WO1998047297A2

WO1998047297A2 - Advance indication of a transcoding change

Info

Publication number: WO1998047297A2
Application number: PCT/GB1998/001104
Authority: WO
Inventors: Richard Driscoll; Nigel Jonathan King; Paul Maxwell Martin; Neil Philip Piercy; Nigel Roles
Original assignee: Ionica International Limited
Priority date: 1997-04-16
Filing date: 1998-04-15
Publication date: 1998-10-22
Also published as: GB9707726D0; WO1998047297A3; RU99123710A; BR9809102A; AU7061698A; EP0983695A2; JP2001521693A

Abstract

A method is shown for transmitting a digital data message in time slots within fixed length time frames between a sending unit and a receiving unit. The message is transcoded for transmission according to a transcoding function and when the transcoding function is due to be changed, one of the units sends to the other unit a signal indicating at which frame the transcoding will change.

Description

A METHOD AND APPARATUS FOR DIGITAL DATA COMMUNICATIONS INCLUDING ADVANCE INDICATION OF A TRANSCODING CHANGE

Technical Field

The invention relates to digital data transmission in time slots within a fixed length time

frame, in particular, time division multiplexing/time division multiple access (TDM/TDMA)

digital telephony.

Background Art

As described* in International patent application WO96/08934, a known TDM/TDMA communications system initially establishes all call connections using a single time slot per frame (a single "bearer") at 32 kilobits per second (kbps) encoded according to an Adaptive Differential Pulse Code Modulation (ADPCM) modulation scheme. This provides good quality transmission for speech data, but the rate of information flow is insufficient for voice-band modems. To support use by modems, the communications system adds a

second time slot per frame (a second "bearer") to the call connection such that the two

bearers together provide a 64 kbps rate of data transmission using Pulse Code Modulation

(PCM) as the data encoding scheme. This adding a bearer is one example of a service

change.

A service change is any change whereby the transcoding of input signal data into digital data for transmission is changed.

An other example of a service change is conversion from ADPCM to PCM, or vice versa. This might be done to provide a better quality of transmission.

In the case of adding a bearer, the start of use of the second slot in the call connection occurs whilst the modem at the receiver transmits answertone. Answertone is the high

pitched tone audible when a remote modem answers a call. The second bearer is thus in use before the modem begins its subsequent hand-shaking.

In more detail, when a higher rate of transmission than 32 kbps is required, the known communications system chooses to assign multiple independent slots to the call connection and inter leaves the digital data between the slots. For example, when two slots per frame are used, with a 64 kbps data rate, one of the slots transmits all of the even samples and the other transmits all of the odd samples. This assignment of samples to slots is undertaken assuming minimum delay.

TDM/TDMA communications systems transmit data in bursts at a rate higher than the data

is acquired. Specifically, data is transmitted in a designated time slot or time slots within

a longer time frame. Between the bursts, the data is queued until the next burst, which

introduces a delay in the data transmission. Minimum delay is when the data samples

placed in a particular burst are the latest prior to transmission of that burst.

The assignment of data samples to time slots in the known system described in WO96/08934 is shown in Figure 1. This shows a call connection using two slots per frame, in which even samples, denoted sample 16 to sample 54 are transmitted in slot 3 of a ten slot frame. The odd samples, denoted 29 to 67 are correspondingly transmitted in slot 6 of the same frame. Later odd and even samples are similarly transmitted in slot

3 and slot 6 respectively of the subsequent frame. The slots are transmitted at different predetermined times, and the time relationship of the samples to the slots is known; for

example, as shown in Figure 1, slot 6 transmits samples which are 13 samples later in time than slot 3. This is because slot 6 is transmitted 12 samples later in time, and transmits the odd samples.

Given that both the transmitter and receiver are aware of the time relationship between the slots, the data can be decompressed and reassembled at the receiver.

The digital data for transmission is created using a transcoding function, which takes audio data and converts it into a digital data stream. It is this digital data stream which is compressed into one or more predetermined slots for transmission. This process occurs both in the uplink direction, ie. communications from a subscriber unit to a base station,

and in the downlink direction, ie. from a base station to a subscriber unit. If the service

is changed, then the transcoding function is also changed, as indicated schematically in

Figure 2.

Figure 2 shows an audio input signal switchably connected via a first transcoding function

FI or a second transcoding function F2 to a TDM transmitter. The signal is received by a TDMA receiver to determine the received digital data and this is switchably reverse- transcoded via reverse transcoding function Fl' or reverse transcoding function F2' to provide an audio output signal.

The digital data is transmitted across a radio link, and there are delays both due to the TDM TDMA protocol itself and delays in transcoding. As a result, the switching between transcoding functions must be timed accurately so that the data passes through the correct

transcoding function. If not, the data will be processed by the wrong transcoding function with corruption of the audio output signal resulting.

The known TDM/TDMA communications system as described in WO96/08934 defines a period in time when the data changes and executes a handshake between the transmitter and receiver. The sequence of signals involved in the handshake is shown in Figure 3.

Specifically, the base station which initiates the service change mutes its output of digital data received from a subscriber unit corresponding to information such as audio signals, and transmits a control signal to the subscriber unit requesting a change in transcoding. The subscriber unit (NTE) mutes its output and changes its transcoding, and transmits an

acknowledgement control signal back to the base station (BTE). The base station then switches to the new transcoding function for both uplink and downlink and unmutes (that

is resumes forwarding information data into the network) using the new transcoding function. The base station then sends a message indicating completion to the subscriber

unit which unmutes so as to resume output to terminal equipment of the subscriber. This

operation is described in WO96/08934, to which the reader is now referred.

This known approach allows both the transmitter and receiver reasonable time to transfer to the new transcoding function. However, a key disadvantage is the period of mute which is entailed. The theoretical minimum time for which the mute can last is 2 time frames, but in practice it is typically 5 to 15 time frames allowing for processing delays.

Because of this period of mute, typically 5 to 15 time frames in length, which corresponds to approximately 25 to 75 milliseconds, there is a corresponding break in the signals received by the subscriber unit. This interruption is a problem, when the subscriber unit

is connected to one of a substantial group of types of modems. These modems cannot be

relied on to successfully establish a call connection at the maximum transmission rate which they are capable of because of the break in signals received.

Summary of the Invention

The present invention relates to a method of transmitting a digital data message in time

slots within fixed length time frames between a sending unit and a receiving unit, a time slot or slots in successive frames being assigned for transmission of the message, the message being transcoded for transmission according to a transcoding function, in which, upon the transcoding function being due to be changed, one of the units sends to the other

unit a signal indicating at which frame the transcoding will change.

By indicating a frame number at which the transcoding function will change, it allows time

for both the sending unit and receiving unit to prepare for processing data according to the new transcoding function. Preferably the indication of the frame at which transcoding change will occur is sent more than one time frame in advance. This has the advantage that should the indication not be successfully transmitted, re-transmission in a subsequent frame can still be made, and in

sufficient time.

The indication can be sent up to X time frames in advance, where X can be, for example, between 10 and 20 inclusive, for example 15.

Agreeing the time frame at which to change transcoding, has the advantage that a mute period is no longer required, in particular when switching from a call connection involving a single slot per frame to a call connection involving two more slots per frame.

Preferably the slot or slots in the time frame at which the transcoding function changes and used for transmission according to the old transcoding function, or the new transcoding function contain data such that a first part of a first slot will contain data encoded according to the old transcoding function if the first slot is used for transmitting data

according to the old transcoding function, and a second part of the slot will contain data

coded according to the new transcoding function if the slot is used for transmitting data

according to the new transcoding function.

In slots including data according to both transcoding functions, preferably the data in the

first part occupies the first (N-S) x B/N bits of the slot, where S is the slot number, B is the number of bits transmitted in the slot, and N is the number of slots in the time frame,

and the second part occupies the remainder of the slot. By transmitting according to this rule, the receiving unit can deduce which samples result from the old transcoding function and which result from the new transcoding function.

Preferably the change in transcoding function occurs earlier at the transmitting unit than at the receiving unit to account for processing and transmission delays. Preferably the change in transcoding function occurs at a different time in the uplink direction than in the

downlink direction to account for processing and transmission delays. Furthermore, the

change in transcoding function preferably occurs within 10 to 20 time frames of being indicated. This is to ensure that, where a modem is connected to a receiving unit, the change is complete before the end of an answertone transmission from the modem, but allows sufficient time for the receiver to understand the frame at which the transcoding change will occur and to react in time.

Preferably the transcoding function is changed one frame earlier for data transmitted in the uplink direction, that is, from a subscriber unit to a base station, than in the downlink direction, that is, from the base station to the subscriber unit.

The present invention also relates to a corresponding method of reception. The present invention also relates to a corresponding transmitter. The present invention also relates to

a corresponding receiver. The present invention also relates to a corresponding communications means including a sending unit and a receiving unit. The present

invention also relates to a corresponding base station.

Detailed Description of Preferred Embodiments Figure 1 is a diagram illustrating an example of how samples are assigned to time slots in a frame (prior art),

Figure 2 is a simplified functional diagram indicating switching between transcoding functions (prior art),

Figure 3 is a diagram illustrating communications between a base station (BTE) and a subscriber unit (NTE) for changing transcoding function (prior art),

Figure 4 is a schematic diagram illustrating the system including a base station (BTE - Base Terminating Equipment) and subscriber unit (NTE - Network Terminating Equipment),

Figure 5 is a diagram illustrating frame structure and timing for a duplex link,

Figure 6 is a schematic diagram illustrating an example of transcoding function change by adding a second slot per frame to a call connection, and

Figure 7 is a diagram illustrating communications between a base station (BTE) and a

subscriber unit (NTE) for changing transcoding function.

The System

As shown in Figure 4, the preferred system is part of a telephone system in which the local wired loop from exchange to subscriber has been replaced by a full duplex radio link between a fixed base station and fixed subscriber unit. The preferred system includes the duplex radio link, and transmitters and receivers for implementing the necessary protocol. More specifically, there is a base station BTE connected via a access network an exchange

to the public network. There are also subscriber units NTE each connected to terminal

equipment TE such as a telephone handset, answering machine, facsimile machine or

modem.

There are similarities between the preferred system and digital cellular mobile telephone systems such as GSM which are known in the art. This system uses a protocol based on

a layered model, in particular the following layers; PHY (Physical), MAC (Medium Access Control), DLC (Data Link Control), NWK (Network).

One difference compared with GSM is that, in the preferred system, subscriber units are

at fixed locations and there is no need for hand-off arrangements or other features relating to mobility. This means, for example, in the preferred system directional antennae and

mains electricity can be used.

Each base station in the preferred system provides six duplex radio links at twelve frequencies chosen from the overall frequency allocation, so as to minimise interference

between base stations nearby. The frame structure and timing for the duplex link is

illustrated in Figure 5. Each duplex radio link comprises an up-link from a subscriber unit to a base station and, at a fixed frequency offset, a down-link from the base station to the subscriber unit. The down-links are TDM, and the up-links are TDMA. Modulation for all links is π/4 - DQPSK, and the basic frame structure for all links is ten slots per frame of 2560 bits i.e. 256 bits per slot. The bit rate is 512kbps. Down-links are continuously

transmitted and incorporate a broadcast channel for essential system information. When there is no user information to be transmitted, the down-link transmissions continue to use the basic frame and slot structure and contain a suitable fill pattern.

For both up-link and down-link transmissions, there are two types of slot: normal slots which are used after call set-up, and pilot slots used during call set-up.

Each down-link normal slot comprises 24 bits of synchronisation information followed by 24 bits designated S-field which includes an 8 bit header, followed by 160 bits designated D-field. This is followed by 24 bits of Forward Error Correction and an 8 bit filler, followed by 12 bits of the broadcast channel. The broadcast channel consists of segments in each of the slots of a frame which together form the down-link common signalling channel which is transmitted by the base station, and contains control messages containing link information such as slot lists, multi-frame and super-frame information, connectionless messages and other information basic to the operation of the system.

During call set-up, each down-link pilot slot contains frequency correction data and a training sequence for receiver initialisation, with only a short S-field and no D-field information.

Up-link slots basically contain two different types of data packet. The first type of packet, called a pilot packet, is used before a connection is set up, for example, for an ALOHA call request and to allow adaptive time alignment. The other type of data packet, called a normal packet, is used when a call has been established and is a larger data packet, due to the use of adaptive time alignment.

Each up-link normal packet contains a data packet of 244 bits which is preceded and followed by a ramp of 4 bits duration. The ramps and the remaining bits left of the 256

bit slot provide a guard gap against interference from neighbouring slots due to timing errors. Each subscriber unit adjusts the timing of its slot transmissions to compensate for the time it takes signals to reach the base station. Each up-link normal data packet comprises 24 bits of synchronisation data followed by an S-field and D-field of the same number of bits as in each down-link normal slot.

Each up-link pilot slot contains a pilot data packet which is 192 bits long preceded and followed by 4 bits ramps defining an extended guard gap of 60 bits. This larger guard gap

is necessary because there is no timing information available and without it propagation delays would cause neighbouring slots to interfere. The pilot packet comprises 64 bits of

sync followed by 104 bits of S-field which starts with an 8 bit header and finishes with a 16 bit Cyclic Redundancy Check, 2 reserved bits, 14 FEC bits, and 8 tail bits. There is

no D-field.

The S-field in the above mentioned data packets can be used for two types of signalling. The first type is MAC signalling (MS) and is used for signalling between the MAC layer

of the base station and the MAC layer of a subscriber unit whereby timing is important. The second type is called associated signalling, which can be slow or fast and is used for signalling between the base station and subscriber units in the DLC or NWK layers.

The D-field is the largest data field, and in the case of normal telephony contains digitised speech, but can also contain a non-speech data samples.

Provision is made in the preferred system for subscriber unit authentication using a challenge response protocol. General encryption is provided by combining the speech or data with a non-predictable sequence of cipher bits produced by a key stream generator which is synchronised to the transmitted super-frame number.

In addition, the transmitted signal is scrambled to remove dc components.

Changing Transcoding Function

In a preferred embodiment, to change the transcoding function in a way that avoids the need for a mute period, the transcoding function must be changed:

(1) Between samples,

(2) Earlier at the sending unit than at the receiving unit to account for processing and transmission delays,

(3) At different times in uplink and downlink directions to account for further processing and transmission delays, and (4) Within say 20 frames of the transcoding function change being indicated so that transcoding function changes are undertaken without undue delay.

The approach taken is that both the sending unit and receiving unit understand whether the contents of slots are encoding according to the old or new transcoding function in consequence both the sending unit and receiving unit can handle the data appropriately.

As regards requirement (1) above, the transcoding must be changed between samples to ensure correct transmission and reception. The slot at which the transcoding function change occurs includes data transcoded according to the old function in its first part and data transcoded according to the new function in its second part. The number of binary bits in the first part is chosen to be a multiple of the sample size resulting from the old

transcoding function. The number of binary bits in the second part is chosen to be a multiple of the sample size resulting from the new transcoding function. For example, a 32 kbps ADPCM transcoding produces 4 bit samples whereas a PCM transcoding produces 8 bit samples. Accordingly, the transcoding change in the slot at which ADPCM

transcoding is changed to (or from) PCM transcoding occurs after a multiple of 8 bits in

the first part.

The transmitting unit and receiving unit are respectively one of a base station and a subscriber unit. The base station sends a message to the subscriber unit indicating a future

frame (by its frame number) at which the new transcoding function will be used. This

indicated frame is designated as a time reference for when the new transcoding function

will start. Once both the subscriber unit and base station have knowledge of the frame at which the transcoding function will start, they can configure themselves to handle the data correctly. Furthermore by agreeing in advance at which frame the transcoding function will change,

even if the signal indicating that frame is corrupted, it can be resent in subsequent frames whilst still allowing time for correct reception of the indication of the frame at which the

transcoding function will change.

If the designated frame is F, then the following rules describe the contents of each slot (whether they contain the old service, the new service, or a proportion of both):

(1) all slots of frame F, F + 1, F + 2 etc. in use for the new service contain data encoded

according to the new service,

(2) all slots of frame F - 2, F - 3. F - 4 etc. in use for the old service contain data

encoded according to the old service,

(3) all slots of frame F - 1 in use for the old or new service or both are divided into a first part and a second part. The first part contains data encoded according to the old

service if the slot is in use in that service. The second part contains data encoded according to the new service if the slot is in use in that service. The first part occupies the first (N - S)*B/N bits of the slot, where S is the slot number, B is the

number of bits transmitted in the slot, N the number of slots in a frame. The second

part occupies the remainder of the slot. This rule (3) above ensures that the boundary between the old and new service is positioned such that all data is transmitted. This is because where there are 160 bits per slot and 10 slots per frame, rule (3) above indicates that the first part must consist of 160 - 16S bits where S is the slot number (an integer between 0 and 9 inclusive). Thus the

transcoding change must occur with some multiple of 16 bits in the first part to satisfy rule

(3). In the case of an ADPCM to PCM transcoding change, there must be a multiple of 8 bits in the second part to ensure the change occurs between samples. Thus this requirement is also satisfied.

An example is shown in Figure 6 where a digital data stream (a) is compressed into selected time slots (b) for transmission and them decompressed upon reception to reconstitute the digital data stream B.

Before the transcoding function change, slot 3 carries samples numbered 16 to 55 encoded using ADPCM. Thus slot 3 carries 20 bytes of digital data. Before the transcoding

function change slot 6 does not carry data from this digital data message. This occurs in

frame F - 2.

In the next time frame denoted F - 1 the transcoding function change occurs. Slot 3 carries

samples numbered 56 to 83 using ADPCM coding. This is 14 bytes of digital data. Slot 3 also carries, in its later portion, even PCM coded samples numbered 84 to 94. This

constitutes 6 bytes of data. In this frame, slot 6 carries nothing in its first part (8 bytes) and odd PCM samples numbered 85 to 107 in its second part, which is equivalent to 12

bytes of data. In the next frame, denoted F, slot 3 carries the even PCM samples numbered 96 to 134 and slot 6 carries the odd PCM samples numbered 109 to 147. These same rules are applied

both in the uplink direction from subscriber unit to base station, and in the downlink direction from base station to subscriber unit, accept that the transcoding function change occurs one frame later for the uplink than the downlink. If the change occurs at frame F on the downlink, it occurs at frame F + 1 on the uplink. The reason for this is apparent from Figure 6, namely that there is approximately a single frame delay between

transmission and reception, so the sending unit has to change its transcoding function approximately one frame earlier than the receiving unit.

The series of control signals between the base station and subscriber unit to effect the

transcoding change is shown in Figure 7. The first step is that the base station selects a frame at which the transcoding function will change. This frame is denoted F. The base station sends a control signal namely an enable service control signal which includes an indication of the frame F rather than sending a complete identifier it is sufficient to send F modulo 2^K where K is chosen so as to trade-off the size of the signal transmitted against

the amount of time in the future which may be represented. F modulo 2^K is the remainder when F is divided by 2^K, and is the equivalent to the last K binary bits of F, which is more

conveniently handled than the complete F number. The value for K of 4 gives up to 15

frames notice.

On receipt of the enable service request, the subscriber unit prepares to effect the

transcoding change at frame F, which is calculated as the next frame number at which the

lower K bits will match the K bits of F just received, and sends an enable service acknowledgement to the base station in response. The base station receives the enable service acknowledgement and sends a further control signal to indicate receipt of the acknowledgement, at which point the signalling exchange is complete.

In an alternative embodiment the signalling exchange is not completed until the transcoding function change has been completed. More specifically, in this alternative embodiment the base station does not send the final control signal until it has both received the enable service acknowledgement and has actually effected the transcoding change.

Transcoding Delay

To perform a transcoding function change without any mute period, the delay due to transcoding must be the same for both old and new transcoding functions. Otherwise there

will be a phase shift in the signal. If the phase shift is significant, it must be compensated

for by the transmitter responsible. For example, the subscriber unit may need to delay transmission uplink to the base station using PCM coding by one sample so as to compensate for the ADPCM transcoding delay. This is because PCM transcoding delay is less than ADPCM coding delay. So as to ensure that the delays with old and new

transcoding are matched, the delay of the transcoding function which is greater is applied as the delay to both. It is possible to decide for particular transcoding functions whether

truly seamless switching (ie. with zero mute period when a change in transcoding function is effected) is appropriate. Synchronisation of Encode/Decode Pairs

In ADPCM, and some other transcoding functions which significantly compress the data, the encoded data is dependent on the current input signal and previous signals also.

Parameters which represent longer term variations (such as general amplitude) in the signal are independently derived at the transmitting unit and receiving unit from the previous signals. The encoded data contains information about the short term variations only. The received encoded data is combined with the derived parameters to reconstitute the signal.

For the signal to be accurately conveyed, both transmitting and receiving unit derive the same values for each parameter, and ensues that these values be optimal. The methods by

which the parameters are derived are chosen so that the transmitter and receiver converge over time to the same values.

When a transmitting unit/receiving unit pair are started from an initial state, they have no

previous signals from which to calculate parameters, so must use agreed estimates. Since these are not optimal, a certain amount of time must pass before these estimates have converged to optimal values. In consequence, the signals decoded at the receiver are often distorted for a short period after the change of transcoding function has occurred until the parameter values have converged.

PCM is not a transcoding function which takes time to converge. This means for an ADPCM to PCM transcoding change the time required to converge is not a cause of significant distortion. For a PCM to ADPCM change, there is about a lmS distortion of the signal. This distortion is handled by arranging for both the transmitting unit and receiving unit to run their transcoders for the new transcoding function for an agreed number of samples before the transcoding change, so that they are already convergent. This requires both transmitter and receiver to have the capacity to encode or decode using both transcoding functions at the same time.

Alternative Embodiment

In an alternative embodiment, the subscriber unit cannot independently control switching of transcoding functions used for uplink and downlink but must change both at the same time. Also, it is not possible to precisely assign within a slot data encoded according to more than one transcoding function, in particular as described by rule (3) above. However, causing the transcoding function to change one frame later on the uplink compared to the downlink is useful in minimising the mute period and hence minimising the risk of a modem attached to a subscriber unit not establishing a call correctly. In consequence, the period for which the subscriber unit mutes output to the subscriber terminal equipment

need only last for a period of a single frame to avoid a corrupted received signal being sent to the subscriber terminal equipment.

In a subscriber unit according to the second embodiment, the transcoding happens as soon

as the input signal value is received. It is then buffered along with subsequent transcoded samples until the next transmit burst. Equivalently, when receiving, the encoded data is

buffered, and not decoded until it is required to form the output signal. So, referring to Figure 6, the transcoding must be changed at the transmitting unit A around the start of slot 1 of frame F - 1, and the inverse transcoding changed at the

receiving unit B around the start of slot 0 of frame F. Considering that a subscriber unit

is both a transmitting unit (on the uplink) and a receiving unit (on the downlink), the transcoding at the uplink encoder of the subscriber unit occurs roughly 9 slots earlier than at the downlink decoder of the subscriber unit.

In fact, they are between (1 + 0.04) frames and (1 + 0.08) frames apart because of other

factors. The main other factor is that the uplink is transmitted one whole slot in advance, see Figure 5, so this makes the difference close to 1 frame. Another factor is the adaptive timing advance (the amount by which the subscriber unit advances its transmit timing to compensate for propagation delay) which can take a value between 0 bits and 104 bits. A third factor is that the transmitted data occupies only 160 of the 256 bits in the slot; this

adds another 96 bits of difference.

Since a subscriber unit according to the second embodiment can only change its transcoding both uplink and downlink at the same time, by offsetting the uplink switch by 1 frame, the period of corruption of data received on the downlink and uplink is reduced

to a maximum of 0.08 of a frame, which is approximately a few samples. The base station

and subscriber unit both mute their transmission of digital data corresponding to information, such as audio signals, for a period of 1 frame so as to avoid corrupted data being sent. On both the uplink and the downlink, the mute period is sufficiently short that

many types of modems can successfully establish a call connection despite the break in

signals received.

Claims

1. A method of transmitting a digital data message in time slots within fixed length time frames between a sending unit and a receiving unit, a time slot or slots in successive

frames being assigned for transmission of the message, the message being transcoded for

transmission according to a transcoding function, in which, upon the transcoding function being due to be changed, one of the units sends to the other unit a signal indicating at which frame the transcoding will change.

2. A method as claimed in claim 1, wherein the signal indicating at which frame the transcoding will change is sent more than one time frame in advance.

3. A method as claimed in claim 2, wherein the signal is sent between 10 and 20 time

frames in advance.

4. A method as claimed in claim 3, wherein the signal is sent 15 time frames in advance.

5. A method as claimed in any one of the preceding claims, wherein the slot or slots, in

the time frame at which the transcoding function changes and which is used for

transmission according to the old transcoding function or the new transcoding function, contain data such that a first part of a first slot will contain data encoded according to the old transcoding function if the first slot is used for transmitting data according to the old transcoding function, and a second part of the first slot will contain data coded according

to the new transcoding function if the slot is used for transmitting data according to the

new transcoding function.

6. A method as claimed in any one of the preceding claims, wherein in a slot including data according to both transcoding functions, the data in the first part occupies the first (N- S) x B/N bits of the slot, where S is the slot number, B is the number of bits transmitted

in the slot, and N is the number of slots, in the time frame, and the second part occupies

the remaining bits of the slot.

7. A method as claimed in any one of the preceding claims, wherein the change in transcoding function occurs earlier at the sending unit than at the receiving unit to account for processing and transmission delays.

8. A method as claimed in any one of the preceding claims, wherein the change in transcoding function preferably occurs within between 10 and 20 time frames of being

indicated.

9. A method as claimed in any one of the preceding claims, wherein the units are a

subscriber unit and a base station.

10. A method as claimed in claim 9, wherein the change in transcoding function occurs

at a different time for data transmitted from a subscriber unit to a base station than for data transmitted from the base station to the subscriber unit.

11. A method as claimed in claim 10, wherein the transcoding function is changed one

frame earlier for data transmitted from a subscriber unit to a base station than for data transmitted from the base station to the subscriber unit.

12. A method of receiving a transcoded digital data message transmitted in time slots within fixed length time frames from a sending unit to a receiving unit, a time slot or slots in successive frames being assigned for reception of the message, the message being

reverse transcoded for reception according to a reverse transcoding function, in which, upon the transcoding of the message being due to be changed, the receiving unit receives from the sending unit a signal indicating at which frame the reverse transcoding will need

to change.

13. A transmitter unit for transmitting a digital data message in time slots within fixed length time frames, comprising means for assigning a time slot or slots in successive frames for transmission of the message, means for transcoding the message according to a transcoding function, and means responsive to a change being due in the transcoding function for generating a signal indicating at which frame the transcoding will change.

14. A receiver unit for receiving a transcoded digital data message in time slots within

fixed length time frames, comprising means for assigning a time slot or slots in successive frames for reception of the message, means for reverse transcoding the transcoded message according to a reverse transcoding function, and means responsive to a signal indicating at which frame the transcoding will change for changing the reverse transcoding function

associated with the reverse transcoding means.

15. A communication system including a transmitter unit as claimed in claim 13 and/or

a receiver unit as claimed in claim 14.