3GPP TSG RAN WG1 Meeting #51 R1-074704
`Jeju, Korea, November 05 – 09, 2007
`
`InterDigital Communications, LLC
`Source:
`Uplink power control procedures and Text Proposal for E-UTRA
`Title:
`Agenda Item: 6.4.2
`Document for: Discussion and decision
`
`
`1. Introduction
`
`This contribution describes UE and eNodeB behavior for uplink power control. We first present a
`description of the procedures, incorporating agreements reached at #50bis [2], [3], and then give
`text modifications for TR36.213 [1].
`2. General procedures for uplink power control
`2.1 Physical uplink shared channel
`The PUSCH power control has two components, an open loop and a closed-loop component. Both
`open and closed loop components run consecutively, but asynchronously. The procedure can be
`viewed as the following flow diagram:
`
`
`Figure 1 Flow chart for PUSCH power control
`
`
`
`HTC/ZTE EXHIBIT 1009
`
`

`

`In a given uplink subframe i, PUSCH may be transmitted
`-
`on a dynamically assigned resource (by an uplink grant on PDCCH)
`-
`or on a persistently assigned resource
`-
`or not at all
`In a given downlink subframe n, a power control command ΔPUSCH (n) for PUSCH power control
`may be provided
`- within an UL scheduling grant on PDCCH
`-
`or on a TPC-PDCCH (referred to as the TPC-PDCCHPUSCH)
`-
`or not at all
`If, in a given downlink subframe n, power control commands are provided within an UL
`scheduling grant and on a TPC-PDCCHPUSCH, the valid power control assumed to be delivered in
`subframe n should be the power control command provided within the UL scheduling grant (i.e. the
`power control command on TPC-PDCCHPUSCH should be ignored).
`If PUSCH is transmitted in subframe i, it is transmitted with the power PPUSCH (i). In the power
`control formula below, KPUSCH is the delay in the PUSCH power control, i.e. a power control
`command provided in downlink subframe n will not impact the PUSCH transmit power in
`subframes prior to subframe n+KPUSCH. Exact value of KPUSCH is TBD.
`The power control formula for PUSCH, as modified in [3], is outlined below:
`=
`+
`Δ+
`
`⋅+ α
`Δ+
`P
`
`)(i
`P
`
` ) PM
`PL
`f
`min(
`log
`10,
`(
`[
`o
`PUSCH
`10
`max
`
`K-i(
`
`mcs
`
`PUSCH
`
`)]
`
`
`
`PUSCH
`
`where:
`•
` is the maximum allowed power (in dBm) that depends on the UE power class
`maxP
`• M is the number of assigned resource blocks as indicated in the UL scheduling grant
`•
`oP is a UE specific parameter (in dBm) with 1 dB resolution over the range: [-126dBm,
`24dBm]
`• α is cell specific path loss compensation factor (can be set to one to allow full path loss
`compensation) that has 8 values from 0.4 to 1 in steps of 0.1 with one of the possible
`values being zero.
`• PL is the downlink pathloss calculated in the UE from a RSRP measurement and signaled
`RS transmit power
`mcsΔ
` is signaled by RRC ( mcsΔ
` table entries can be set to zero)
`o MCS signaled in each UL scheduling grant
`Δ
` is a UE specific correction value and is included in every Nth UL scheduling grant,
`PUSCH
`(where N can be 1), or jointly coded with other UE specific correction values on a TPC-
`PDCCHPUSCH.
`o The UE attempts to detect a TPC-PDCCHPUSCH on every subframe except when in
`DRX.
`
`•
`
`•
`
`
`
`HTC/ZTE EXHIBIT 1009-2
`
`

`

`The power control formula is applied differently dependent on scheduling as given by:
`• Dynamically scheduled PUSCH
`][∗f
`o Function
` represents either accumulation or absolute value, The mode is
`Δ
`signaled semi-statically via higher layers. When a new value of
` is
`received in the Nth scheduling grant,
` For absolute control
`Δ
`f
`
` K-(i
`[
`)]
`PUSCH
`PUSCH
` For accumulation control
`
`Δ=
`
`
`
` K-(i
`
`)
`
`
`
`PUSCH
`
`PUSCH
`
`PUSCH
`
`Δ
`{
`
`PUSCH
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`i
`
`=
`
`m
`
`0
`
`Δ
`
`[
`
`f
`
`PUSCH
`
`−
`i K(
`
`
`
`PUSCH
`
`=
`
`)]
`
`• Persistently scheduled PUSCH
`][∗f
`o Function
` represents accumulation only. When a new value of
`received in the TPC-PDCCHPUSCH,
`
`Δ
`
` is
`
`PUSCH
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`PUSCH
`
`Δ
`{
`
`i
`
`=
`
`Δ
`
`f
`
`[
`
`PUSCH
`
`−
`i K(
`
`
`
`=
`
`)]
`
`PUSCH
`
`m
`0
`• Combined dynamically and persistently scheduled PUSCH
`o When UL grant is configured for accumulation commands (for dynamically
`scheduled PUSCH), the UE combines the accumulation commands received in
`both the SG and the TPC-PDCCHPUSCH
`o When UL grant is configured for absolute commands (for dynamically scheduled
`PUSCH), the UE resets the accumulation immediately after receiving each
`absolute TPC and then combines the absolute with the accumulation
`i
`
`PUSCH
`
`,
`
`TPC
`
`−
`
`PDCCH
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`+
`
`+
`
`)
`
`−
`Kj
`
`(
`
`PUSCH
`
`Δ
`
`f
`
`[
`
`PUSCH
`
`−
`i K(
`
`
`
`PUSCH
`
`Δ=
`
`)]
`
`PUSCH
`
`UL
`,
`
`_
`
`grant
`
`Δ
`{
`=
`jm
`1
`where j is the subframe index of the last absolute command.
`If no power control step ΔPUSCH (k) is provided on neither an UL scheduling grant or on a TPC-
`PDCCHPUSCH in subframe k, then
`• For absolute control, ΔPUSCH (k) in the equations above should be set to the latest value of ΔPUSCH.
`• For accumulation control, ΔPUSCH (k) in the equations above should be set to zero.
`2.1.1 Open loop component
`The UE first determines the open loop component based on a filtered linear pathloss estimate, pl,
`from the serving eNodeB to the UE. The pathloss is updated in the power control formula after
`each downlink RSRP measurement. The UE continuously (or periodically) measures the
`instantaneous pathloss at the k-th instance based on the DL RSRP whose transmit RS power is
`signaled to the UE.
`
` filtering method is then applied to the pathloss measurements, such as
`
` A
`
`HTC/ZTE EXHIBIT 1009-3
`
`

`

`
`
`⋅= ρ
`
`⋅
`
`−+
`ρ
`pl
`pl
`pl
`inst
`1(
`)
`
`−
`k
`k
`k
`1
`where plk and plk-1 represents the filtered pathloss at the k-th instance and (k-1)-th instant,
`is the instantaneous pathloss at the k-th instant. ρ is a filter coefficient,
`kpl
`inst
`respectively.
`
`≤≤ ρ , which is generally determined by the UE, possibly depending on pathloss variation, fast
`0
`1
`fading rate, the time of UL transmission, etc. Then, the pathloss in dB is obtained as
`PLk = 10log10 (plk).
`Alternatively, a moving averaging method may be considered for the pathloss filtering.
`2.1.2 Closed loop component
`Additionally, the UE applies a closed-loop power correction factor relative to the open loop power,
`primarily in order to compensate for open loop errors, including the pathloss estimation error due
`to non-perfect reciprocity in UL and DL in FDD, and the UE impairments due to power amplifier
`and receiver non-linearities.
`The correction factor is derived at the UE, based on the recent received correction command(s).
`Which UL scheduling grants convey the correction command is a higher layer configurable
`parameter per UE basis, so that the UE knows which UL grants to look at for the correction
`command. For example, the command signaling is done in particular UL scheduling grants such as
`the UL grant associated with a pre-defined HARQ process, say, HARQ process #1; not every grant
`corresponding to that HARQ process has to carry a TPC command. In this case, it is assumed that
`multiple control channel formats (such as one with a PC correction command and another without
`it) are supported in the downlink. No additional control signaling to indicate a control channel
`format in use in a given TTI is required, since the UE knows in advance which UL grants carry the
`correction command. We also assume the UE applies BD (blind detection) to determine which
`grant format was sent. This BD can be based on message length, CRC mask, code or other methods.
`
`BD protects against the problem of a lost TPC command if a SG (scheduled grant) is not detected.
`If the kth grant containing ΔPUSCH (k) is lost, the UE would not transmit on the assigned allocation
`due to the grant failure. The eNodeB would, of course, not receive the expected transmission. The
`eNodeB may then append ΔPUSCH to the next scheduling grant. Upon successful detection the UE,
`having not received the previous grant, would consider this one the kth grant.
`
`Another scenario is when the UE correctly receives the kth grant with ΔPUSCH (k), resets its counter,
`transmits its message, but the eNodeB fails to correctly receive the UE’s transmission. Following
`the logic above, the eNodeB may resend ΔPUSCH (k) in the next grant. However, the UE knows to
`ignore this redundant power control command.
`
`[f Δ
`
`]
` to the Tx
`For scheduled UEs there are two alternatives for applying the correction
`PUSCH
`power, accumulated or absolute value. The eNodeB sends to each scheduled UE (or a sub-group of
`scheduled UEs) a power correction factor using multiple command bits, 2 bits for accumulation
`control or TBD [2 or 3] bits for absolute control, in the UL grant (or on the TPC-PDCCHPUSCH for
`accumulation), where the correction command is determined based on link quality (such as
`received PSD or SINR) of the UL power controlled data channel (and possibly UL sounding
`reference symbol, if available). For instance, assuming absolute control, the correction factor may
`be determined such as
`[
`SINR
`Tar
`
`ESINR
`
`Δ
`
`PUSCH
`
`=
`
`−
`
`get
`
`]est
`
`
`
`HTC/ZTE EXHIBIT 1009-4
`
`

`

`
`
`
`
`0.1e
`
`where ESINRest and SINRTarget denote the effective SINR (ESINR) estimate at the receiver and
`target SINR, respectively, of the power controlled channel(s) in dB. [x] denotes a correction value
`in the correction set which is nearest to x. The observed samples at the eNodeB for the ESINR
`estimation include (some of or all) SC-FDMA symbols of the UL power controlled channel(s),
`which have been received over an averaging window. Because slow power control is to be used, a
`linear block average since the last correction command signaling may be used when commands are
`sent infrequently, or a moving average over a suitable window may be used if updates are sent in
`every grant.
`
`For persistenlty scheduled UEs only accumulation control is used. For accumulated control the
`eNodeB sends a command in every N grants or in the TPC-PDCCHPUSCH. Assuming block
`averaging is used, the command is derived as follows:
`
`For each reception from the UE, the eNodeB computes an error given by
`−
`=ke
`SINR
`ESINR
`
`
`
`Tar
`get
`k
`ESINR is the effective SINR, in dB, for the kth reception since the last command
`where
`k
`was sent. The eNodeB computes a block average over M receptions during N intervals:
`= M
`/)k M
`
`(
` E
`10
`1
`The command sent is obtained as
`
`[
`]E
`=
`Δ
`
`log
`10
`
`
`
`PUSCH
`10
`where [x] denotes a correction value in the correction set which is nearest to x. If the command is
`sent frequently, e.g. every grant, then a moving average over a window M can be used.
`2.1.3 Timing of closed loop corrections
`To reduce the signaling overhead of the power correction command, the correction command
`signaling is not required in every UL grant. Assuming that there exist multiple DL control formats,
`we can reduce the signaling overhead by applying the following rules:
`• A correction command signaling timing is configured at the eNodeB (or on a RRC level)
`per UE basis and is then known at both the eNodeB and the UE via higher layer signaling.
`• When the correction command is signaled in the UL grant, assuming that UL HARQ is
`synchronous, the signaling timing configuration can be simplified such that the command
`signaling is done in particular UL grants such as the UL grant associated with a pre-defined
`HARQ process, say, HARQ process #1. However, even in this case it is not necessary to
`signal the correction commands in all the associated UL grant channels. For example, the
`signaling may occur in every N associated grant channel for N >= 1, which would be
`equivalent to one command signaling in every N HARQ cycle period. Figure 2Figure 2
`shows an example of the proposed PC scheme when the PC correction command is
`conveyed in the UL grant associated with HARQ process #1 and N is set to 2. In this
`example, the PC update rate is 8 msec, assuming the number of HARQ processes is 4 and
`the inter-TTI is equal to 1.
`• The signaling timing (or associated parameters) may be reconfigured on a semi static basis.
`
`HTC/ZTE EXHIBIT 1009-5
`
`

`

`Time (msec)
`
`
`
`DL control
`signaling
`
`UL grant &
`ACK/NACK
`(HARQ 1)
`
`UL grant &
`ACK/NACK
`(HARQ 2)
`
`UL grant &
`ACK/NACK
`(HARQ 3)
`
`UL grant &
`ACK/NACK
`(HARQ 4)
`
`UL grant &
`ACK/NACK
`(HARQ 1)
`
`UL grant &
`ACK/NACK
`(HARQ 2)
`
`UL grant &
`ACK/NACK
`(HARQ 3)
`
`UL grant &
`ACK/NACK
`(HARQ 4)
`
`UL grant &
`ACK/NACK
`(HARQ 1)
`
`UL grant &
`ACK/NACK
`(HARQ 2)
`
`UL grant &
`ACK/NACK
`(HARQ 3)
`
`UL grant &
`ACK/NACK
`(HARQ 4)
`
`UL grant &
`ACK/NACK
`(HARQ 1)
`
`UL grant &
`ACK/NACK
`(HARQ 2)
`
`UL grant &
`ACK/NACK
`(HARQ 3)
`
` No recent PC
`command available
`(in the initial Tx or
`right after DTX)
`
`UL data
`channel
`
`ESINR estimation
`&
`Correction command
`(with 3 bits, 0, +/- 1, 3, 5, -7 dB)
`
`HARQ 1
`
`HARQ 2
`
`HARQ 3
`
`HARQ 4
`
`HARQ 1
`
`HARQ 2
`
`HARQ 3
`
`HARQ 4
`
`HARQ 1
`
`HARQ 2
`
`HARQ 3
`
`HARQ 4
`
`HARQ 1
`
`HARQ cycle period
`
`Potential options without a recent closed loop correction
`
`- Open loop PC only
`- Based on pathloss change between the time before the DTX and the time before resuming the UL Tx
`- Apply a power offset relative to the PSD of PUCCH (if the recent PSD of PUCCH is available)
`- Open loop + UL grant based closed loop
`
`Open loop + closed loop PC
`
`
`
`UL Tx PSD
`level
`Figure 2. Example of the proposed scheme when the PC correction command is conveyed only in the
`UL grant associated with HARQ process #1 and N is set to 2.
`
`When the UE receives one correction command from the serving eNodeB in a UL grant since the
`last Tx power adjustment, it derives a correction factor, ΔPUSCH, from the received correction
`command for the next power adjustment.
`
`The UE then adjusts the transmit power of the data channel using the derived correction factor, the
`most recent open loop power, and a power offset associated with the granted MCS, ΔMCS. The
`resulting Tx power is applied to the very beginning (first SC-FDMA symbol) of the next UL TTI
`for the data channel and remain constant until the next power adjustment, as shown in Figure
`2Figure 2.
`
`Figure 3Figure 3 and Figure 4Figure 4 show examples of the proposed power control scheme using
`different PC system configurations. In Figure 3Figure 3, we assume a HARQ cycle period of 4
`TTIs and that the correction command is signaled in the respective UL grant only associated with
`HARQ process #1 (N=1). In Figure 4Figure 4, a HARQ cycle period of 8 TTIs is considered as an
`example.
`
`
`
`HTC/ZTE EXHIBIT 1009-6
`
`

`

`
`
`
`
`Figure 3. Example of the proposed PC scheme when the PC correction command is conveyed in the
`respective UL grant only associated with HARQ process #1 and N is set to 1
`
`
`
`Figure 4. Example of the proposed PC scheme when the PC correction command is conveyed in the
`respective UL grant only associated with HARQ process #1, N is set to 1, and the inter-TTI is set to 2.
`
`
`2.2 Physical uplink control channel
`In a given downlink subframe n, a power control command ΔPUCCH (n) for PUCCH power control
`may be provided
`- within a downlink assignment on PDCCH
`-
`or on a TPC-PDCCH (referred to as the TPC-PDCCHPUCCH)
`-
`or not at all
`If, in a given downlink subframe n, power control commands are provided within a downlink
`assignment and on a TPC-PDCCHPUCCH, the valid power control command assumed to be
`delivered in subframe n should be the power control command provided within the downlink
`assignment (i.e. the power control command on TPC-PDCCHPUCCH should be ignored).
`
`HTC/ZTE EXHIBIT 1009-7
`
`

`

`The procedures and power control formula for the PUCCH are the same as for the PUSCH except
`for these restrictions:
`• α = 1
`• M=1
`• only accumulation control is used
`Different power offset (such as P0) and ΔMCS may be used. Also, the TPC-PDCCHPUCCH has a one-
`bit format: [-1,1] in addition to the two-bit formats used for the PUSCH.
`
`The two TPC-PDCCHs for the PUSCH and PUCCH are identified by two separately configured
`RNTIs. Nothing prevents the assignment of identical RNTIs for TPC-PDCCHPUSCH and TPC-
`PDCCHPUCCH.
`
`
`2.3 Sounding Reference Symbol
`srsP for the SRS is set equal to the
`The current working assumption is that the UE transmit power
`PUSCH power level. While this may be modified, the SRS power will be at least proportional to
`the PUSCH power.
`
`When there is no recent closed loop correction command for the PUSCH (for example, due to no
`recent scheduled UL data transmission, i.e., UL DTX), there may be several options for the UE to
`set its SRS Tx power as follows:
`Option-1) Relying on the open loop component.
`Option-2) Based on the pathloss variation between the time before the DTX and the time before
`resuming the UL transmission: if the UL DTX is short, the UE may adjust its Tx power such as
`(
`=
`+−
`−
`−
` )(nP
`
` (nP
`
`
`)(n
`P
`n
`P
`)1
`(
`open
`open
`Tx
`Tx
`where n is the Tx power setting time before resuming the UL transmission and (n-1) is the power
`setting time before the DTX. An example of this case is shown in Figure 5.
`Option-3) Applying a power offset relative to the most recent PSD for PUCCH, if available: even
`though there was no UL data transmission, there may be UL control signaling (such as CQI and
`ACK/NACK) for DL. In this case, since the UL control channel (PUCCH) is also power controlled
`(but using different parameters and update rate), we may use the PUCCH Tx power for the PUSCH
`Tx power as follows:
`(
`)
`P
`P
`PUCCH
`SRS
`PUCCH
`SRS
`(
`)
`(
`,
`)
`
`Tx
`Tx
`control
`where PTx(PUCCH) is the most recent power (or averaged over the recent updates) for PUCCH and
`Δcontrol( SRS , PUCCH) represent the PUCCH power offset relative to the Tx power for PUSCH.
`This offset accounts for the number of assigned resource blocks, the P0 offset and Δmcs.
`
`))1
`
`
`
`=
`
`Δ+
`
`HTC/ZTE EXHIBIT 1009-8
`
`

`

`
`
`
`n)(
`PSD
`Tx
`(
`=
`PSD
`
`=
`
`open
`
`'
`Tx
`
`PSD
`−
`
`n)(
`
`+−
`n
`)1
`(
`PSD
`(
`
`open
`
`)
`
`Δ+
`
`n
`
`−
`
`)1
`
`MCS
`
`Figure 5. Option-2) Tx PSD setting based on the pathloss variation between the time before the DTX
`and the time before resuming the UL transmission
`
`
`
`
`
`References:
`[1] 3GPP TR 36.213 V8.0.0: “Physical layer procedures”
`[2] R1-074511, “Summary of Power Control Discussion”, Nokia
`[3] R1-074489, “Uplink Power Control – Way forward”, Ericsson
`
`
`HTC/ZTE EXHIBIT 1009-9
`
`

`

`
`Modified text for TR36.213
`
`Based on these procedures the following text changes are proposed, note following section
`numbering is based on TR36.213 v8.0.0.
`5.1 Uplink power control
`Uplink power control consists of open and closed loop components and controls energy per resource element
`applied for a UE transmission. For intra-cell uplink power control the closed loop component adjusts a set
`point determined by the open loop power control component.
`
`Upon reception of an a-periodic transmit power command in an uplink scheduling grant the UE shall adjust
`its transmit EPRE accordingly.
`
`EPRE is set in the UE.
`
`A cell wide overload indicator (OI) is exchanged over X2 for inter-cell power control.
`
`[Note: Above line regarding OI and X2 to be moved to an appropriate RAN3 spec when it becomes
`available]
`
`Physical uplink shared channel
`
`UE behaviour
`5.1.1.1
`In a given downlink subframe n, a power control command ΔPUSCH (n) for PUSCH power control may be
`provided
`- within an UL scheduling grant on PDCCH
`-
`or on the TPC-PDCCHPUSCH
`-
`or not at all
`
` for the physical uplink shared channel (PUSCH)
`
`
`P
`The setting of the UE Transmit power PUSCH
`transmissions is defined by
`=
`P
`P
`
`)(i
`min(
`PUSCH
`max
`=
`P
`min(
`pusch
`
`10,
`P
`max
`
`⋅+ α
`
`+
`PL
` ) PM
`
`log
`(
`o
`10
`
`⋅+ α
`+
`PL
` ) PM
`
`(
`10,
`log
`o
`
`10
`
`Δ+
`Δ+
`
`mcs
`
`mcs
`
`Δ+
`f
`[
`K-i(
`PUSCH
`Δ
`+
`f
` (dBm)
`(
`))
`
`i
`
`)]
`
`PUSCH
`
`Wherewhere,
`•
`maxP
` is the maximum allowed power (in dBm) that depends on the UE power class
`• M is the number of assigned resource blocks as indicated in the UL scheduling grant
`•
`oP is a UE specific parameter (in dBm) with 1 dB resolution over the range: [-126dBm, 24dBm]
`• α is cell specific path loss compensation factor (can be set to one to allow full path loss
`compensation) that has 8 values from 0.4 to 1 in steps of 0.1 with one of the possible values being
`zero.which can take on these eight values: {0, 0.4, 0.5,…,1}
`• PL is the downlink pathloss (in dB) calculated in the UE from a RSRP measurement and signalled
`RS transmit power
`mcsΔ
` is signaled (in dB) by RRC ( mcsΔ
`
` table entries can be set to zero)
`
`•
`
`HTC/ZTE EXHIBIT 1009-10
`
`

`

`•
`
`o MCS signaled in each UL scheduling grant
`Δ
`iΔ is a UE specific correction value (in dB) and is included in every Nth UL scheduling
`PUSCH
`grant, (where N can be 1), or jointly coded with other UE specific correction values in a
`TPC-PDCCHPUSCH.
`o The UE attempts to detect a TPC-PDCCHPUSCH on every subframe except when in
`DRX.
`o The TPC command from a UL scheduling grant overrides any command from a
`TPC-PDCCHPUSCH when both are received in a given subframe.
`• KPUSCH is the delay in the PUSCH power control, i.e. a power control command provided in
`downlink subframe i will not impact the PUSCH transmit power in subframes prior to subframe
`i+KPUSCH. Exact value of KPUSCH. is TBD.
`The power control formula and is definedis applied differently dependent on scheduling as given by:
`• Dynamically Scheduled PUSCH
`Δ
`•
`PUSCH
`be 1 .
`][∗f
`)(∗f
`o Function
` represents either accumulation or absolute value, and is signaled
`Δ
`semi-statically via higher layers. When a new value of
` is received,
`
`iΔ is included in every Nth UL scheduling grant, where N can
`
`PUSCH
`
`When a new value of
`
`Δ
`
`PUSCH
`
` is received,
`
` For absolute control
`Δ
`−
`f
`
`( Ki
`[
`
`PUSCH
`
`PUSCH
`
`Δ=
`
`)]
`
`−
`( Ki
`
`
`)
`
`
`
`PUSCH
`
`PUSCH
`
`Δ
`{
`
`PUSCH
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`i
`
`=
`
`m
`
`0
`
` For accumulation control
`Δ
`−
`f
`
` Ki(
`)]
`[
`
`PUSCH
`
`PUSCH
`
`=
`
`
`
` represents either accumulation or current absolute value
`
`
`
`)(∗f
`o NotPersistently scheduled PUSCH
`Δ
` is included in each DL scheduling assignment or jointly coded with other UE
`PUSCH
`specific correction values on a TPC-PDCCHPUSCH
`The UE attempts to detect a TPC-PDCCHPUSCH and a DL scheduling frame on every subframe
`except when in DRX.
`Δ
` from a DL scheduling assignment overrides any command from a TPC-
`The
`PUSCH
`PDCCHPUSCH when both are received in a given subframe.
`][∗f
` Function
` represents accumulation only. When a new value of
`
`Δ
`
`PUSCH
`
` is received,
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`PUSCH
`
`Δ
`{
`
`i
`
`=
`
`Δ
`
`f
`
`[
`
`K-i(
`
`PUSCH
`
`=
`
`)]
`
`PUSCH
`
`m
`0
`• Combined dynamically and persistently scheduled PUSCH
`
`HTC/ZTE EXHIBIT 1009-11
`
`

`

`o When UL grant is configured for accumulation commands (for dynamically
`scheduled PUSCH), the UE combines the accumulation commands received in
`both the SG and the TPC-PDCCHPUSCH
`o When UL grant is configured for absolute commands (for dynamically scheduled
`PUSCH), the UE resets the accumulation immediately after receiving each
`absolute TPC and then combines the absolute with the accumulation
`i
`
`PUSCH
`
`,
`
`TPC
`
`−
`
`PDCCH
`
`−
`Km
`
`(
`
`)}
`
`
`
`PUSCH
`
`Δ
`{
`1
`
`+
`
`=
`jm
`
`Δ
`
`[
`
`f
`
`PUSCH
`
`−
`i K(
`
`
`
`PUSCH
`
`Δ=
`
`)]
`
`PUSCH
`
`
`
`UL,
`
`_
`
`grant
`
`−
`Kj
`
`(
`
`+
`
`)
`
`PUSCH
`
`where j is the subframe index of the last absolute command.
`If no power control step ΔPUSCH (k) is provided on neither an UL scheduling grant or on a TPC-
`PDCCHPUSCH in subframe k, then
`• For absolute control, ΔPUSCH (k) in the equations above should be set to the latest value of
`ΔPUSCH (k-1).
`• For accumulation control, ΔPUSCH (k) in the equations above should be set to zero
` iΔ is included in each DL scheduling assignment or jointly coded with other UE
`specific correction values on a TPC PDCCH
`The UE attempts to detect a TPC PDCCH and a DL scheduling frame on every
`subframe except when in DRX.
`iΔ from a DL scheduling assignment overrides any command from a TPC
`The
`PDCCH when both are received in a given subframe.
`)(∗f
`Function
` represents accumulation only
`
`Open loop component
`The UE first determines the open loop component based on a filtered linear pathloss estimate from the
`serving eNodeB to the UE. The UE continuously measures the instantaneous pathloss based on the DL RSRP
`whose transmit RS power is signaled to the UE. A filtering method, such as moving averaging, is then
`applied to the pathloss measurements. The pathloss is updated in the power control formula after each
`downlink RSRP measurement, multiplied by the signaled value of α and applied in the next transmission.
`Closed loop component
`Δ
` and mcsΔ
`in the transmission in the downlink subframe with a delay of KPUSCH after
`The UE applies
`PUSCH
`receiving either or both commands.
`After reading the MCS signaled in each UL scheduling grant, the UE directly applies mcsΔ
`table signalled by RRC, in the power control formula.
`
`, obtained from a
`
`
`
`5.1.1.2
`
`eNodeB behaviour
`
`f Δ
`[ PUSCH
`]
` to the Tx
`For dynamically scheduled UEs there are two alternatives for applying the correction
`power, accumulated or absolute value. The eNodeB sends to each scheduled UE (or a sub-group of scheduled
`UEs) a power correction factorcommand using multiple command bits in the UL grant (or in the TPC-
`PDCCHPUSCH), where the correction command is determined based on link quality (such as received PSD or
`SINR) of the UL power controlled data channel (and possibly UL sounding reference symbol, if available).
`
`HTC/ZTE EXHIBIT 1009-12
`
`

`

`For absolute control, the power control command may be determined such as
`[
`Δ
`SINR
`Tar
`
`−
`
`ESINR
`
`get
`
`]est
`
`
`
`=
`
`PUSCH
`
`where ESINRest and SINRTarget denote the effective SINR (ESINR) estimate at the receiver and target SINR,
`respectively, of the power controlled channel(s) in dB. [x] denotes a correction value in the correction set
`which is nearest to x. The observed samples at the eNodeB for the ESINR estimation include (some of or all)
`SC-FDMA symbols of the UL power controlled channel(s), which have been received over an averaging
`window. Because slow power control is to be used, a linear block average since the last power control
`command signaling may be used when commands are sent infrequently, or a moving average over a suitable
`window may be used when updates are sent in every grant.
`
`The power control command consists of 2 bits for accumulation control and TBD [2 or 3] bits for absolute
`control, using these formats:
`•
`accumulation control has two sets of 2 bit commands (configured by higher layers): [-1,0,1,3] and [-
`3,-1,1,3]
`absolute control: [x1, x2, x3, …,xn]
`
`•
`
`
`For persistenly scheduled UEs only accumulation control is used. For accumulated control the eNodeB sends
`a command every N grants or in the TPC-PDCCHPUSCH. Assuming block averaging is used, the command is
`derived as follows:
`
`For each reception from the UE, the eNodeB computes an error given by
`]k
`[
`−
`=ke
`SINR
`ESINR
`
`
`
`Tar
`get
`ESINR is the effective SINR for the kth reception since the last command was sent. The
`where
`k
`eNodeB computes a block average over M receptions during N intervals:
`= M
`k M
`/)
`10
`(
` E
`
`
`
`
`
`0.1e
`
` for the physical uplink control channel (PUCCH)
`
`1
`The command sent is obtained as
`
`[
`]E
`Δ
`=
`
`log
`10
`
`
`
`PUSCH
`10
`If the command is sent frequently, e.g. every grant, then a moving average over a window M can be used.
`
`Physical uplink control channel
`5.1.2
`UE behaviour
`5.1.2.1
`In a given downlink subframe n, a power control command ΔPUCCH (n) for PUCCH power control may be
`provided
`- within a downlink assignment on PDCCH
`-
`or on the TPC-PDCCHPUCCH)
`-
`or not at all
`P
`The setting of the UE Transmit power PUCCH
`transmissions is defined by
`=
`P
`P
`log
`10,
`min(
`PUCCH
`max
`
`(
`
`M
`
`10
`
`pucch
`
`+
`
`)
`
`P
`o
`
`_
`
`pucch
`
`+
`
`PL
`
`Δ+
`
`mcs
`
`_
`
`pucch
`
`+
`
`where
`•
`
`pucchM
`
` is the number of assigned resource blocks for the PUCCH
`
`Δ
`{
`
`PUCCH
`
`−
`
`( Ki
`
`))}
`
`
`
`PUCCH
`
`i
`
`=
`
`m
`
`0
`
`HTC/ZTE EXHIBIT 1009-13
`
`

`

`oFunction
`
`•
`
`Δ
`
` is signaled by RRC (
`
`Δ
`
` table entries can be set to zero)
`
`pucch
`mcs _
`pucch
`mcs _
`o MCS is signaled using higher layer signaling
`•
`oP _
` is a UE specific parameter with 1 dB resolution over the range: [-126dBm, 96dBm]
`pucch
`• ΔPUSCH
`jΔ is a UE specific correction value, also referred to as a TPC command, which can have
`one of two sets of 2 bit commands (configured by higher layers): [-1,0,1,3] and [-3,-1,1,3]. ΔPUSCH
`is included in a DL scheduling assignment or sent jointly coded with other UE specific correction
`values on a TPC-PDCCHPUCCHTPC PDCCH.
`o The UE attempts to detect a TPC-PDCCHPUCCH TPC PDCCH and a DL scheduling frame
`on every subframe except when in DRX.
`o The TPC command from a DL scheduling assignment overrides any command from a
`TPC-PDCCHPUCCH TPC PDCCH when both are received in a given subframe.
`)(∗g
` represents accumulation
`Open loop component
`The UE performs the same functions as described for the PUSCH, except that α = 1.
`Closed loop component
`The UE performs the same functions as described for the case of accumulated control for the PUSCH.
`
`eNodeB behaviour
`5.1.2.2
`The eNodeB performs the same functions as described for the case of accumulated control for the PUSCH.
`
`5.1.3
`5.1.2.1
`
`Sounding Reference Symbol
`UE behaviour
`
`P
`srsP for the SRS is set equal to the PUSCH power level PUCCH
`
`P
`pusch
`
`. plus an
`
`The UE Transmit power
`P
`offset
`.
`srs
`offset
`_
`If the time since the last command for the PUSCH exceeds TPUSCH_To , a signaled parameter, there are several
`options for the UE to set its SRS Tx power as follows:
`• Option 1) Applying a power offset relative to the most recent PSD for PUCCH, if the time since the last
`command does not exceed TPUCCH_T1, a signaled parameter, as follows:
`=
`Δ+
`P
`P
`SRS
`PUCCH
`(
`,
`PUCCH
`SRS
`
`control
`
`)
`
`
`
`P
`is the most recent power (or averaged over the recent updates) for PUCCH and
`where
`pusch
`Δcontrol( SRS , PUCCH) represent the PUCCH power offset relative to the Tx power for PUSCCH.
`This offset accounts for the number of assigned resource blocks, the P0 offset and Δmcs.
`
` Option-2) If the time since the last command exceeds TPUCCH_T1 but does not exceed TSRS_To , a signalled
`parameter, the SRS PSD is adjusted based on the pathloss variation between the time before the DTX
`and the time before resuming the UL transmission as follows:
`(
`))1
`=
`+−
`−
`
`)(n
`P
`n
`P
`
` )(nPL
`)1
`(
`SRS
`SRS
`
`
`
`(nPL
`
`−
`
`
`
` •
`
`HTC/ZTE EXHIBIT 1009-14
`
`

`

`where n is the Tx power setting time before resuming the UL SRS transmission and (n-1) is the power setting
`time for the last previous transmission.
`• Option-3) If the time since the last command exceeds TSRS_To, then open-loop power control is applied
`using the procedures described in 5.1.2.1.
`
`
`eNodeB behaviour
`5.1.2.2
`There is no specific action taken by the eNodeB for SRS power control because SRS is controlled via
`PUSCH power control.
`
`
`
`
`
`HTC/ZTE EXHIBIT 1009-15
`
`