The overall structure of Radio Bearer Configurtion for WCDMA and TDSCDMA are almost same. The major difference between WCDMA and TDSCDMA is physical layer (DPCH Uplink, DPCH Downlink). So in this page, I would describe mostly on physical layer configuration of TDSCDMA RAB. For the layers above the physical layer, refer to WCDMA RAB Overview and WCDMA R99 Radio Bearer.

Since most of the details above the physical layer is almost same as WCDMA, I would put some bullets here only for physical layer. TD-SCDMA is the 1.28 Mcps option of UTRA TDD. So the physical layer rules on this page come from the 1.28 Mcps clauses of 25.221 and 25.222, not from the FDD specifications. Let's go from the frame to the spreading factor of each channel. After that, we put all channels into one slot plan and finish with a rule for changing a bearer.

• Frame Structure
• Spreading Factor
• Master Plan for Physical Layer
• Guidelines for RAB Modification
• Examples of common Radio Bearer with focus on Physical Layer Configuration
• Reference

Frame Structure

Before you look into the details of Radio Bearer, you need to have very detailed understanding of Frame Structure. This is pretty big topics. So I posted it in separate page titled as 'Frame Structure'. Here is the short version that the rest of this page relies on.

A 10 ms radio frame of 1.28 Mcps TDD is split into two sub-frames of 5 ms, and both sub-frames have the same structure. Each sub-frame has seven traffic time slots, TS0 to TS6, of 864 chips each. Between TS0 and TS1 there are three special fields: DwPTS of 96 chips, the main guard period GP of 96 chips, and UpPTS of 160 chips. Together they add up to 6400 chips, which is exactly 5 ms at 1.28 Mcps.

TS0 is always a downlink slot, and TS1 is always an uplink slot. A switching point divides the other slots between uplink and downlink, so one sub-frame has two switching points in total. This is why a TD-SCDMA bearer is described by time slots and codes, not by codes alone. The network must tell the UE which slot, which channelisation code and which slot format to use in each direction.

A traffic burst in one slot has two data fields of 352 chips, a midamble of 144 chips and a guard period of 16 chips. The spreading factor decides how many symbols fit into those two data fields. So the next section, on the spreading factor, is really about how many bits one code carries in one slot.

• A TD-SCDMA sub-frame is 5 ms long : two sub-frames make one 10 ms radio frame, and both have the same layout.
• TS0 is always downlink and TS1 is always uplink : the switching point decides the direction of TS2 to TS6.
• A bearer is allocated in slots and codes : each physical channel needs a time slot, a channelisation code and a slot format.

Spreading Factor

Most of details in this section is based on 3GPP 25.221 V8.7.0 (2010-02). The same rules are still in 25.221 v19.0.0, in clause 5A for the 1.28 Mcps option and in clause 5.2.1, which clause 5A reuses for spreading. The spreading factor sets the number of symbols in a burst, so it sets the number of bits that one code carries in one slot. Keep in mind that SF = 1 gives the whole code space of a slot to a single code.

DPCH - Downlink

You can use only two different SF for downlink (SF = 16, SF = 1). Refer to 25.221 5.2.1.1 Spreading for Downlink Physical Channels for details. SF = 1 is one code that fills the slot, while SF = 16 lets the Node B split the slot among up to 16 codes. The choice also changes the slot format of the bearer.

• With SF = 16
• Multiple parallel physical channels can be used to support higher data rates
• These parallel physical channels shall be transmitted using different channelisation codes
• With SF = 1
• Only single code can be used

DPCH - Uplink

You can use any spreading factor from 16 to 1. (Refer to 25.221 5.2.1.2 Spreading for Uplink Physical Channels for details). The UE may use SFmin all the time, or it may raise the spreading factor by itself when the current TFC needs fewer bits. The network decides which of the two options applies.

• For each physical channel an individual minimum spreading factor SFmin is transmitted by means of the higher layers
• For multicode transmission a UE shall use a maximum of two physical channels per timeslot simultaneously. These two parallel physical channels shall be transmitted using different channelisation codes

P-CCPCH

You can use only one spreading factor (SF = 16). The P-CCPCH carries the BCH, and its position is fixed in 1.28 Mcps TDD. It always sits on the first two codes of TS0, so a UE can read the BCH before the network has configured anything.

• P-CCPCH1 can use only one channelization Code C(k=1, Q=16)
• P-CCPCH2 can use only one channelization Code C(k=2, Q=16)

S-CCPCH

You can use only one spreading factor (SF = 16) except MBSFN slots. The S-CCPCH carries the PCH and the FACH, and its time slot and codes are broadcast on the BCH. In an MBSFN time slot of 1.28 Mcps TDD, 25.221 v19.0.0 allows SF = 1, 2 or 16 instead.

FPACH

You can use only one spreading factor (SF = 16). The FPACH is the Node B answer to a SYNC-UL code on the UpPCH. One burst carries the acknowledgement together with a timing and power adjustment, and its code and time slot are broadcast on the BCH.

PRACH

You can use SF=16, SF=8 or SF=4 and SIB (BCH message) will specify which SF should be used. The UE sends the PRACH only after the FPACH has given it a timing correction. So the PRACH already arrives aligned with the Node B timing, unlike the UpPCH before it.

HS-PDSCH

You can use SF = 16 or SF = 1 if it is not for MIMO. The HS-PDSCH carries the HS-DSCH, and its slot formats are listed on the HSDPA Slot Format page. With SF = 16, every 1.28 Mcps TDD HS-DSCH category in 25.306 can receive up to 16 codes in one time slot.

• For the UEs configured in MIMO mode, if SF=16 is configured by higher layers to be not supported for dual stream transmission, the HS-PDSCH shall use spreading factor SF=1 only. Otherwise, the HS-PDSCH shall use either spreading factor SF = 16 or SF=1

HS-SCCH

You can use only one spreading factor (SF = 16). The HS-SCCH tells the UE which HS-PDSCH slots and codes carry its data. In 1.28 Mcps TDD it uses two physical channels, HS-SCCH1 and HS-SCCH2, and the HS-SCCH page shows its fields.

HS - SICH

You can use only one spreading factor (SF = 16) with some exception on MIMO case. The HS-SICH is the uplink feedback channel of HSDPA. It carries the ACK or NACK and the channel quality report back to the Node B.

• When MIMO dual-stream is transmitted, the HS-SICH shall use spreading factor SF=8 which shall utilize an additional SF=16 channelisation code along the branch with the higher code numbering of the allowed OVSF sub tree.

PLCCH

You can use only one spreading factor (SF = 16). The PLCCH carries only TPC and SS commands, and one PLCCH can serve several UEs. The network uses it, for example, for a UE that has HS-DSCH but no downlink DPCH to carry these commands.

• Speading Code is specified by higher layer

E-PUCH

You can use SF = 1,2,4,8,16 but you have to use the same SF for all E-PUCH within the same TTI. The E-PUCH carries the E-DCH data of HSUPA, together with the E-UCCH and a TPC field.

For scheduled transmissions, E-PUCHs use the spreading factor specified by CRRI on E-AGCH.

E-RUCCH

You can use SF = 16 or SF = 8. The E-RUCCH carries E-DCH uplink control signalling when the UE has no E-PUCH resource. It is mapped to the random access physical resources, and that is why its codes follow the PRACH.

The set of admissible spreading codes used on the E-RUCCH are based on the spreading codes of PRACH

E-AGCH

You can use only one spreading factor (SF = 16). The E-AGCH carries the absolute grant of HSUPA. The grant gives the code resource through the CRRI, and that is how the network sets the E-PUCH spreading factor described above.

E-HICH

You can use only one spreading factor (SF = 16). The E-HICH carries the HARQ acknowledgement for E-DCH transmissions. Several users share one channelisation code through different signature sequences, as the bullet below says.

• Multiple users’ signature sequences (including the inserted spare bits) sharing the same channelisation code are combined

Summary of Spreading Factors

Let's put all fourteen channels side by side before we place them in a slot plan. The table below lists the spreading factors that 25.221 v19.0.0 allows for each channel in 1.28 Mcps TDD.

• Most downlink channels use SF = 16 only : among the channels on this page, only the DPCH and the HS-PDSCH can also use SF = 1 outside MBSFN slots.
• The uplink data channels have the full range : DPCH and E-PUCH can use any spreading factor from 16 down to 1.
• The random access channels have their own set : PRACH uses 16, 8 or 4, and E-RUCCH uses 16 or 8.

Master Plan for Physical Layer

Once the spreading factors are known, the next step is to place every channel into the slots of a sub-frame. The master plan below does that for one cell and one UE with a dedicated bearer. It shows the slot, the code and the slot format of each channel, and the message that configures it.

Each column is a time slot of one sub-frame, and each row is one of the 16 codes at SF = 16. The narrow columns DW, G and P are DwPTS, GP and UpPTS. The colour of a cell names the channel, and the text in the cell is its slot format. The legend also names the message that carries each configuration. SIB5 configures the common channels, RRC Connection Setup configures the SDCCH on DPCH, and Radio Bearer Setup configures the DTCH on DPCH.

Channel plan of a 1.28 Mcps TDD cell over two sub-frames. The common channels sit in TS0 and TS1, and the dedicated bearer uses TS3 in the uplink and TS6 in the downlink.

• TS0 carries the common downlink channels : P-CCPCH on codes 1 and 2, PICH on codes 3 and 4, S-CCPCH on codes 5 and 6, and FPACH on code 15.
• TS1 carries the PRACH : the PRACH with format 10 takes the space of codes 15 and 16, which is one code at SF = 8.
• The DTCH uses TS3 in the uplink : format 48 takes the space of codes 1 to 8, which is one code at SF = 2.
• The DTCH uses TS6 in the downlink : codes 1 to 8 are used, with format 8 on code 1 and format 0 on codes 2 to 8.
• The SDCCH uses code 16 with format 7 : the uplink one is in TS3 of sub-frame 0, and the downlink one is in TS6 of sub-frame 1.
• TS2, TS4 and TS5 are free : this plan leaves them for other UEs or other bearers.

Guidelines for RAB Modification

One of the common request that I was getting was to create a new radio bearer or change a certain parameter of existing bearer. Unfortunately, this kind of request comes with a lot of missing information. If they provide full details of radio bearer parameters as described in 3GPP sample radio bearer definition, it would be relatively easy (too be honest.. it is not easy even with all those information.. but at least it is possible), but if they put out those request with a lot of missing information. If have to figure out on my own following things.

i) Is the parameter that is requested to be changed correlated to any other parameters ?

ii) If the parameter has correlation to other parameter, it is one-to-one type of correlation ? or if any changes are ok as long as it falls into a certain range

Let's suppose we have following request.

a) I want you to change the number of channelization code from 8 to 10 in my existing bearer.

b) I want you to change the number of slots from 1 to 2 in my existing bearer.

c) I want you to change the slot fromat from 8 to 3 in my existing bearer.

If I take the request a), do I only have to change the number of channelization code for each slot ? or do I have to change some other parameter like TFS definition or any other transport/higher layer parameter ?

You may ask similar question for the request b) or c).

General guide line for this situation is as follows.

First, figure out the maximum total databits per radio frame(10 ms) based on TFCS setting. This become the input data size of transport channel process.

Second, figure out the totoal databits for radio frame (10 ms) based on physical layer configuration. This becomes the size of the output of transport channel process.0

Theoretically, if the output data size (2) of transport process is greater than input data size (1), any change in PHY layer or transport layer parameter are acceptable.

In reality, you may need a little bit further detailed guide lines case by case, but this can be a good first step guide line and at least it will relieve some panic -:)

Let's make the comparison a little more exact. Quantity (1) counts transport block bits, before CRC attachment and channel coding. Coding adds bits, so the real input to rate matching is larger than (1). For example, rate 1/3 turbo coding roughly triples the number of bits. So a bearer can pass the rough test above and still not fit.

Rate matching then adapts the number of coded bits to the number of physical channel bits. It repeats bits when the physical channel has room to spare. It punctures bits when there are too many coded bits, but only down to the puncturing limit PL. 25.222 allows at most (1 - PL) x 100 percent of puncturing. So the Puncturing Limit row of the DPCH Uplink table, 0.56 for Physical 1, allows up to 44 percent of the coded bits to be punctured.

Now we can check the three requests at the start of this section in the same way. Requests a) and b) change the number of codes or slots, so they change quantity (2). Request c) changes the slot format, which changes the bits per slot, so quantity (2) moves again. In all three cases, recompute (2) and confirm that the coded bits from the TFCS still fit after rate matching. If they do not fit, the TFS or the TFCS must change too.

• Compare the TFCS bits with the physical channel bits : this is the first test for any change to a bearer.
• Coding adds bits before rate matching : the transport block size alone underestimates what the physical layer must carry.
• The puncturing limit sets how far the bits can be reduced : with PL = 0.56, at most 44 percent of the coded bits can be punctured.

Examples of common Radio Bearer with focus on Physical Layer Configuration

The pages below apply the rules of this page to real bearers. Each one shows the configuration of one TD-SCDMA radio bearer with the focus on the physical layer. You can compare its slot and code plan with the master plan and the spreading factors above.

• Radio Bearer - Packet DL 64K/UL 64K
• Radio Bearer - Packet DL 144K/UL 64K
• Radio Bearer - Packet DL 384K/UL 64K
• Radio Bearer - HSDPA

Reference

• 3GPP TS 25.221 v19.0.0 : clause 5A, physical channels for the 1.28 Mcps option, and clause 5.2.1, spreading
• 3GPP TS 25.222 v19.0.0 : clause 4.2.7, rate matching
• 3GPP TS 25.306 v19.0.0 : Table 5.1c, 1.28 Mcps TDD HS-DSCH physical layer categories