5 Internal test loops

04.143GPPIndividual equipment type requirements and interworkingSpecial conformance testing functionsTS

A number of internal test loops are required providing access to isolated functions of the MS without introducing new physical interfaces just for the reason of type approval testing. Fig 5-1 shows a functional block diagram of a reference MS containing the different test loops.

NOTE: It should be emphasized that these test loops only describe the functional behaviour of the MS with respect to its external interfaces; physical implementation of the loops is completely left open to the manufacturer.

A particular loop is activated in an MS by transmitting the appropriate command message to the MS.

NOTE: In the case of loops A and B, when a TCH/EFS is used, the MS loops back 244 bits instead of 260 bits, see subclauses 5.1.2.1 and 5.1.3.1.

Figure 1: Test loops in the MS

5.1 Single-slot TCH loops

5.1.1 Purpose of Single-slot TCH loops

To establish a transparent loop for TCH blocks a TCH must be active between the SS and MS. The TCH may be full or half rate, speech or data of any rate specified in the GSM system.

Six types of Single-slot TCH loop back are defined.

The first (A) includes the signalling of erased frames and is used to determine Frame Erasure Ratio (FER) and Residual Bit Error Ratio (RBER) for speech TCH and Bit Error Ratio (BER) for any data TCH.

The second type (B) is required to determine Class II bit error ratio for the speech TCH.

With the third loop (C) the 114 information bits of each TCH burst (excluding stealing flags) prior to applying benefit of the channel decoder, but after decryption, shall be transmitted in an uplink burst. (Equivalent error rate to TCH/FS Class II). All that is received shall be re-transmitted regardless of the state of the received midamble. The midamble in the uplink bursts shall be the normal midamble used by the MS. SACCH and idle bursts are not looped back.

The fourth loop (D) includes the signalling of erased frames and unreliable frames and is used to determine Unreliable Frame Ratio (UFR) and Residual Bit Error Ratio (RBER) for TCH/HS.

The fifth loop (E) includes the signalling of erased SID frames and is used to determine Erased SID Frame Rate (ESIDR) and Residual Bit Error Ratio (RBER) for TCH/HS.

The sixth loop (F) includes the signalling of erased valid SID frames and is used to determine Erased Valid SID Frame Rate (EVSIDR) and Residual Bit Error Ratio (RBER) for TCH/HS.

NOTE: Measurement of TCH/FS chip BER is approximately five times faster using loop C rather than loop B.

5.1.2 TCH loop including signalling of erased frames (A)

5.1.2.1 Procedure

The SS orders the MS to close its TCH loop by transmitting a CLOSE_TCH_LOOP_CMD message, specifying the TCH to be looped and that erased frames are to be signalled by the MS. The SS then starts timer TT01.

If no TCH is active, or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_TCH_LOOP_ACK message. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, every good speech frame or any user data frame received by the MS on the specified TCH (downlink) shall be taken from the output of the channel decoder, input to the channel encoder and transmitted on the same TCH (uplink).

In the case where TCH is TCH/FS or TCH/HS, the MS shall loop back the 260 bits after normal channel decoding.

In the case where TCH is TCH/EFS, the MS shall loop back the 244 bits after normal and preliminary channel decoding.

If the channel decoder detects a bad speech frame , then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH (uplink).

If the MS decodes stealing flags as indicating an FACCH frame, then there is no defined response for the MS to the channel encoder for transmission on the TCH (uplink). The FACCH channel shall operate as normal.

5.1.3 Speech TCH loop without signalling of erased frames (B)

5.1.3.1 Procedure

The SS orders the MS to close its TCH loop by transmitting a CLOSE_TCH_LOOP_CMD message, specifying the TCH to be looped. The SS then starts timer TT01.

If no TCH is active or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_TCH_LOOP_ACK. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, any speech frame received by the MS on the specified TCH (downlink) shall be taken from the output of the channel decoder, input to the channel encoder, and transmitted on the same TCH (uplink).

In the case where TCH is TCH/FS or TCH/HS, the MS shall loop back the 260 bits after normal channel decoding.

In the case where TCH is TCH/EFS, the MS shall loop back the 244 bits after normal and preliminary channel decoding.

The SS should avoid using the FACCH downlink in this situation until the test is complete.

5.1.4 TCH burst-by-burst loop (C)

5.1.4.1 Applicability

The test loop shall be implemented by all ME, supporting any TCH.

5.1.4.2 Procedure

Establishment and clearing of the loop is performed at ideal radio conditions.

5.1.4.3 Establishment

– The establishment shall be commanded by transmitting a CLOSE_TCH_LOOP_CMD message. The SS then starts timer TT01. This command shall be acknowledged by the MS with a CLOSE_TCH_LOOP_ACK message. Upon receipt of that message the SS stops timer TT01. The MS shall establish the loop within one reporting period [SACCH-block = 104 frames] from the sending of the CLOSE_TCH_LOOP_ACK.

– If no TCH is active or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

5.1.4.4 Operation

– The round trip delay (RTD), which is the number of TCH frames between the reception of one burst at the MS, and the transmission of the same burst (on the uplink) shall be less than 26 TDMA frames. The actual value shall be declared for the implementation to be tested.

NOTE 1: The RTD can be as long as required to receive the number of interleaved burst for the relevant TCH.

NOTE 2: Example of RTD = 5

TDMA Frame No.

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Downlink

T1

T2

T3

T4

T5

T6

T7

T8

T9

T10

T11

T12

Sd

Uplink

T1

T2

T3

T4

T5

T6

T7

Su

T8

T9

T10

T11

T12

Sd = Downlink SACCH frame, Su = Uplink SACCH frame, Tn = TCH frame

Note from the above that TCH frames looped back prior to the uplink SACCH (or Idle) frame are delayed on the air interface by 5 TDMA frames, but the TCN frames following the SACCH frame are delayed by 6 TDMA frames. The RTD is therefore not to be confused with the TDMA frame delay for a TCH burst, which varies depending on whether the TCH burst is before or after the uplink SACCH frame. The reason for the variable TDMA time delay is to preserve the uplink SACCH frame position in the multi-frame. Note also that the uplink SACCH data is not a looped back version of the downlink SACCH data.

5.1.5 TCH loop including signalling of erased frames and unreliable frames (D)

5.1.5.1 Procedure

The SS orders the MS to close its TCH loop by transmitting a CLOSE_TCH_LOOP_CMD message, specifying the TCH to be looped and that erased frames and unreliable frames are to be signalled by the MS. The SS then starts timer TT01.

If no TCH is active, or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_TCH_LOOP_ACK. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, every reliable speech frame (UFI = 0) received by the MS on the specified TCH/HS (downlink) shall be taken from the output of the channel decoder, input to the channel encoder and transmitted on the same TCH (uplink).

If the channel decoder detects a bad speech frame or an unreliable frame (BFI = 1 or UFI = 1) or if the MS decodes the stealing flags as indicating an FACCH frame, then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH/HS (uplink).The FACCH channel shall operate normally.

5.1.6 TCH loop including signalling of erased SID frames (E)

5.1.6.1 Procedure

The SS orders the MS to close its TCH loop by transmitting a CLOSE_TCH_LOOP_CMD message, specifying the TCH to be looped and that erased SID frames are to be signalled by the MS. The SS then starts timer TT01.

If no TCH is active, or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_TCH_LOOP_ACK. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, every valid SID frame (SID = 2) or invalid SID frame (SID = 1) received by the MS on the specified TCH/HS (downlink), shall be taken from the output of the channel decoder, input to the channel encoder and transmitted on the same TCH/HS (uplink).

If the channel decoder detects an erased SID frame (SID = 0), then this shall be signalled to the SS, by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH/HS (uplink).

If the MS decodes the stealing flags as indicating an FACCH frame, then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH/HS (uplink).The FACCH channel shall operate normally.

5.1.7 TCH loop including signalling of erased valid SID frames (F)

5.1.7.1 Procedure

The SS orders the MS to close its TCH loop by transmitting a CLOSE_TCH_LOOP_CMD message, specifying the TCH to be looped and that erased valid SID frames are to be signalled by the MS. The SS then starts timer TT01.

If no TCH is active, or any test loop is already closed, the MS shall ignore any CLOSE_TCH_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_TCH_LOOP_ACK. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, every valid SID frame (SID = 2 and BFI = 0) received by the MS on the specified TCH/HS (downlink), shall be taken from the output of the channel decoder, input to the channel encoder and transmitted on the same TCH/HS (uplink).

If the channel decoder detects an erased valid SID frame (SID = 1) or (SID = 0) or ((BFI or UFI) = 1)), then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH/HS (uplink).

If the MS decodes the stealing flags as indicating an FACCH frame, then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH/HS (uplink).The FACCH channel shall operate normally.

5.1.8 Additional non-mandatory operating characteristics for single-slot loops

In order to optimise the speed and flexibility of mobile manufacturing and repair, the following non-mandatory characteristics of the test loops are suggested:

– The normal FACCH downlink and uplink functions should ideally be maintained when the test loop is closed. In particular, channel assignments or handovers, and call termination from either the mobile or the base station simulator.

– Following an assignment or handover, the loop should not open if it was closed prior to the handover.

– Following call dropping or deliberate call termination, the loop should be re-opened.

– The loopback functions should ideally operate with or without (i.e. no SIM) the test SIM present, but should not operate with a network SIM present.

– Audio muting should be enabled when the loop is closed.

5.2 Multi-slot TCH loops

5.2.1 Purpose of Multi-slot TCH loops

To establish a transparent loop for TCH blocks, from multiple slots, a TCH must be active between the SS and MS.

Two types of Multi-slot TCH loop back are defined.

With the first loop (G) the 114 information bits of each multi-slot TCH burst (excluding stealing flags) prior to applying benefit of the channel decoder, but after decryption (see Figure 1), shall be transmitted in an uplink burst. (Equivalent error rate to TCH/FS Class II). All that is received shall be re-transmitted regardless of the state of the received midamble. The midamble in the uplink bursts shall be the normal midamble used by the MS. SACCH and idle bursts are not looped back.

The second loop (H) includes the signalling of erased frames and is used to determine Frame Erasure Ratio (FER), Residual Bit Error Ratio (RBER) and Bit Error Ratio (BER) for any multi-slot configuration TCH.

Each of the two loops shall support the following mechanisms:

The first (Multi-slot mechanism 1) is used to loop the TCH data of slot X of the downlink onto the TCH of the main uplink slot (for HSCSD). This mechanism is needed to cover the case where there are more downlink slots than uplink slots.

The second (Multi-slot mechanism 2) is used to loop as many downlink slots as possible to the corresponding uplink slots, based on the following rules for HSCSD:

Loop back all bi-directional timeslots, and leave the unidirectional slots not looped back. This maintains the logical association with bi-directional timeslots.

It should be further noted:

The order of the data on the downlink shall be preserved on the uplink.

The OPEN_Multi-slot_LOOP_CMD message shall open all Multi-slot loops.

Assignment to a new multi-slot configuration shall be preceded by an OPEN_Multi-slot_LOOP_CMD message to open all loops.

It is the responsibility of the System Simulator (SS) to ensure that the correct configuration is enabled for the test. Test loops will be opened by the receipt of a OPEN_Multi-slot_LOOP_CMD or by disconnecting the call. Other behaviour, such as receiving a new TxLev or a channel assignment or handover to a new ARFCN will not affect the test loops. The SS should ensure that a new multislot configuration affecting an existing test loop is not included within channel assignment, handover or configuration change commands.

If the Multi-slot mechanism 1 is used and a downlink slot that is not part of the current multi-slot configuration is specified, the MS shall ignore the command and send a negative acknowledgment. The loopback state should not change.

Once a loop is closed, a further loopback command shall over-ride a previous command – multiple CLOSE_Multi-slot_LOOP_CMD messages are not additive.

Call disconnect for whatever reason shall open all loops. No OPEN_Multi-slot_LOOP_ACK message shall be sent.

The multi-slot loopback is restricted to the TCH logical channel only. The downlink and uplink FACCH and SACCH should work as if loopback did not exist.

The Multi-slot TCH loops are in addition to any Single-slot TCH loops already specified for the type of MS.

Support of the Multi-slot loops is mandatory for any MS supporting HSCSD.

Any MS supporting the Multi-Slot loops shall activate the functions defined in this section of the specification regardless of the presence or not of a test SIM.

5.2.2 Multi-slot TCH burst-by-burst loop (G)

5.2.2.1 Procedure

The establishment shall be commanded by transmitting a CLOSE_Multi-slot_LOOP_CMD message. The SS then starts timer TT01. This command shall be acknowledged by the MS with a CLOSE_Multi-slot_LOOP_ACK message. Upon receipt of that message the SS stops timer TT01. The MS shall establish the loop within one reporting period [SACCH-block = 104 frames] from the sending of the CLOSE_Multi-slot_LOOP_ACK.

If no TCH is active or any test loop is already closed, the MS shall ignore any CLOSE_Multi-slot_LOOP_CMD message

RTD is as the same as subclause 5.1.4.4.

5.2.3 Multi-slot TCH loop including signalling of erased frames (H)

5.2.3.1 Procedure

The SS orders the MS to close its Multi-slot TCH loop by transmitting a CLOSE_Multi-slot_LOOP_CMD message, specifying the TCH to be looped and that erased frames are to be signalled by the MS. The SS then starts timer TT01.

If no TCH is active, or any test loop is already closed, the MS shall ignore any CLOSE_Multi-slot_LOOP_CMD message.

If a TCH is active, the MS shall close its TCH loop for the TCH specified and send back to the SS a CLOSE_Multi-slot_LOOP_ACK message. Upon reception of that message the SS stops timer TT01.

After the MS has closed its TCH loop, every good speech frame or any user data frame received by the MS on the specified TCH (downlink) shall be taken from the output of the channel decoder, input to the channel encoder and transmitted on the same TCH (uplink).

If the channel decoder detects a bad speech frame , then this shall be signalled to the SS by setting the input frame to the channel encoder to zero’s, and transmitting on the TCH (uplink).

If the MS decodes stealing flags as indicating an FACCH frame, then there is no defined response for the MS to the channel encoder for transmission on the TCH (uplink). The FACCH channel shall operate as normal.

5.3 Deactivating loops

5.3.1 Deactivating Single-slot TCH loops

The SS orders the MS to open any Single-slot TCH loop by transmitting an OPEN_LOOP_CMD message.

If no loop is closed the MS shall ignore any OPEN_LOOP_CMD message.

If a Single-slot TCH is looped, the MS shall open the loop.

If the loop opened was type C, the MS shall send an OPEN_LOOP_CMD message to the SS with bit 0 of the optional acknowledgement element set to 1.

All channels shall be open for normal use again.

5.3.2 Deactivating Multi-slot TCH loops

The SS orders the MS to open any Multi-slot TCH loop by transmitting an OPEN_Multi-slot_LOOP_CMD message.

If no loop is closed the MS shall ignore any OPEN_Multi-slot_LOOP_CMD message.

If a Multi-slot TCH is looped, the MS shall open the loop and send a OPEN_Multi-slot_LOOP_ACK message to the SS.

All channels shall be open for normal use again.

5.4 Multi-slot test mode for GPRS & EGPRS

The following test mode applies to both GPRS and EGPRS

5.4.1 Initiation

5.4.1.1 MS Declaration

The manufacturer shall declare the MS test mode capability before test mode initiation. The possible declarations are as follows:

a) MS is capable of transmitting a pseudo-random data sequence in RLC data blocks

b) MS is capable of transmitting looped-back RLC data blocks

c) MS is capable of both a) and b).

The specifics of test mode initiation and operation depend on what capability the MS has declared. In addition to options a),b) or c), an EGPRS MS shall also support the EGPRS Radio Block Loopback Mode defined in section 5.5.

5.4.1.2 Establishment of uplink TBF

The MS is assumed to be GPRS attached, in packet idle mode.

The SS establishes a downlink TBF on one timeslot, according to normal procedures as defined in 04.60.

The SS orders the MS into GPRS test mode by transmitting a GPRS_TEST_MODE_CMD (Layer 3 message, SAPI 1) with parameter PDU Description set to define the number of PDUs and number of octets within the PDUs that the MS is to transmit in the uplink during the test. The SS then starts timer TT02.

If the MS has declared c) capability in section 5.4.1.1, then the M bit in the GPRS_TEST_MODE_CMD message shall determine whether the MS operates the test mode in mode a) or mode b).

This commands the MS to request the establishment of an uplink TBF, in RLC unacknowledged mode, according to normal procedures as defined in 04.60. Upon receipt of the MS request for uplink resources, the SS stops timer TT02 and proceeds with the establishment of the uplink TBF, by assigning a Dynamic Allocation uplink TBF.If the MS is operating in mode a), the SS releases the downlink TBF according to normal procedures as defined in 04.60. If the MS is operating in mode b), the downlink TBF continues throughout test mode operation.

The SS shall not send a new GPRS_TEST_MODE_CMD to the MS unless the currently activated multi-slot test mode for GPRS is terminated.

When the MS has activated the multi-slot test mode for GPRS, the received RLC data blocks (on the MS side) shall not be passed to LLC.

5.4.2 Operation

5.4.2.1 MS Operating in mode a)

The SS sets the USF field in blocks transmitted on the downlink to address the MS. The MS shall transmit RLC data blocks obeying USF, according to the normal rules for transmission, as defined in 04.60.

For the uplink the data payload of the RLC data blocks shall contain a pseudorandom data sequence, as specified in clause 5.4.4. The blocks shall have valid MAC headers and may have valid RLC headers. The blocks shall be processed by Layer1 in the normal manner.

Where multiple transmit timeslots are active, the same data as is carried in the RLC data block in the first timeslot may be used in RLC data blocks carried in subsequent time-slots. The blocks shall have valid MAC headers and may have valid RLC headers. The blocks shall be processed by Layer 1 in the normal manner.

5.4.2.2 MS Operating in mode b)

The SS shall transmit RLC data blocks on the downlink TBF containing a pseudorandom data sequence in the data payload of the block, as specified in clause 5.4.4. The blocks shall have valid MAC and RLC headers. The SS shall apply the same channel coding scheme on the downlink data blocks as the commanded coding scheme on the uplink. The blocks shall be processed by Layer 1 in the normal manner.

The SS sets the USF field in blocks transmitted on the downlink to address the MS. The MS shall transmit RLC data blocks obeying USF, according to the normal rules for transmission, as defined in 04.60.

For the uplink, the data payload of the RLC data blocks shall contain the data payload of the RLC data blocks transmitted on the downlink TBF. The blocks shall have valid MAC headers and may have valid RLC headers. The countdown procedure shall not be used for the uplink RLC data blocks (CV=15).

When RLC/MAC control blocks are received on the downlink, the MS may repeat on the uplink the pseudorandom sequence carried in the previous uplink RLC data blocks.

In the event where the USF field is correctly decoded but there is a CRC error on the payload data the MS shall, if required by the USF, transmit the received payload data.

When RLC/MAC control blocks are sent on the uplink, the MS may discard the pseudorandom sequence that would otherwise have been transmitted at that time.Where multiple transmit timeslots are active, and only one downlink timeslot is active, the same data as is carried in the RLCdata block in the first timeslot may be used in RLC data blocks carried in subsequent time-slots. The blocks shall have valid MAC headers and may have valid RLC headers. The blocks shall be processed by Layer 1 in the normal manner.

5.4.2.3 Operational constraints applying to both modes

During test mode operation, the MS shall continue to receive RLC/MAC control blocks sent on the downlink, and shall respond normally.

5.4.3 Termination

Termination of the test mode occurs for test mode operating in mode a), either when the requested number of PDUs have been transmitted on the uplink TBF, or when the SS initiates the TBF release. For test mode operating in mode b), termination of test mode occurs when the SS terminates the downlink TBF, or when the SS initiates PDCH release for the uplink TBF.

5.4.4 PN Sequence Definition

The data to be inserted into the data part of the RLC/MAC data blocks is generated using any binary pseudorandom sequence generator with a cycle of 32,767 bits or greater (for example CCITT defined PN15, PN22 etc.).

Example test patterns may be found in CCITT recommendation O.153 Fascicle IV.4, (Basic parameters for the measurement of error performance at bit rates below the primary rate, Melbourne 1988) clause 2.1.

5.4.5 Optional Multi-slot operation

To facilitate production tests and for other purposes, the MS may optionally implement the following extension to this test mode.

If the downlink TBF is established on more than one timeslot, the MS shall transmit in the second uplink timeslot (if present) RLC/MAC blocks received on the second downlink timeslot, and shall transmit in the third uplink timeslot (if present) RLC/MAC blocks received in the third downlink timeslot and so on.

If more transmit timeslots are present than receive timeslots, then the contents of uplink timeslots that do not map to downlink timeslots shall be the same as in the last timeslot that maps to downlink.

However, if the downlink TBF contains only a single timeslot the MS must fill all uplink timeslots as defined in subclause 5.4.2 above.

In this description, downlink timeslots are counted from the "Downlink Timeslot Offset" in the mode flag of the GPRS_TEST_MODE_CMD. For example, if the "Downlink Timeslot Offset" is set to 3, TN3 shall be treated as the first downlink timeslot if a TBF is established in TN3. If TN3 does not support a TBF, the first active timeslot after TN3 shall be treated as the first downlink timeslot. The counting sequence is continuous through TN7 and TN0.

Uplink timeslots are always counted from TN0.

5.5 EGPRS Switched Radio Block Loopback Mode

The EGPRS Switched Radio Block Loopback mode must be supported by an EGPRS MS. It is a Physical RF layer loopback performed before channel decoding designed to support BER testing.

The following loopback path is used:

5.5.1 Initiation

The MS is assumed to be GPRS attached, in packet idle mode.

The SS establishes a downlink TBF on one timeslot, according to normal procedures as defined in 04.60.

The SS orders the MS into EGPRS Switched Radio Block Loopback Mode by transmitting a EGPRS_START_RADIO_BLOCK_LOOPBACK_CMD (Layer 3 message, LLC SAPI 1). The SS then starts timer TT02.

This will force the MS to request the establishment of an uplink TBF, in RLC unacknowledged mode, according to normal procedures as defined in 04.60. Upon receipt of the MS request for uplink resources, the SS stops timer TT02 and proceeds with the establishment of the uplink TBF, by assigning a Dynamic Allocation uplink TBF.

The downlink and uplink TBF continue throughout test mode operation, as described in 5.5.2.

After the MS has been assigned a Dynamic Allocation uplink TBF, the SS shall start timer TT03. After this timer expires the SS shall start to transmit radio blocks to the MS using the same downlink resources as the existing downlink TBF. The radio blocks shall contain a valid RLC/MAC header (addressing the MS) and an RLC data block (or blocks) to be filled with pseudorandom data as specified in 5.4.4, or contain an RLC/MAC control message.

When the MS has activated the EGPRS Switched Radio Block Loopback mode, any received RLC data blocks (on the MS side) shall not be passed to LLC.

5.5.2 Operation

The downlink and uplink TBFs shall remain open throughout switched radio block loopback mode operation. Timers T3180, T3182 and T3164 shall be disabled throughout switched radio block loopback mode operation to prevent the uplink TBF from expiring.

EGPRS Switched Radio Block Loopback mode defines two Sub-modes of operation – Radio Block Loopback ON and Radio Block Loopback OFF. When the EGPRS Switched Radio Block Loopback mode is initiated, the MS shall enter the Radio Block Loopback ON sub-mode.

Switching between the two Sub-modes is controlled by the use of the Payload Type field in the MAC header of RLC/MAC control messages. During EGPRS Switched Radio Block Loopback mode, the Payload Type field conveys two pieces of information – whether or not the optional octets of the RLC/MAC control header have been included (as for normal operation) and an instruction to change between Radio Block Loopback Sub-modes. If in Radio Block Loopback Sub-mode ON, an instruction to turn ON the Radio Block Loopback Sub-mode shall be ignored, similarly an instruction to turn OFF the Radio Block Loopback Sub-mode when in Radio Block Loopback Sub-mode OFF. The MS shall stay in EGPRS Switched Radio Block Loopback Mode whether in Radio Block Loopback Sub-mode ON or OFF until it is terminated by the procedure detailed in 5.5.3.

bit
8 7

Payload Type meaning during normal operation

Payload Type meaning in EGPRS Switched Radio Block Loopback Mode

Loopback Sub-Mode Control

0 0

RLC/MAC block contains an RLC data block

RLC/MAC block contains an RLC/MAC control block that does not include the optional octets of the RLC/MAC control header

Loopback Sub Mode ON

0 1

RLC/MAC block contains an RLC/MAC control block that does not include the optional octets of the RLC/MAC control header

Loopback Sub Mode OFF

10

In the downlink direction, the RLC/MAC block contains an RLC/MAC control block that includes the optional first octet of the RLC/MAC control header.
In the uplink direction, this value is reserved.

In the downlink direction, the RLC/MAC block contains an RLC/MAC control block that includes the optional first octet of the RLC/MAC control header.
In the uplink direction, this value is reserved

Loopback Sub-Mode ON

1 1

Reserved. In this version of the protocol, the mobile station shall ignore all fields of the RLC/MAC block except for the USF field

Loopback Sub-Mode OFF

If in Radio Block Loopback Sub-mode ON, the MS shall receive all radio blocks and, before they pass through the decoding process (see figure in section 5.5), shall send them to the SS using the same uplink resources as the existing uplink TBF. The radio blocks shall be resent on the very next block period regardless of USF or TFI decoding. The radio blocks shall also pass through the decoding process and if the RLC/MAC header is successfully decoded shall be sent to RLC/MAC. When in Radio Block Loopback Sub-mode ON, the MS should not send any control messages or data from RLC/MAC to its own Physical Link layer for transmission.

If in Radio Block Loopback Sub-mode OFF, the MS shall not loopback any radio blocks received. It can send control messages or data to be transmitted, following the normal procedures in 04.60 for a dynamic allocation.

Whether in Radio Block Loopback Sub-mode ON or OFF, the MS must obey any RLC/MAC control messages referring to the uplink or downlink TBF.

5.5.3 Termination

The SS orders the MS to terminate EGPRS Switched Radio Block Loopback Mode by transmitting a Packet TBF release, releasing the uplink and downlink TBFs. If timer T3190 expires EGPRS Switched Radio Block Loopback Mode shall also be terminated. If EGPRS test mode is terminated, the MS shall stop transmitting on the uplink, discard any data associated with the uplink TBF and return to packet idle mode.

5.5.4 Support of EGPRS MS without 8PSK modulation capability in uplink

EGPRS Switched Radio Block Loopback Mode makes provision for EGPRS MS without 8PSK modulation capability in uplink in the following manner. If the uplink TBF was established using a GMSK modulation scheme, and the downlink is 8PSK modulated, a radio block sent by the SS on a downlink timeslot using 8PSK modulation should be followed by two radio blocks on the same timeslot where the SS transmits GSM dummy bursts. The MS shall retransmit the received 8PSK data over the following three radio blocks on the matching uplink timeslot.

The EGPRS MS without 8PSK modulation capability in uplink shall not retransmit the last 16 encrypted bits received in an 8PSK radio block when retransmitting it using GMSK modulation.

The stealing bits of the GMSK blocks shall be ignored by the SS.

The SS shall select this mode by setting bit 1 of the Mode Flag in the EGPRS_START_RADIO_BLOCK_LOOPBACK_CMD message. See sub-clause 8.14.

5.5.5 Optional Multi-slot operation

To facilitate production tests and for other purposes, the MS may optionally implement the following extension to this test mode.

If the downlink TBF is established on more than one timeslot, the MS shall transmit in the second uplink timeslot (if present) radio blocks received on the second downlink timeslot, and shall transmit in the third uplink timeslot (if present) radio blocks received in the third downlink timeslot and so on.

If more transmit timeslots are present than receive timeslots, then the contents of uplink timeslots that do not map to downlink timeslots shall be the same as in the last timeslot that maps to downlink.

However, if the downlink TBF contains only a single timeslot the MS must fill all uplink timeslots as defined in subclause 5.5.2 above.

In this description, downlink timeslots are counted from the "Downlink Timeslot Offset" in the mode flag of the EGPRS_START_RADIO_BLOCK_LOOPBACK_MODE_CMD. For example, if the "Downlink Timeslot Offset" is set to 3, TN3 shall be treated as the first downlink timeslot if a TBF is established in TN3. If TN3 does not support a TBF, the first active timeslot after TN3 shall be treated as the first downlink timeslot. The counting sequence is continuous through TN7 and TN0.

Uplink timeslots are always counted from TN0.