`Warsaw, Poland, June 30 – July 4, 2008
`
`Source:
`Title:
`Agenda Item:
`Document for:
`
`
`Panasonic
`Consideration on Multicarrier Transmission scheme for LTE-Adv uplink
`12
`Discussion
`
`R1-082398
`
`Introduction
`1.
`In RAN1#53, several contributions proposed to study multicarrier transmission, i.e. Clustered SC-FDMA,
`OFDMA, in addition to the legacy SC-FDMA for LTE-Advanced uplink [1] –[9] .
`In this paper, we review the properties of the proposed multicarrier transmission schemes and discuss selection
`criteria for the multicarrier transmission schemes.
`2. Proposed multicarrier transmission schemes
`As discussed in [7] [8] , one of the main reasons for supporting non-continuous resource block allocation and
`transmission for UE with wider bandwidth capability (i.e. beyond 20MHz) is to accommodate the situation that
`PUCCHs are transmitted at the middle of the transmission bandwidth for backward compatibility as shown in
`Figure 1.
`In addition, more flexible PUSCH assignment and PUSCH/PUCCH multiplexing compare to the legacy SC-
`FDMA would be beneficial for efficient uplink data transmissions [2] [4] [9] .
`Therefore, at least non-continuous resource block allocation and transmission capability has to be supported for
`LTE-Advanced uplink.
`
`
`PUCCH
`
`e.g. 40MHz
`PUCCH
`
`PUCCH
`
`UE#1
`
`UE#1
`
`non-continuous resource block allocation for a UE
`
`Figure 1 non-continuous resource block allocation due to PUCCH transmission.
`
`frequency
`
`
`In order to support non-continuous resource block allocation and transmission, two schemes having different
`properties were proposed as follows;
`Clustered SC-FDMA (Figure 2) [2] [7] [8] :
`• Spectrum element after DFT precoding is divided into two or more parts called as "cluster", and each cluster
`is mapped to allocated resource blocks.
`Pros
`-
`
`-
`
`CM property of Clustered SC-FDMA is lower than OFDM and CM depends on the number of clusters,
`cluster size and cluster placement,
`Less complexity, implementation and test efforts because most of components of the legacy SC-
`FDMA except sub-carrier mapping can be reused.
`OFDMA (Figure 3) [1] - [4] [8] [9] :
`• OFDMA transmission can be carried out by skipping DFT precoding. PUSCH symbols to be transmitted are
`directly mapped to allocated resource elements, i.e. sub-carriers.
`Pros
`-
`
`Better performance for higher order modulation [1] and MIMO with advanced receiver e.g. MLD
`based receiver [8] ,
`- More flexible PUSCH assignment [2] [4]
`
`-1/3-
`
`IPR2018-01476
`Apple v. INVT
`INVT Exhibit 2006 - Page 1
`
`
`
`
`
`
`
`S/P
`
`DFT
`
`IFFT
`
`P/S
`
`CP
`insertion
`
`Figure 2 Clustered SC-FDMA.
`
`DFT
`
`S/P
`
`IFFT
`
`P/S
`
`CP
`insertion
`
`freq.
`
`
`
`freq.
`
`
`
`Figure 3 OFDMA.
`
`
`3. Discussion
`Since Clustered SC-FDMA and OFDMA have quite different properties from each other, the target usage would
`be different. Some contributions are proposing to support both Clustered SC-FDMA and OFDMA depending on
`usages. Supporting both Clustered SC-FDMA and OFDMA would require much additional implementation and
`test efforts. Therefore, current our preference is to choose one of them for additional uplink multiple access
`scheme to minimize additional standardization, implementation and test efforts.
`In contributions proposing Clustered SC-FDMA[2] [7] [8] , the followings seemed to be assumed;
`- single codeword transmission for PUSCH
`- transmission power limited situation
`If these assumptions are really target condition of the optimization point for LTE-Advanced, Clustered SC-
`FDMA would become attractive solution, although it is necessary to confirm whether OFDMA provides
`significant performance improvement.
`Hence, before starting detailed comparison between Clustered SC-FDMA and OFDMA, we need to discuss the
`target conditions where multicarrier transmission will be mainly used. We see the need of the following
`discussion.
`• Importance of CM property for multicarrier transmission
`• Number of codewords supported for uplink transmission in frequency domain
`
`
`Importance of CM property for multicarrier transmission
`Low Cubic Metric (CM) was one of the important properties for the uplink multiple access design in LTE to
`increase coverage. However, it is not clear whether low CM property is still important criterion for the selection
`of multicarrier transmission scheme for LTE-Advanced. The difference of the CM values between Clustered SC-
`FDMA and OFDMA is up to around 2.4dB[7] . As the optimization point of the performance for LTE-Advanced
`is for relatively small cells such as case 1 (ISD=500m), the most of the situation might not be transmission
`power limited condition.
`If non-continuous resource allocation under transmission power limited environment is really required, low CM
`property will be one of the most important criteria for the selection of multicarrier transmission scheme.
`Meanwhile, if non-continuous resource allocation is mainly used under non-power limited environment,
`performance and/or complexity should be the most important criteria for the selection since the legacy SC-
`FDMA can still be used for UEs with transmit power limited situation.
`Therefore, we should make clear whether non-continuous transmission with low CM property is really required
`from the optimization point of view.
`
`
`
`-2/3-
`
`IPR2018-01476
`Apple v. INVT
`INVT Exhibit 2006 - Page 2
`
`
`
`
`Number of codewords supported in frequency domain
`Contributions proposing OFDMA and/or Clustered SC-FDMA seem to assume multicarrier transmission only
`for PUSCH and also assume single codeword transmission even for wider system operating bandwidth beyond
`20MHz.
`If we keep single codeword transmission for PUSCH, Clustered SC-FDMA would achieve less complexity,
`implementation and test efforts compared to OFDMA as discussed in [7] . If we support multiple codewords
`transmission for PUSCH, for instance, in wider system bandwidth beyond 20MHz, OFDMA would become
`simpler solution than Clustered SC-FDMA because Clustered SC-FDMA has to equip additional DFT precoder
`as illustrated in Figure 4.
`Therefore, we think OFDMA is more flexible for both single and multi codeword transmission with less
`complexity if multiple codeword is supported. However, further discussion on the number of codewords in
`frequency domain for uplink transmission would be necessary.
`
`
`CW#1
`
`S/P
`
`DFT
`
`CW#2
`
`S/P
`
`DFT
`
`IFFT
`
`P/S
`
`CP
`insertion
`
`PUCCH
`
`UE#1 CW#1
`
`e.g. 40MHz
`PUCCH
`
`UE#1 CW#2
`
`PUCCH
`
`non-continuous resource block allocation for a UE
`(a) In case of Clustered SC-FDMA.
`
`frequency
`
`CW#1
`
`S/P
`
`IFFT
`
`P/S
`
`CP
`insertion
`
`PUCCH
`
`UE#1 CW#1
`
`e.g. 40MHz
`PUCCH
`
`UE#1 CW#2
`
`PUCCH
`
`
`
`
`
`CW#2
`
`S/P
`
`frequency
`
`non-continuous resource block allocation for a UE
`(b) In case of OFDMA.
`Figure 4 Example illustration of multiple codewords transmission in frequency domain.
`
`
`4. Conclusion
`We discussed the properties of the proposed multicarrier transmission schemes for uplink, i.e. Clustered SC-
`FDMA and OFDMA. Our current view is that either Clustered SC-FDMA or OFDMA should be selected to
`support flexible resource allocation without too much additional implementation and test efforts. However,
`further discussion of the following items would be necessary for the down selection.
`• Importance of CM property for multicarrier transmission
`• Number of codewords supported for uplink transmission in frequency domain
`
`
`References
`[1] R1-081722 Samsung, "Future 3GPP Radio Technologies for LTE-Advanced"
`[2] R1-081752 NEC, "Proposals on PHY related aspects in LTE Advanced"
`[3] R1-081773 ZTE, "Technical points for LTE-advanced"
`[4] R1-081791 Panasonic, "Technical proposals and considerations for LTE advanced"
`[5] R1-081809 LG Electronics, "On the Consideration of Technical Candidates for LTE-advanced"
`[6] R1-081838 Huawei, "Physical layer technologies for LTE-Advanced"
`[7] R1-081842 Nokia Siemens Networks, Nokia, "LTE-A proposals for evolution"
`[8] R1-081948 NTT DoCoMo, "Proposals for LTE-Advanced Technologies"
`[9] R1-081957 Qualcomm Europe "Categorization of technical proposals for the PHY layer of LTE-A"
`
`
`-3/3-
`
`IPR2018-01476
`Apple v. INVT
`INVT Exhibit 2006 - Page 3
`
`