`RX-3185
`
`3GPP TSG RAN WG 1 #23 Meeting
`January 8-11, Espoo, Finland, 2002
`
`TSGR1#(23)02-O018
`
`Agenda Item:
`
`5.2
`
`Source:
`
`Nokia
`
`Title:
`
`Compact signalling of multi-code allocation for HSDPA, version 2
`
`Document for: Discussion and approval
`
`This document is contributed to both WG1 and WG2 meetings and includes a text proposal for WG2
`TS. The approach was already discussed in WG1 [7] but here we include a further clarification as
`well as a text proposal for version 5. O. 0 of TS 25. 308.
`
`In order to reduce the signaling overhead it is important to find bit-efficient ways of signaling while
`maintaining performance and flexibility. In this contribution, we propose a method to do efficient
`downlink signaling of multi-code allocation in high speed downlink packet access. The method is very
`simple from a Node B architecture point of view and allows the Node B to allocate any number of
`codes at any possible offset (128 different combinations are supported using a 7 bit self-decodable
`representation).
`
`1 Introduction
`Using the Shared Control Channel High Speed (SCCH-HS), the Node B needs to signal to the UE
`exactly how- many multi-codes have been allocated and at which offset the set of codes begin (all
`defined at spreading factor 16 level). The current working assumption is that only clusters of
`consecutive codes are allocated to one user at a time. However, any number of codes and any number
`of code offsets are possible within the physical limits imposed by the code tree. This is illustrated in
`Figure 1. The number of multi-codes currently allocated is denoted by m and the code offset is denoted
`by A.
`
`Code ’0’ is
`
`rceoSmemWoe~ f°r
`
`channels
`
`A=7
`
`m=5
`
`Code offset 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
`
`Figure 1 - Example code allocation to HS-DSCH with m codes located at code-oflg’et A.
`
`SF=16
`
`In general, up to 15 multi-codes will be supported by the most capable UEs and for the single-code
`case there is up to 15 different code offsets possible since a single code (offset 0) is reserved to the
`common channels P-CPICH/P-CCPCH [4]. Using specific signaling of the offset and the number of
`codes would require 2x4=8 bits. However, when noting the dependence between the code offset and
`the number of multi-codes (e.g. when using 15 multi-codes only one code offset is possible), it can be
`calculated that only 120 different possible combinations exist. Using a lookup table approach, these
`combinations can be represented by using only 7 bits. Hence, in WG1 meeting #22, it was decided that
`7 bits axe allocated to represent the possible multi-code representations allowing up to 128 different
`representations.
`
`N K8681TC015910670
` Ex. 2006-0001
`
`InterDigital Tech. Corp.
`Exhibit 2006
`ZTE Corp. v. InterDigital Tech. Corp.
`IPR2014-00275
`
`
`
`
`3GPP TSG RAN WG 1 #23 Meeting
`January 8-11, Espoo, Finland, 2002
`
`TSGR1#(23)02-O018
`
`In this contribution, we suggest a method that has the following advantages:
`
`¯
`
`It only uses 7 bits but allows full flexibility in terms of code allocation and code offsets. This way,
`we ensure high spectral efficiency potential as well as high RRM flexibility (QoS, code/time
`multiplexing) for HS-DSCH.
`¯
`It is an absolute approach that gives direct signaling of number of multi-codes and code offset.
`¯ The method is on-the-fly encodable and self-decodable at the Node B and UE sides respectively,
`thereby alleviating the need for lookup tables for code allocation.
`¯ The signaling is consistent and avoids the use of conditioned signaling (e. g. signaling only
`possible code offsets).
`
`2 Multi-code representation
`The method is based on clustering of multi-codes which in total has a number of possible offset
`combinations of 15. In this sense, the 1 multi-code and 15 multi-code situations are grouped together
`as one, the 2 multi-code situation is grouped with the 14 multi-code operation, and so forth. This way,
`a total of 8 groups axe formed (1/15, 2/14, 3/13, 4/12, 5/11, 6/10, 7/9, and 8/8) which can be
`represented by 3-bit signaling. The next issue is then to identify which of the two code situations is
`active and what is the exact code offset. This aspect is achieved by utilizing that the total number of
`code-offsets is 16 for each cluster allowing a 4-bit representation. The encoding algorithm can be
`described as follows:
`
`Codeword needed to signal m multi-codes with code-offset A:
`(Code group indicator)
`Three first bits: CW1 = min(m, 16-m)-I
`Four last bits: CW2 = IA-l-(m>7)*151
`(Tree offset indicator)
`
`Where (m>7) is either 1 or 0 depending on whether the condition is fulfilled or not - true is denoted by
`1 and visa versa. As an example, the codeword needed to signal 12 multi-codes with code-offset 2 is:
`
`CW1 = min(12,4)-I = 3 =’011’
`CW2 = 12-1-(12>7)’151 = 14 =’1110’
`
`Such that the total codeword becomes ’0111110’. The total encoding matrix for all possible
`combinations is given in Figure 2.
`
`One advantage of the method, is that the decoding can be done by a simple algorithm at the UE side:
`
`Extraction of number of multi-codes (m) and code-offset (A) from 7-bit codeword (CWI+CW2):
`m
`= ICWI+I-(CW2 _> (15-CW1))*161
`A
`= ICW2+l-(m _> 8)*171
`
`For the example shown above we have
`
`m
`
`: 13+1-(14 _> 12)*161 : 12
`: 114+1-(12 > 7)*171 : 2
`
`Which corresponds to the original signaled values.
`
`N K8681TC015910671
` Ex. 2006-0002
`
`
`
`3GPP TSG RAN WG 1 #23 Meeting
`January 8-11, Espoo, Finland, 2002
`
`TSGR1#(23)02-O018
`
`Tree offset indicator (4 bits)
`0 1 2 3 4 5 6 7 8 0 1(3 11 12 13 14 15
`
`Decoding notation
`
`J Number of
`multi-codes
`Offset from
`left/right in code
`tree (SF=16)
`
`0(1/15)
`
`1(2/14)
`
`2 (3/13)
`
`3 (4/12)
`
`4 (5/11)
`
`5 (6/10)
`
`6 (7/9)
`
`7 (8/8)
`
`Figure 2 - Total encoding matrix for signaling multi-code allocation in HSDPA.
`
`In traditional computer pseudo language, the encoding and decoding algorithms can be represented as:
`
`Encoding:
`if(m<8)
`BEGIN
`
`CWI=m-1
`CW2=A-1
`
`END
`else
`BEGIN
`
`END
`
`CWI=15-m
`CW2= 16-A
`
`Decoding:
`m = CWI+I
`A = CW2+I
`if(A > 16-m)
`BEGIN
`
`A- 16-CW2
`m = 16-(CWI+I)
`
`END
`
`N K8681TC015910672
` Ex. 2006-0003
`
`
`
`3GPP TSG RAN WG 1 #23 Meeting
`January 8-11, Espoo, Finland, 2002
`3 Recommendation forstandardisation
`Based on the findings presented in this document, we make the following recommendations for further
`HSDPA standardization work:
`
`TSGR1#(23)02-O018
`
`¯
`
`It is recommended that full multi-code flexibility be maintained such that high HS-DSCH spectral
`efficiency and RRM flexibility are retained.
`
`¯ Compared to standaxd 8-bit, 7-bit mapping table approaches, or conditioned signaling, it is
`recommended to use the proposed 7-bit self-decodable and consistent method for HS-DSCH
`multi-code signaling.
`
`4 References
`[1] R1-01-0874, "DL Signalling for HSDPA", Samsung
`
`[2] R 1-01-0619 (R2-011150), "Signalling requirements for HS-D SCH", Ericsson.
`
`[3] R2-01-1177, "HSDPA related signalling parameters in downlink, version 2", Nokia.
`
`[4] 3GPP TS 25.213, "Spreading and modulation (FDD)."
`
`[5] R1-01-1105, "Draft minutes of RAN WG1 HSDPA Ad Hoc, including joint sessions with TSG
`RAN WG2", RAN WG1 chairman and secretary.
`[6] R1-01-1030, "Structure of the downlink shared control channel for HSDPA", Siemens.
`[7] R 1-01-1184, "Compact signalling of multi-code allocation for HSDPA", Nokia.
`
`5 Text proposal for 25.308
`The current version 5.0.0 of 25.308 only contains the following description related to multi-code
`signaling:
`
`.................................................... subsection 9.1.4 in TS 25.308
`
`8.1.3 Code channels in case of code multiplexing (FDD only)
`This identifies to the UE (or UEs) flae codes it (flaey) should receive and decode.
`
`In annex 1 a new proposal is listed.
`
`N K8681TC015910673
` Ex. 2006-0004
`
`
`
`3GPP TSG RAN WG 1 #23 Meeting
`January 8-11, Espoo, Finland, 2002
`
`TSGR1#(23)02-O018
`
`Annex 1. text proposal for TS25.308, v.5.0.O, subsection 8.1.3
`
`8.1.3 Code channels in case of code multiplexing (FDD only)
`This identifies to the UE (or UEs) flae codes it (flaey) should receive and decode. The parameters to be signaled
`are the number of multi-codes (m) and the code offset in the code tree (A). Code 0 cannot be allocated to HSDPA
`since it is partially reserved to P-CPICH. Hence, m is an integer from 1-15 and the code offset is an integer from
`1-15.
`
`Code ’0’ is ~
`
`A=7 m=5
`
`Code offset 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
`
`Figure 3 - Example code allocation to HS-DSCH with m codes located at code-offSet
`
`SF=16
`
`The number of multi-codes (m) and the code offset (A) are compressed into 7 bits using the following encoding
`mechanism:
`
`Three firstbits: CW1 = min(m,16-m)-i
`Four last bits:
`CW2 = IA-l-(m>7)*15l
`
`(Code group indicator)
`(Tree offset indicator)
`
`The expression "(m>7)" evaluates to 1 if the expression is true and 0 if it is false. The configuration of the multi-
`code field is shown in Figure 4.
`
`MSB
`
`LSB
`
`cwl
`
`cw2
`
`~ 3 bits ~ 4 bits ~
`Figure 4 - 7 bit codeword used to represent multi-code allocation.
`
`N K8681TC015910674
` Ex. 2006-0005