2 switch (DynaUpdtPO2Stch) turned on,
dynamic update the PO2 as following: Get the PO2(DpchPO2)value based on DPCH
data rate and traffic class, if DPCH bears multi-services, get the DpchPO2 value
respectively for each service, and then get the minimumvalue. And then the DpchPO2
value is send to Node B through control frame. If the DPCH data rate is changed, then
get the new PO2, and send the new PO2 to Node B. In this way, PO2 dynamic update
is achieved.
The parameters involved in this section, such as MaxDlDpchPwr, MinDlDpchPwr,
DpchPO1, DpchPO2 and DpchPO3, are related with the diversity mode and
sub-service types. To obtain the parameters, first, obtain the Diversity PC Index
(DivPcIndex (Utran Cell)) from the configuration items of Utran Cell. Next, query
DivPcIndex (Power Control Related to Service and Diversity Mode), TxDivMod and
SrvType in Power Control Related to Service and Diversity Mode.
2. Related Measurement
TCP: Transmitted Carrier Power. The internal measurement value of NodeB is obtained
fromthe public measurement report and reported to RNC. The measurement is started
after the cell is set up and the public transport channel of cell is set up, and ended after
the cell is deleted. CPICH Ec/N0 is the SNR for reception of CPICHWhen a service is
set up, the measurement result carried in the RRC connection request is used. In the
case of handover, the measurement result of intra-frequency or inter-frequency
measurement report is used.
2.3 Downlink open loop power control of R99 in compressed mode
As the adjustment proportion of DPDCH transmit power is the same as that of the
control domain for the downlink compressed mode, it is unnecessary to change the
values of PO1, PO2 and PO3. That is, the power offset between the control part and
data part in the compressed mode is the same as that in normal mode.


15
3 Upl ink i nner l oop power control of R99
3GPP TS 25.214 specifies the following methods for calculating inner loop power
control.
At the receiving end, first, the SIR measurement (SIR=Eb/No) is done for each
received radio link. Next, the measurement result is compared with the target
SIR (SIR
target
) required by the service.
If SIR SIR
target
, control information is returned to the sender with a transmit
power command (Transmitted Power Control-TPC) whose bit value is 0.
If SIR <SIR
target
, a TPC command whose bit value being 1 is returned through
the downlink control channel to the sender.
The sender judges whether to increase or decrease the transmit power
depending on the received TPC command and specified power control
algorithm. The adjustment extent = TPC_cmdTPC_STEP_SIZE
(TpcStepSize).
This section discusses how to select the proper inner loop power control algorithm, as
the principles of inner loop power control between uplink and downlink are same.
Description of TPC: When UTRAN and UE setup the first radio link, before uplink
synchronization, UTRAN could not work out the TPC in normal way. So UTRAN send
fixed TPC pattern in TPC bit of the downlink DPCH. The TPC pattern shall consist of
DlTpcN instances of the pair of TPC commands ("0", "1"), followed by one instance of
TPC command "1". The TPC pattern continuously repeat but shall be forcibly
re-started at the beginning of each frame where CFN mod 4 =0. And the TPC pattern
shall terminate once uplink synchronisation is achieved.
3.1 Uplink inner loop power control of R99 in normal mode
There are two uplink inner loop power control algorithms (UlIlPcAlg), which are
described as follows.
Algorithm1 (UlIlPcAlg =1):
With algorithm1, the transmit power of sender can be adjusted in every timeslot. Each
timeslot, the receiver judges, whether to increase or decrease the transmit power of the
UMTS Power Control

16
sender depending on the received TPC command,.
Rules for UE to combine the TPC command are as follows:
Suppose the TPCs of all radio link sets are 1, then TPC_cmd=1 (to increase
transmit power).
Suppose one TPC coming from any radio link set is 0, then TPC_cmd=-1
(TPC being 0 indicates the transmit power should be decreased).
Algorithm2 (UlIlPcAlg =2):
With algorithm2, the transmit power of sender is adjusted once every five timeslots.
Rules for UE to combine TPC command are (when single TPC or several TPCs are
received in one timeslot):
When a single TPC is received: Transmit power is not adjusted in the first four
timeslots (TPC_cmd=0). When the TPC command of the 5
th
timeslot is received, a soft
decision is made: TPC_cmd=1 if all five received TPC commands are 1; TPC_cmd=-1
if all five received TPC commands are 0; TPC_cmd=0 in other cases.
When several TPCs are received: Transmit power is not adjusted in the first four
timeslots of the five continuous timeslots (TPC_cmd=0). At the 5
th
timeslot, first
determine TPC
i
(i=1,2,,N, N is the number of radio link TPC commands from
different radio link sets). Next, combine the TPC command respectively as when a
single TPC is received to obtain N number of temporary TPC commands (TPC_temp).
Finally, combine TPC_cmd in the following rule:
If 5 . 0 _
1
1
>

=
N
i
i
temp TPC
N
, then TPC_cmd =1.
If all _
i
TPC temp is -1, then TPC_cmd =-1.
In other cases, TPC_cmd =0.
Description: TPC_cmd =1 indicates to increase the transmit power; TPC_cmd =-1
indicates to decrease the transmit power; TPC_cmd =0 indicates not to adjust the
transmit power.
Principle for selecting the inner loop power control algorithm:
Algorithm 1 is to perform inner loop power control at each timeslot, while
algorithm 2 is to perform inner loop power control only once every five
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
17
timeslots. That is, the frequency is higher to perform inner loop power control
in algorithm 1, When the environment of mobile communication is quite
infavorable and the channel fades very quickly, algorithm 1 helps the transmit
power to converge fast to meet the service quality requirement.
With algorithm 2, the inner loop power control is performed every five
timeslots, that is, the frequency is lower to perform inner loop power control in
algorithm 2. So algorithm 2 is applicable when the environment of mobile
environment is quite favorable (the MS is or will be in static state, for instance)
and the channel fades slowly or hardly fade.
With algorithm 1, when the TPC command is received, the transmit power is
either increased or decreased. With algorithm 2, the transmit power is
increased, decreased or not changed after a soft decision is made for the
TPC command at five different timeslots. In the respect, algorithm 1 is more
applicable in the case when the channel needs the transmit power to be
increased or decreased fast since it is fading fast.
In cases when the channel fades rather slowly, algorithm 2 is more applicable
because the BLER is good enough in a long period even if the transmit power
is not changed during this period and the measured SIR changes very little as
opposed to the target SIR.
Description:
When uplink inner loop power control is being performed, the transmit power
calculated by UE can exceed the maximumtransmit power of uplink DPCH
(MaxUlDpchPwr). In this case, UE can only transmit with this configured
maximumtransmit power. In order to show the differentiation of different basic
priority users, the actual maximumallowed uplink DPCH transmission power is
configured distinctively for different users with different basic priority.
UlIlPcAlg depends on sub-service types and it is obtained in this way: First, the
service-related power control parameter configuration index (TrfPcIndex (Utran
Cell)) is obtained from the configuration item of Utran Cell. Next, query
TrfPcIndex (Traffic-Related Power Control) and SrvType in Traffic-Related
Power Control. All service-related power control parameters can be obtained in
the same way.
3.2 Uplink inner loop power control of R99 in compressed mode
The principle of inner loop power control in compressed mode is the same as that in
normal mode. That is, the serving cell (a cell in the active set) estimates the received
UMTS Power Control

18
SIR
est
of uplink DPCH, and one TPC command is produced and sent in each timeslot
except the downlink transmission gap according to following rules. The rules are: If
SIR
est
>SIR
cm_target
, then the TPC command is 0; if SIR
est
<SIR
cm_target
, then the TPC
command is 1. SIR
cm_target
is the target SIR value during the period when the compressed
mode is adopted.
Way to determine SIR
cm_target
:
SIR
cm_target
=SIR
target
+ASIR
PILOT
+ASIR1_coding +ASIR2_coding
Where,
SIR
target
is the target SIR in normal mode.
ASIR
PILOT
=10Log
10
(N
pilot,N
/N
pilot,curr_frame
):
N
pilot,curr_frame
is the number of pilot bits per timeslot in the current uplink link
frame.
N
pilot,N
is the number of pilot bits per timeslot in normal mode without
transmission gap.
ASIR1_coding and ASIR2_coding are obtained from the parameters of
high-level signal configuration, that is, DeltaSIR1, DeltaSIR2, DeltaSIRafter1
and DeltaSIRafter2 can be calculated using the following methods.
If the current uplink link frame contains the start of the first transmission gap
of the transmission gap pattern, then: ASIR1_coding =DeltaSIR1 (2.3dB).
If the current uplink link frame contains the next frame to the start of the first
transmission gap of the transmission gap pattern, then: ASIR1_coding =
DeltaSIRafter1 (0.3dB).
If the current uplink link frame contains the start of the second transmission
gap of the transmission gap pattern, then: ASIR2_coding =DeltaSIR2 (0dB).
If the current uplink link frame contains the next frame to the start of the
second transmission gap of the transmission gap pattern, then:
ASIR2_coding =DeltaSIRafter2 (0dB).
In other cases, ASIR1_coding =0 dB and ASIR2_coding =0 dB.
As one TGPS (transmission gap pattern sequence) can have only one measurement
value but UE can measure several values at the same time, multi compressed modes can
be activated at the same time in one radio frame. In this case, the ASIR1_coding and
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
19
ASIR2_coding corresponding to each compressed mode can be calculated first and then
summed up to obtain the final available ASIR1_coding and ASIR2_coding.

=
A = A
N
i
i coding SIR coding SIR
1
, _ 1 _ 1

=
A = A
N
i
i coding SIR coding SIR
1
, _ 2 _ 2
Where: N is the type of compressed modes activated at the same time in one radio
frame.
Because no TPC command is sent in the timeslot of the transmissin gap in the
downlink compressed frame, UE sets TPC_cmd to 0 in corresponding receiving
timeslot. Because of the existence of transmission gap of compressed frame in
compressed mode, the format of timeslot used in compressed mode is different from
that in normal mode. As a result, the number of pilots of each timeslot of the uplink
DPCCH may differ between compressed mode and non- compressed mode. To offset
the changes in total power of pilot signals, the transmit power of uplink DPCCH should
be changed. Therefore, at the start of each timeslot, UE calcualtes the power
adjustment volume A
PILOT
.
If the number of pilots per timeslot of uplink DPCCH is different fromthat already sent
in the previous timeslot, then A
PILOT
(dB) is obtained using the following formula:
A
PILOT
=10Log
10
(N
pilot,prev
/N
pilot,curr
);
Where,
N
pilot,prev
is the number of pilot bits of the previous timeslot.
N
pilot,curr
is the number of pilot bits of the current timeslot.
Otherwise, A
PILOT
=0.


21
4 Downli nk Inner Loop Power Control Of
R99
4.1 Downlink inner loop power control of R99 in normal mode
In the case of the downlink inner loop power control, UTRAN adjusts the current
downlink power P(k-1) to the new transmit power P(k) according to the following
formula when it estimates the k
th
number of TPC command.
P(k) =P(k - 1) +P
TPC
(k) +P
bal
(k)
Where,
P
TPC
(k) is the k
th
number of power adjustment volume in the process of inner
loop power control.
P
bal
(k) is a correction value obtained according to the downlink power control
process. It is used to balance the power of radio link so that the value can
approach a common reference power.
Two power control modes are also available to determine P
TPC
(k):
Mode 1: UE sends a TPC command at each timeslot. The UTRAN adjusts the
transmit power at each timeslot according to the TPC command.
Mode 2: UE sends the same TPC command for three timeslots. The UTRAN
adjusts the transmit power once every three timeslots according to the TPC
command.
RNC also considers the requirement of power increase limits at the same time for
downlink inner loop power control. The value of P
TPC
(k) is determined according to
the following principle:

=
> A + A =
< A + A =
A
A +
=
0 ) ( TPC if
e_Limit Power_Rais ) ( and 1 ) ( TPC if
e_Limit Power_Rais ) ( and 1 ) ( TPC if
0 ) (
est
est
est
k
k k
k k
k P
TPC sum
TPC sum
TPC
TPC
TPC

=
= A
1
indow_Size veraging_W DL_Power_A
) ( ) (
k
k i
TPC sum
i P k
Where,
TPC
A is the power adjustment step (TpcDlStep), and Power_Raise_Limit is
PwrRaisLim.
UMTS Power Control

22
Since for the algorithmof downlink inner loop power control, the transmit power is
also adjusted once for one or three timeslots, the selection of this algorithmalso
depends on the channel fading status. That is, algorithm1 is for fast channel fading and
algorithm 2 for slow channel fading. The reason for such selection principle is similar
to that of uplink inner loop power control. At present, RNC is only support UE send
TPC command at each timeslot.
4.2 Downli nk i nner loop power control i n compressed mode
The inner loop power control of UE in compressed mode works in the same way as that
in normal mode, except that both downlink DPDCH and DPCCH stop transmission
during the transmission gap of compressed frames. The transmit power of the first
timeslot after the transmission gap of DPCCH is the same as that of the timeslot prior
to the transmission gap.
During the period when the compressed mode is adopted, the UTRAN adjusts the
current downlink transmit power P(k-1) [dB] of each timeslot except the downlink
transmission gap to a new power value P(k) [dB] based on the TPC command received
at the number k-1
th
timeslot and the following formula.
P(k) =P(k - 1) +P
TPC
(k) +P
SIR
(k) +P
bal
(k)
Where,
P
TPC
(k) is the k
th
time of power adjustment value according to inner loop
power control.
P
SIR
(k) is the k
th
time of power adjustment value used for the reason that in
compressed mode, the downlink SIR
Target
changes as opposed to that in
normal mode (this change is reflected in inner loop instead of outer loop).
P
bal
(k) [dB] is a correction value obtained according to the downlink power
control process. It is used to balance the power of radio link so that the value
can approach a common reference power.
Because of the existence of transmission gap in uplink compressed frames, the uplink
TPC command may fail to be received. In this case, NodeB sets P
TPC
(k) as 0.
Otherwise, P
TPC
(k) is calculated in the same way as that in normal mode except that
A
TPC
is replaced with A
STEP
in the formula.
During the recovery period (RPL number of timeslots) of the transmission gap, the
common power transmission control algorithmis adopted but A
STEP
= A
RP-TPC
. In a
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
23
non-recovery period, A
STEP
=A
TPC
.
Where,
RPL is the length of the recovery period that is expressed in number of
timeslots. RPL=minimum (out of the transmission gap length, 7). If the next
transmission gap starts again before the recovery period ends, then the
recovery period ends at the start of the next transmission gap. RPL depends
on the length of the new transmission gap. RPL=7.
A
RP-TPC
is the step (dB) of power control during the recovery period. A
RP-TPC

=minimum (3dB, 2A
TPC
).
Power offset P
SIR
(k) =P
curr
- P
prev
, P
curr
and P
prev
respectively indicate the P value
of the current timeslot and the latest transmission timeslot. The formula for calculating
P is as follows:
P=max (AP1_compression,,APn_compression) +AP1_coding +AP2_coding
Where: n is the type of TTI length of all TrCHs multiplexed to a CCTrCH. AP1_coding
and AP2_coding are obtained from the uplink parameters, including DeltaSIR1,
DeltaSIR2, DeltaSIRafter1 and DeltaSIRafter2, which are notified by the upper level
and also according to following relations:
If the current frame contains the start of the first transmission gap, then
AP1_coding =DeltaSIR1 (2.3dB).
If the current frame is next to the frame that contains the start of the first
transmission gap, then AP1_coding =DeltaSIRafter1 (0.3dB).
If the current frame contains the start of the second transmission gap, then
AP2_coding =DeltaSIR2 (0dB).
If the current frame is next to the frame that contains the start of the second
transmission gap, then AP2_coding =DeltaSIRafter2 (0dB).
In other cases, AP1_coding =0 dB, AP2_coding =0 dB.
APi_compression is defined as follows:
If the compressed mode with half spectrum spread factor is adopted,
APi_compression =3 dB.
In other cases, ASIR_compression =0.
UMTS Power Control

24
When several compressed modes are used at the same time, P of each compressed mode
is calculated separately. The P adopted for the current frame is the summation of all
P values.
No transmit power of any timeslot in compressed mode can be higher than the allowed
maximum transmit power or lower than the allowed minimum transmit power.
ASIR_compression is used to offset the influence of high SIR needed by the rate
increase of transmission bit in compressed mode.


25
5 Upl ink Outer Loop Power Control of R99
Outer loop power control differs between uplink and downlink directions. The
downlink outer loop power control is realized in the UE and it is unrelated with RNC.
This section describes the uplink outer loop power control algorithmin the UTRAN
only. The principle is: The initial SIR
Target
value (ULInitSIR) is determined upon service
access, and the quality information (such as CRCI and BLER) is obtained fromthe
measurement report and produce the decision command. If adjustment is necessary,
SIR
Target
is adjusted slowly and the signaling OUTER LOOP PC is used to notify
NodeB. NodeB compares the SIR in the dedicated measurement report with the latest
SIR
Target
and makes the single link SIR approach to SIR
Target
through inner loop power
control. In this way, the service quality will not fluctuate drastically in a changing radio
environment. The outer loop power control algorithm based on threshold report is
adopted in RNC. The threshold report algorithmis described as follows.
1. CRC-based outer loop power control algorithm
The principle of the CRC-based outer loop power control algorithmis: The number of
error blocks is counted acording to the CRC result of transport channel. In addition,
total number of transmitted data blocks is also counted (referred to as error block
tolerance counter).
Principle for increase: When the tolerance BLER period (BLERAccpPeriod) (with its unit
being number of data blocks, instead of a time measurement unit) has not expired yet,
but the number error blocks has already exceeded the error transport block number threshold
(ErrorThresh), now increase SIR
Target
(meanwhile, clear the error block counter and
error block tolerance counter to 0).
Principle for decrease: When the error block tolerance counter is no less than the
tolerance BLER period (BLERAccpPeriod).
(1)decrease SIR
Target
if now the received number of error blocks is less than the error
transport block number threshold (ErrorThresh).
(2)keep the SIR
Target
same if now the received number of error blocks equals to the error
transport block number threshold (ErrorThresh).
After SIR
Target
is decreased, it is necessary to clear the error block counter and error
UMTS Power Control

26
block tolerance counter to 0.
Here the step (UlSirTargUpStep) for increasing SIR
Target
can be set to a value greater
than the step (UlSirTargDnStep) for decreasing the SIR
Target.
For example, the increase
step is 0.3dB, the decrease step is 0.1dB. Different QoS are reflected in the values of
tolerance BLER period (BLERAccpPeriod) and error transport block number threshold
(ErrorThresh).
As the loop delay is at least 4~5 frames, the effect of increase will be shown after 4~5
frames. Therefore, if CRC indication error occurs again in 4~5 frames after the
increase, no error block is counted (a shield period (CoverPrd) is used here to shield out the
adjustment function). If the CRCI indication error occurs again after the shield period
expires, the error blocks are counted again. If no CRC error is detected during the shield
period, the counter is decreased by 1. Next the decrease counter is checked to see
whether it has reached the period of SIR
Target
decrease. If yes, SIR
Target
is decreased; if
no, SIR
Target
is kept unchanged.
To prevent that SIR
Target
is increased or decreased too much, the maximumvalue
(ULMaxSIR) and minimumvalue (ULMinSIR) of SIR
Target
is configured in the OMCR.
If the calculated SIR
Target
is greater than ULMaxSIR or smaller than ULMinSIR,
ULMaxSIR or ULMinSIR will be taken as a result.
2. Outer loop power control combination strategy for mixed services
The common outer loop power control algorithm described above is designed for one
transport channel. For mixed services (that is, several transport channels are
multiplexed to one CCtrCH), some special treatment is needed for the outer loop power
control algorithm.
For mixed services (that is, several transport channels are multiplexed to one CCtrCH),
if any one service type fails to satisfy the service quality requirement, SIR
Target
is
increased. SIR
Target
is not decreased unless all services valid for statistics indicate to
decrease SIR
Target
. Services invalid for statistics are excluded from the combination of
power control. Services invalid for statistics are those services that cannot serve as the
reference for SIR
Target
adjustment because their data volume is not enough. If a service
do not have enough data volume, it means the total number of packets received in the
valid time window (ValidTimeWin) is smaller than the error block tolerance period for
the service.


27
6 Downli nk Outer Loop Power Control of
R99
The downlink outer loop power control is realized in the UE. RNC provides
BLERtarget to UE.
BLERtarget corresponding to different downlink traffic are listed in the following table.
Traffic BLERtarget
DL CS 64kbps Conversational 0.1%
DL PS Conversational Video 0.1%
Other downlink services 1%
For F-DPCH, TPC command error rate target (TpcErrTarget) should be provided, it is
used for adjusting SIR target of F-DPCH.


29
7 Downli nk Power Balancing
7.1 Al gori thm Description
In the soft handover or macro diversity status, a UE can communicate with all cells in
the active set. With downlink inner loop power control, the UE sends the same TPC
command to the cells in the active set. But because each link is available with a
different transmission path, error codes will be produced in the TPC command and
some cells will receive wrong TPC command. As a result, some cells increase
downlink transmit power and some cells decrease downlink transmit power, hence the
drifting power. Power drifting is usually eliminated through the downlink power
balancing approach.
The purpose of downlink power balancing is to balance the downlink transmit power of
one or more radio links used by the NodeB of related RRC connection. In the case a
single link is involved, the downlink average power will be insensitive to the central
value of a power control range if the downlink power control balancing is adopted. In
the case that several links are involved, power balancing can help overcome power
drifting.
A simple formula for calculating P
bal
is as follows:
) )( 1 (
init CPICH P ref bal
P P P r P + =

precision0.5 dB
Where,
P
ref
is a reference power, which equals to DL Reference Power.
P
P-CPICH
is the transmit power of the primary CPICH (PcpichPwr).
P
init
is the code power of the last timeslot in the previous adjustment period. If
the last timeslot in the previous adjustment period coincidentially is included in
the transmission gap (in compressed mode), then P
init
equals to the code
power of the timeslot prior to the transmit gap.
r is the adjustment convergence coefficient (AdjRatio) that ranges 0~1.
A simple method for calculating DL Reference Power is as follows:
The downlink transmit power of each radio link, needed for calculating DL Reference
Power, can be obtained indirectly fromthe Transmitted code power (TCP: transmit
UMTS Power Control

30
power of PILOT domain of DPCH) periodically reported by NodeB using the
following formula:
3 ) ( PO TCP dBm P
DPDCH
j
=
Where,
jPDPDCH is the downlink transmit power of j:th radio link.
PO3 is the power offset between the DL DPCH PILOT domain and the DPCH data
domain (DpchPO3).
The downlink reference power of i:th radio link is:
Power CPICH P P dB Power fer DL
DPDCH
j j
_ _ ) ( _ Re _ =
Where, P_CPICH_POWER is P-CPICH power (PcpichPwr).
Next, RNC takes the average value of the reference power of each radio link as the DL
Reference Power needed:
N
Power fer DL
Power fer DL
N
j
j

=
=
1
_ Re _
_ Re _
Where: N is the number of radio links used by the NodeB.
Method to realize power balancing:
The dedicated TCP values of all links are obtained from the dedicated
measurement report. The DL Reference Power is obtained by computing the
reported values.
When the absolute value of the difference between the DL Reference Power
obtained in the new adjustment period and that obtained in the previous period
exceeds the downlink reference power adjustment threshold (DlRefPowUpdtTh),
the signaling of DL Power Control Request message which contains the
information of DL Reference Power sent to NodeB.
NodeB uses this value to implement link balancing through the inner loop power
control algorithm.
Adjustment Type (AdjType):
AdjType is used to select whether to performdownlink power balancing adjustment and
the adjustment type. Power Adjustment Type can take the value of None, Common
or Individual.
Chapter Error! Use the Home tab to apply 1 to the text that you want to appear here. Error! Use th
e Home tab to apply 1 to the text that you want to appear here.
31
When the value of AdjType is None,it means NodeB not need to balance the DL
power.
When the value of AdjType is Common, it means NodeB balance the DL power but
the balanced radio links use common reference power.
When the value of AdjType is Individual, it means NodeB balance the DL power but
the balanced radio links use Individual reference power.
Adjustment Period (AdjPeriod):
The value of AdjPeriod usually does not change once it is selected. It is a value
determined through actual tests.
Adjustment Ratio r (AdjRatio):
AdjRatio can be 0.96 by default. The smaller the value of AdjRatio is, the quicker the
offsets of transmit power of base stations are converged to be as the power offset as
opposed to the common pilot channel of cells. But as the adjustment volume of power
balancing is limited by the maximumadjustment step, the value of convergence is also
limited. The AdjRatio can be set as a fixed value while the period is adjusted as Max
Adjustment Step, which never change after configuration.
Max Adjustment Step (1~10 slots) (MaxAdjStep):
MaxAdjStep defines a time period, in terms of number of slots, in which NodeB can
make power adjustment for balancing purpose by no more than 1dB .
7.2 Related measurement
The measurement of TCP (transmit code power) is reported periodically. The
measurement of TCP gets started after UE changes status frommacro diversity to
non-macro diversity, and is terminated after UE changes status from non-macro
diversity to macro diversity.