8 Radio Interface Synchronisation
25.4023GPPRelease 17Synchronisation in UTRAN Stage 2TS
8.1 General
This subclause describes the Radio Interface Synchronisation for FDD and TDD.
8.2 FDD Radio Interface Synchronisation
8.2.1 General
This section is fully applicable to both the DL DPCH and the F-DPCH. As such, wherever "DL DPCH" appears in this section (in text, figure and equations), it has to be replaced with "F-DPCH" in the context of "F-DPCH".
FDD Radio Interface Synchronisation assures that UE gets the correct frames when received from several cells. The UE measures the Timing difference between its DPCH and SFN in the target cell when doing handover and reports it to SRNC. SRNC sends this Time difference value in two parameters Frame Offset and Chip Offset over Iub to Node B. Node B rounds this value to the closest 256 chip boundary in order to get DL orthogonality (regardless of used spreading factor). The rounded value is used in Node B for the DL DPCH or the F-DPCH.
DOFFFDD is selected by the SRNC considering the interleaving period (e.g. 10, 20, 40 or 80 ms) when entering in dedicated state from common channel state.
Services are scheduled by using DOFFFDD in order to average out the Iub traffic load and the Node B processing load. DOFFFDD (FDD Default DPCH Offset value) is only used when setting up the first RL in order to initialise Frame Offset and Chip Offset and to tell UE when frames are expected.
UE uses the UL DPCH as it is a more defined time instant compared with DL DPCH.
The handover reference is the time instant TUETx -To, which is called DL DPCHnom in the timing diagram.
Tcell is used to skew cells in the same Node B in order to not get colliding SCH bursts, one SCH burst is 1/10 of a slot time.
The timing diagram in Figure 15 shows an example with two cells connected to one UE where handover is done from source cell (Cell 1) to target cell (Cell 2).
Figure 15: FDD Radio Interface Synchronisation timing diagram
SFN1 is found in Cell 1 at Node B1 and SFN2 at Cell 2 and Node B2. SFN1 is sent T_cell1 after the Node B1 reference BFN1. CFN is the frame numbering that is related to each DL and UL Dedicated Physical Channel (DPCH). UL DPCH is sent from UE to both Cells (both Node B’s in this example). UL DPCH2 at Node B2 is shown to indicate the difference to the DL DPCH2 at Node B2.
The new RL (DL DPCH2) which is setup at the HO will face some deviation from nominal position due to the rounding of Frame Offset and Chip Offset to 256 chip boundary in Node B. Time dispersion and UE movements are examples of other factors affecting this phase deviation.
The nominal DL DPCH timing at UE is To before the TUETX time instant, which could be expressed:
DL DPCHnom = TUETX -To (8.1)
In UE dedicated state, OFF and Tm are measured at UE according to the following equation:
OFF + Tm = (SFNtarget –DL DPCHnom) mod 256 frames [chips] (8.2)
NOTE: OFF has the unit Frames and Tm the unit Chips.
EXAMPLE 1: Assume that OFF + Tm equals "3.3300" frames (as given as an example in
Figure 15). Then OFF = 3 and Tm = "0.33" which corresponds to Tm = 12672 chips.
In other words (referring to the timing diagram in Figure 15):
– How to determine Tm at UE: Select a time instant 1) where frame N starts at DL SFN2 e.g. frame number 3, the time from that time instant to the next frame border of DL DPCHnom 2) equals Tm
(if these are in phase with each other, Tm is zero).
– How to determine OFF: The difference between the frame number selected for time instant 1) and the frame number starting at instant 2) mod 256 frames equals OFF.
EXAMPLE 2: (3 –0) mod 256 = 3, another example is (1 –254) mod 256 = 3.
8.2.2 Neighbour cell list timing information
A cell can optionally broadcast a neighbouring cell list that indicates timing information for neighbouring cells. The list contains the inter cell timing difference to neighbour cells with associated estimated uncertainty. The inter cell timing uncertainty depends on what timing difference estimating means that are used in the system (No means at all, Node synchronisation measurements, UE inter-cell measurements, Cells belonging to the same Node B or even GPS). The purpose with the neighbouring cell list timing information is to enable shorter cell search time for UE, to save UE battery and to potentially lower BCH Tx power for cells in a synchronised cluster.
8.3 TDD Radio Interface Synchronisation
8.3.1 General
The TDD Radio Interface Synchronisation relates to the following two aspects:
– Intercell Synchronisation;
– Timing Advance for 3.84Mcps and 7.68 Mcps TDD, and Uplink Synchronisation for 1.28Mcps TDD.
In TDD mode Intercell Synchronisation may be achieved by means of:
– Inter Node B Node Synchronisation that allows to achieve a common timing reference among Node B’s.
The Radio Interface Synchronisation between UE and UTRAN is achieved by means of the Timing Advance mechanism.
8.3.2 Intercell Synchronisation
Intercell Synchronisation ensures that the frame boundaries are positioned at the same time instant in adjacent cells (see Figure 16).
This requirement is necessary to minimise the interference between UEs in neighbouring cell.
In addition it automatically ensures that the slots of different cells are synchronised, i.e. they do not overlap at the UE.
Figure 16: Intercell Synchronisation
Furthermore, Intercell Synchronisation assures the synchronisation of the last 8 bits of the SFN, that is required if frame wise hopping mechanisms among cells are used. It also can be used to keep more efficient and faster all procedures involving a switch from one cell to another, such as searching for new cells, locking to new cells or handover.
8.3.3 Multi Frame Synchronisation
Void.
8.3.4 Timing Advance for 3.84Mcps and 7.68Mcps TDD
Timing Advance is used in uplink to align the uplink radio signals from the UE to the UTRAN both in case of uplink Dedicated Physical Channels (DPCH), E-PUCH, E-RUCCH and of Physical Uplink Shared Channels (PUSCH).
The handling of timing advance can be divided in four main categories: measurement, initial assignment, updates during operation, and setting on handover. For each category, a number of different cases can be distinguished.
1. Measurement of the timing deviation on the physical channels:
– On PRACH transmissions;
– On DPCH transmissions;
– On PUSCH transmissions;
– On E-PUCH transmissions;
– On E-RUCCH transmissions
2. Assignment of correct timing advance value when establishing new channels:
– At transition to CELL_DCH state;
– When establishing an USCH in CELL_FACH state.
3. Update of timing advance value for channels in operation:
– UE in CELL_DCH state;
– UE with USCH in CELL_FACH state.
4. Setting of timing advance value for target cell at handover:
– Handover from TDD to TDD with synchronised cells;
– Handover from TDD to TDD with unsynchronised cells;
– Handover from FDD to TDD;
– Handover from other systems to TDD.
8.3.4.1 Measurement of the timing deviation on the physical channels
Timing deviation measurements are always performed in the physical layer in Node B. These measurements have to be reported to the higher layers, where timing advance values are calculated and signalled to the UE. For this reporting, a number of different ways are foreseen, depending on the used physical channels.
PRACH: The Node B physical layer measures the timing deviation of the received PRACH signal (RX Timing Deviation) and passes this together with the transport block to the CRNC (by means of the Iub RACH Frame Protocol). In case the RACH carries a DDCH or DTCH, the measured timing deviation may be passed from DRNC to the SRNC over Iur interface (by means of the Iur RACH Frame Protocol).
NOTE: PRACH transmissions themselves are transmitted with a large guard period so they do not require timing advance.
PUSCH: The Node B physical layer measures the timing deviation of the received PUSCH signal (RX Timing Deviation) and passes this together with the transport block to the CRNC (by means of the Iub USCH Frame Protocol).
DPCH: The Node B physical layer measures the timing deviation of the received DPCH signal (RX Timing Deviation) and passes this value, if the conditions for reporting the measurement are met, to the SRNC (by means of the Iub & Iur DCH Frame Protocols – Rx Timing Deviation Frame Protocol).
E–PUCH: The Node B physical layer measures the timing deviation of the received E-PUCH signal (RX Timing Deviation) and passes this value, if conditions for reporting the measurement are met, to the SRNC (by means of the Iub & Iur DCH Frame Protocols – Rx Timing Deviation Frame Protocol).
E-RUCH: The Node B physical layer measures the timing deviation of the received E-RUCCH signal (RX Timing Deviation) and passes this value, if conditions for reporting the measurement are met, to the SRNC (by means of the Iub & Iur DCH Frame Protocols – Rx Timing Deviation Frame Protocol). When the E-RUCCH represents a Timing Advance Request message TS 25.321 [26] the Node B calculates the timing advance and responds with a Timing Advance Response message TS 25.321 [26].
8.3.4.2 Assignment of correct timing advance value when establishing new channels
8.3.4.2.1 Transition to CELL_DCH State
The transition to CELL_DCH state from CELL_FACH state or Idle Mode operates in the following manner:
– The SRNC checks whether an up to date timing deviation measurement is available. Such a measurement can be available from a recent RACH access (e.g. from initial access) or from a recent USCH transmission. If no up to date timing deviation measurement is available, e.g. because of lack of uplink transmissions, or during USCH over Iur, the SRNC is not informed about RX Timing Deviations, and has to trigger an uplink transmission from the UE before it can assign a DCH (for example, a RRC procedure requiring a response from the UE). The SRNC calculates the required timing advance value and saves it in the UE context in the SRNC for later use in dedicated or shared channel activation.
– The SRNC attaches the timing advance value to the channel allocation message that it signals to the UE via FACH (RRC message CONNECTION SETUP or RRC message RADIO BEARER SETUP).
– When the UE receives the channel allocation message it configures its physical layer with the given absolute timing advance value. When a timing advance command is signalled to the UE, the CFN that the new timing advance is to be applied is always signalled.
8.3.4.2.2 When establishing an USCH in CELL_FACH state
For uplink traffic using the USCH, short time allocations are sent to the UE regularly. Therefore establishing an USCH in CELL_FACH state is very similar to handling of timing advance updates during USCH operation. The UTRAN shall use a recent timing deviation measurement. Such a measurement shall be available from a recent USCH burst or a recent RACH access (e.g. from the RRC message PUSCH CAPACITY REQUEST).
8.3.4.2.3 When establishing E-DCH in CELL_DCH state (E-DCH/HS-DSCH operation with no UL DPCH)
The UTRAN shall use timing deviation. Such measurements shall be available in relation to recent transmissions on E-RUCCH and/or E-PUCH.
8.3.4.3 Update of timing advance value for channels in operation
8.3.4.3.1 UE in CELL_DCH state
An UE that is operating a dedicated channel (CELL_DCH state) has to update the timing advance from time to time to keep the received signal at the Node B within the required time window. Under reasonable assumptions the worst case update frequency is in the order of 8 seconds.
The timing advance update procedure operates in the following manner:
1. The SRNC determines whether a new timing advance value has to be transmitted to the UE taking into account the timing deviation measurements. The new timing advance value is calculated taking into account the UE’s current timing advance value.
2. The new timing advance value and the CFN in which it is to take effect are signalled to the UE via RRC signalling on FACH or DCH (PHYSICAL CHANNEL RECONFIGURATION, TRANSPORT CHANNEL RECONFIGURATION, RADIO BEARER RECONFIGURATION or UPLINK PHYSICAL CHANNEL CONTROL are examples of possible messages on the DCCH).
3. The SRNC shall also send the updated timing advance value and the CFN in which it is to take effect to the Node B, using a user plan control message. The Node B may adjust its physical layer to take the change in uplink transmission into account.
4. When the UE receives a new timing advance value, it shall configure its physical layer so that the updated timing advance value takes effect on the given CFN specified within the RRC message. The timing advance value shall be applied to all DPCHs and, if present, to all PUSCHs.
There is no need for the UE to acknowledge the timing advance update: the Node B continually measures and reports the UE timing deviation and the UE reports the received timing advance value as part of its measurement reporting. The SRNC is thus able to detect if a timing advance update has not been received and needs to be resent.
8.3.4.3.2 UE with USCH Traffic in CELL_FACH state
If the UE uses an USCH in CELL_FACH state (no DCH), the timing advance update procedure operates in the following manner:
1. The CRNC determines whether a new timing advance value has to be transmitted to the UE taking into account when the last timing advance update was signalled. Two cases are possible:
– If the data transfer is uplink after a longer idle period then the UE has to transmit a capacity request on the RACH. The CRNC is therefore informed of any timing deviation on this RACH.
– If a new allocation follows an USCH transmission, the timing deviation is already known to the CRNC from measurements of the last uplink transmission.
2. If a Timing Advance update is needed, the CRNC includes a new timing advance value and the CFN in which it will take effect in the next USCH allocation message to the UE (PHYSICAL SHARED CHANNEL ALLOCATION).
3. The CRNC shall also send a user plane control message indicating the CFN and the updated timing advance value to the Node B so the Node B can adjust its physical layer averaging to take the change in uplink transmission into account.
4. When the UE receives a new timing advance value, the UE shall configure its physical layer, so that the updated timing advance value takes effect on the given CFN specified within the PHYSICAL SHARED CHANNEL ALLOCATION message. The timing advance value shall be applied to all present PUSCHs.
8.3.4.4 Setting of timing advance value for target cell at handover
8.3.4.4.1 General
Since the uplink radio signals need to be adjusted only because of large enough distances between the UE and the cell transmission, certain cells will have a small enough radius that timing advance needs to not be used. In those cells the timing advance value in the UE is set to zero and UE autonomous adjustment of timing advance upon handover is disabled in the handover messages to the UE.
In these cells, where TA is not applied, the “RX Timing Deviation” measurement can be omitted if no other procedure (e.g. LCS) requires it.
8.3.4.4.2 Handover from TDD to TDD with synchronised cells
When two TDD cells are involved in handover and the two cells are sufficiently synchronised, a UE is able to measure the time offset between P-CCPCH reception of the two cells and, consequently, is able to autonomously correct its timing on handover without UTRAN assistance. However to improve the accuracy for the UE calculated timing advance, the SRNC can include an updated timing advance based on the timing deviation measured by the old cell in the messages triggering the handover in the UE. Note that this update shall apply in the old cell at the specified CFN if handover is performed on a later CFN or if the handover fails and falls back to the old cell. The UE shall use this new value as the basis for the UE autonomous update.
After a successful handover, a response message is transmitted in the new cell. In this message, if the UE autonomously updated its timing advance it shall report the calculated timing advance value, which it is using for access to the new cell. By this way, the SRNC is informed as fast as possible about the absolute timing advance value in the UE, and it can correct the timing advance immediately or in the future based on this value, if necessary.
8.3.4.4.3 Handover from FDD to TDD, Handover from other systems to TDD, or Handover from TDD to TDD with unsynchronised cells
In these cases, synchronisation between the handover cells is not possible. In the case where DPCH(s) are assigned, the new TDD cell must use a burst type with a large enough transmission window to allow the immediate transmission of data without the need of timing advance adjustment in the new cell, since timing adjustment can only be performed in these cells after the first uplink transmission. In the case where E-DCH/HS-DSCH operation is configured without a DPCH then the UE obtains timing advance in the new cell (using TA Request, see TS 25.321 [26]) before transmitting data.
8.3.5 UL Synchronisation for 1.28Mcps TDD
This section describes the details of the UL synchronisation including the establishment of UL synchronisation and maintenance of the UL synchronisation.
8.3.5.1 The establishment of uplink synchronisation
8.3.5.1.1 Preparation of uplink synchronisation by downlink synchronisation
When a UE is powered on, it first needs to establish the downlink synchronisation with the cell. Only after the UE can establish and maintain the downlink synchronisation, it can start the uplink synchronisation procedure.
8.3.5.1.2 Establishment of uplink synchronisation
Although the UE can receive the downlink synchronisation signal from the Node B, the distance to Node B is still uncertain which would lead to unsynchronised uplink transmission. Therefore, the first transmission in uplink direction is performed in Uplink Pilot Channel (UpPCH), to avoid interference in traffic time-slots.
The timing used for the SYNC_UL code are set e.g. according to the received power level of DwPCH and/or
P-CCPCH.
At the detection of the SYNC_UL sequence in the searching window, the Node B will evaluate the received power levels and timing, and reply by sending the adjustment information to UE to modify its timing and power level for next transmission and for establishment of the uplink synchronisation procedure. Within the next 4 sub-frames, the Node B will send the adjustment information to the UE (in a single subframe message in the FPACH). The uplink synchronisation procedure, normally used for a random access to the system, can also be used for the re-establishment of the uplink synchronisation when uplink is out of synchronisation.
8.3.5.2. Maintenance of uplink synchronisation
For the maintenance of the uplink synchronisation, the midamble field of each uplink burst can be used.
In each uplink time slot the midamble in each UE is different. The Node B can estimate the power level and timing shift by measuring the midamble field of each UE in the same time slot. Then, in the next available downlink time slot, the Node B will signal the Synchronisation Shift (SS) and the Power Control (PC) commands to enable the UE to properly adjust respectively its Tx timing and Tx power level.
These procedures guarantee the reliability of the uplink synchronisation. The uplink synchronisation can be checked once per 1.28Mcps TDD subframe. The step size in uplink synchronisation is configurable and re-configurable and can be adapted from 1/8 chip to 1 chip duration. The following updates for UL synchronisation are possible: 1 step up;
1 step down; no update.
For 3.84Mcps and 7.68Mcps TDD option, uplink synchronisation is mentioned in 4.3 of TS 25.224 [16]. But the implementation method is a little different with the 1.28Mcps TDD option. For 1.28Mcps TDD option, the establishment of the UL synchronisation is done by using the UpPCH and the FPACH.
UE will select one of the set of SYNC_UL codes which can be used in the cell to establish uplink synchronisation in the access procedure. The benefit of this method is when the UE wants to do random access, the PRACH will have minimum interference to other traffic channel. Vice versa, it will also reduce the interference from traffic channels to PRACH.