3GPP TSG RAN WG1 #50bis Meeting
`Shanghai, China, October 08-12, 2007
`
`
`
`Nokia Siemens Networks, Nokia
`Source:
`Closed loop power control corrections for PUSCH
`Title:
`Agenda item: 6.4.2
`Document for: Discussion/Decision
`
`R1-074347
`
`1. Introduction
`In this contribution we address the meaning of closed loop power corrections for the PUSCH. We basically propose to
`only have specified accumulated power control corrections in order maximize the power control dynamic range from
`closed loop corrections, while minimizing the number of power control options and bits for power control in the UL
`grant (PDCCH).
`
`2. Interpretation of closed power control corrections
`The agreement on the power control formula for PUSCH from [1] is summarized below:
`
`PC formula: P = min ( Pmax , 10 log M + Po + α x PL + delta_mcs + f(delta_i))
`o UE obeys the power setting formulation based on the parameters signaled by the network
`o M is the number of assigned RBs (based on UL grant)
`o Po is a cell specific parameter that is broadcasted (default value)
`o α is cell specific path loss compensation factor (can be set to one to allow full path loss compensation)
`o PL is downlink pathloss calculated in the UE
`o delta_mcs is signaled by RRC (table entries can be set to zero)
` MCS signaled in UL grant
`o delta_i is UE specific correction value included in the UL grant
` Function f(*) signaled via higher layers
` Only two possibilities
` Accumulated vs. absolute value
`
`-
`
`
`
`As listed above, one of the open issues is the exact interpretation of the closed loop power control commands, as well as
`how many bits are used for each power control command. For the case with “absolute value”, the function, f(delta_i),
`only depends on the latest received closed loop correction (delta_i). If N bits are used for signaling the closed loop
`correction, then f(delta_i) can take log2(N) different values. As example, if N=2, then one possibility is to configure
`f(delta_i) to the following values; [-3dB, -2 dB, 1 dB, 3 dB].
`
`If “accumulated value” is assumed, then the value of f(delta_i) depends also on previously received closed loop
`corrections. As an example, if we assume only a 1-bit closed loop power correction, then we have
`
`f(delta_i)=f_old+P_step*delta_i,
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`(1)
`
`where f_old is the old value of f(delta_i), P_step is the power control step size, and delta_i is taken values of -1 and +1
`depending on the value of the received closed loop correction bit. Hence, with this approach the power is increased or
`decreased by P_step decibels whenever a new closed loop correction is received by the UE. Notice that the expression
`in (1) is just a simple example, which is easily extendable to the more general cases where different step-sizes are used
`for power-up and power-down, or to cases with multiple step up/down sizes in case several bits are used for each closed
`loop correction.
`
` The advantage of using the approach with “accumulated values” is that 1-2 bits is estimated to be sufficient for the
`closed loop power correction. By sending several closed loop power control corrections, the UE Tx power can be
`
`HTC CORPORATION, ET AL. v. CELLULAR COMMUNICATIONS EQUIPMENT LLC
`CCE EX2011 (R1-074347) – 001
`IPR2017-01508
`
`

`

`adjusted over a larger dynamic range. On the contrary, using the approach with “absolute value”, the dynamic range and
`granularity from using closed loop commands is hard limited by the number of bits for each closed loop command, as
`well as the set values for f(delta_i).
`
`The need for additional UE transmit power adjustments via closed loop power corrections is in practice expected to
`depend on many factors such as; QoS requirements for the users, cell load, interference from other cells, etc.. It is
`therefore desirable to have standardized a scheme which offers; (i) high Tx power dynamic range from using closed
`loop corrections, (ii) reasonable power control granularity, and (iii) a low number of bits from sending closed loop
`corrections. The scheme which best meets these design goals seem to be the solution with “accumulated values”, as it
`only requires few number of bits per closed loop correction (say 1-2 bits), it offers high dynamic range by sending
`several closed loop corrections, and it offers a reasonable power control granularity by choosing an appropriate power
`control step size.
`
`As a specific proposal, we therefore suggest to only use 2 bits for closed loop corrections. As an example, the power
`control granularity for closed loop commands could be in steps of +/-1 and +/-3 dB.
`
`3. Conclusions
`In this contribution we have addressed the meaning of closed loop power control corrections for the PUSCH. Our
`recommendation is to only standardize accumulated closed loop corrections, where the value of the closed loop
`adjustment depends also on the previously received closed loop corrections. We furthermore suggest using 2-bits for
`each closed loop correction, assuming a power control granularity (step-size) of +/-1 and +/-3 dB.
`
`4. References
`[1]
`3GPP Tdoc, R1-073224, “Way Forward on Power Control of PUSCH”, June 2007 (Orlando meeting)
`
`
`
`HTC CORPORATION, ET AL. v. CELLULAR COMMUNICATIONS EQUIPMENT LLC
`CCE EX2011 (R1-074347) – 002
`IPR2017-01508
`
`