11 The Split/Combine and Padding-functions

3GPP44.021Rate adaption on the Mobile Station - Base Station System (MS - BSS) interfaceTS

The split/combine and padding functions shall be used with multislot connections as described below. The Split/Combine function splits/recombines the overall data stream to/from the substreams. The Padding function inserts filling into one of the substreams in cases where the total capacity of the substreams is larger than necessary to achieve the required AIUR.

11.1 Data frame distribution into the substreams/channels by the Split/Combine function

11.1.1 Data frame distribution into the substreams/channels by the Split/Combine function (TCH/F9.6 and TCH/F4.8 channel codings)

a) In the transparent case the Split/Combine-function distributes the V.110-frames into the substreams and recombines the overall data stream from the substreams according to the following rules:

In the overall data stream

1) the frame in position p in substream q precedes the frame in position p in substream q+1,0 q  n-1

2) the frame in position p in substream n-1 precedes the frame in position p+1 in substream 0;

where in the rules above n is the number of substreams.

b) In the non-transparent case the Split/Combine-function distributes the RLP-frames  or the four V.110-frames making up an RLP-frame (Reference: 3GPP TS 48.020, Subclause 15.1)  into channels so that one whole RLP-frame is carried through one channel. Furthermore the RLP-frames are distributed into the available channels so that the resulting delay in the overall data stream is kept as small as possible. The receiving Split/Combine-function recombines the overall data stream according to the inherent RLP-frame numbering, i.e. the N(S)-numbers in the RLP-frame header (3GPP TS 24.022).

11.1.2 Data block distribution into the substreams by the Split/Combine function (TCH/F14.4 channel coding)

a) Transparent services

The Split/Combine-function distributes the user data carried in the 290-bit blocks (Refer to subclause 10.3.2) into the substreams and recombines the overall data stream from the substreams according to the following rules:

In the overall data stream:

1) the data block in position m of multiframe in substream q precedes the data block in position m of multiframe in substream q+1, 0 q  n-1, 0m30.

2) the data block in position m of multiframe in substream n-1 precedes the data block in position m+1 of multiframe in substream 0;

where in the rules above n is the number of substreams.

Figure 2a: Distribution of data frames or data blocks into the substreams in transparent operation

b) Non-transparent services

In the non-transparent operation the Split/Combine-function distributes the RLP-frames into substreams so that one whole RLP-frame is carried through one substream. This means that the two 290-bit air-interface blocks carrying one RLP-frame are transmitted through the same substream. Furthermore the RLP-frames are distributed into the available substreams so that the resulting delay in the overall data stream is kept as small as possible. The receiving Split/Combine-function recombines the overall data stream according to the inherent RLP-frame numbering, i.e. the N(S)-numbers in the RLP-frame header (3GPP TS 24.022).

11.2 Substream numbering in transparent operation

11.2.1 Substream numbering for TCH/F4.8 and TCH/F9.6 channel codings

In transparent multislot data configurations of more than one TCH/F the parallel data substreams between the Split/Combine-functions carry inband substream numbering. The status bits S1, S3, and the X-bit between data bits D12 and D13 (Figures 5 and 6) are used for transferring this substream numbering information (S1 is the MSB and S3 the LSB). The substreams are numbered 0, 1, 2, 3 etc. regardless of the physical channels through which the substreams are transmitted. The highest substream number is one less than the number of physical channels in use at a given time; i.e. the numbering cycle changes when physical channels are either added to or removed from a connection.

The S4-bit is used for frame synchronisation between the parallel substreams. This bit follows a 31-bit PN-sequence of 0000 1001 0110 0111 1100 0110 1110 101. This thirty one bit sequence is used for substream resynchronisation in cases where delay has occurred on one or more substream(s); the position of a frame in a substream can be determined modulo 31 by the values of the S4-bit in a sequence of 5 consecutive frames including the frame in question. Provided that the relative delay between substreams is less than 75 ms (i.e. less than a 15-frame displacement), this and the frame distribution rules given in subclause 11.1 are sufficient to determine the correct order of the frames.

Bits S6, S8, S9, and the other X-bit are used for conveying channel control information according to the relevant terminal adapter function specification.

These rules apply to all multislot data AIURs up to and including 48 kbit/s. When the received 48 kbit/s AIUR is converted into 64 kbit/s rate, the bits extracted from the 60-bit radio interface frames (Figure 5) are mapped into the 32-bit frame format of Figure 13. Here the values for the status bits S1, X, S3, and S4 in the 32-bit frame shall be derived from status bits S6, X, S8, and S9 in the radio-interface frame because status bits S1, X, S3, and S4 in the upper right hand corner of the 60-bit frame have been used for data substream numbering as described above.

In the 11,2 kbit/s frames used for AIURs 56 and 64 kbit/s (Figure 10) the T1, T2, T3 (T1 the MSB and T3 the LSB) are used for carrying the substream numbering as status bits S1, X, and S3 do according to the definition given in the first paragraph of subclause 10.1. Bit T4 is used for carrying the substream synchronisation sequence just as status bit S4 does in the description given in the second paragraph of this subclause.

11.2.2 Substream numbering for TCH/F14.4 and TCH/F28.8 channel codings

Bit M1 carries the multiframe sequence 0000 1001 0110 0111 1100 0110 1110 101. The number of the substream in which a multiframe is sent is carried four times in a 31-bit period of the M2-sequence. In the three-bit number code the bit in the lowest bit position is the MSB (See table in subclause 10.3.1).

11.3 Substream Synchronisation

Two interfaces are involved in the TAF regarding the need for the data frame synchronisation, i.e the TAF/multichannel interface and the TAF/TE interface.

The Split/Combine function is responsible for controlling the synchronisation and resynchronisation procedures as described in 3GPP TS 27.001.

11.4 Void

11.5 Padding Functions

11.5.1 Padding for TCH/F9.6 and TCH/F4.8

When the required AIUR is not a multiple of the rates supported by TCH/F4.8 or TCH/F9.6, padding is used for producing the required AIUR.

To achieve the required AIUR the data bits are distributed across the substreams 1 to n as follows:

– Substream(s) 1 (to n-1) carry multiples of the rate supported by the channel coding used.

– Substream n carries the remaining amount of data bits required to achieve the required AIUR. The remainder of data stream n carries padding bits set to binary value ‘1’.

11.5.1.1 Padding for AIUR 14,4 kbit/s:

The frame of the lower substream number carries full 9,6 kbit/s. The frame of the higher substream number carries 4,8 kbit/s of user data in bit positions D1-D24 while bit positions D25-D48 are inserted with binary "1"s. (Figure 15)

11.5.1.2 Padding for AIUR 64 kbit/s:

The frames numbered 1-5 carry full 11,2 kbit/s. Frame number 6 carries 8.0 kbit/s of user data in bit positions D1-D40 while bit positions D41-D56 are inserted with binary "1"s. (Figure 16)

11.5.2 Padding for TCH/F14.4 channel coding

11.5.2.1 Padding for AIURs up to 38,4 kbit/s

Padding is not necessary for AIURs 14,4 and 28,8 kbit/s.

For AIUR 38.4 kbit/s the 290-bit blocks in the substreams of the lower substream numbers carry 288 user data bits while the 290-bit blocks in the substream of the highest substream number carries 192 user data bits; this means that five of the eight 36-bit frames making up the block carry 36 user data bits whereas the sixth frame carries 12 user data bits (536+12). Frames seven, eight, and the rest of the sixth frame are padded with ‘1’s.

11.5.2.2 Padding for AIURs of 48 kbit/s

The 290-bit blocks in the highest numbered substream carry 96 user data bits (236+24).

The M2-bit sequences are used, for the applicable parts, as described in subclauses 10.3.1.1.

11.5.2.3 Padding for AIURs of 56 kbit/s

The 290-bit blocks in the highest numbered substream carry 256 user data bits (736+4).

The M2-bit sequences are used, for the applicable parts, as described in subclauses 10.3.1.1.

11.5.2.4 Padding for AIURs of 64 kbit/s

The 290-bit blocks in the highest numbered substream carry 128 user data bits (336+20).

The M2-bit sequences are used, for the applicable parts, as described in subclauses 10.3.1.1.

11.6 Handling of the E1-E3 bits in multislot operation

Between the Split / Combine functions the substreams carry the code associated with the substream rate as defined in Figure 4. When the substreams are combined the code is set to correspond to the overall AIUR according to Figure 4.