AU711434B2

AU711434B2 - Transmission burst for discontinuous transmission

Info

Publication number: AU711434B2
Application number: AU40950/96A
Authority: AU
Inventors: Laurent Cruchant; Pierre Dupuy
Original assignee: Alcatel NV
Current assignee: Alcatel Lucent NV
Priority date: 1995-01-24
Filing date: 1996-01-15
Publication date: 1999-10-14
Anticipated expiration: 2016-01-15
Also published as: FR2729806A1; FR2729806B1; FI960300A; DK0724342T3; ATE200369T1; EP0724342B1; EP0724342A1; AU4095096A; ES2155582T3; DE69612314D1; FI960300A0; DE69612314T2

Abstract

The stuffing packet is used in a radio telecommunications system. This includes a transmitter with a packet forming circuit (BU) introducing a synchronisation sequence in the synchronisation field taken from a synchronisation set (S1-S8). A selection circuit (SC) produces this sequence as a training sequence (S1) or a distinct identification sequence (S2) according to whether the packet is respectively a speech packet or a stuffing packet. A receiver calculates the correlation between a sequence received in the synchronisation field and a reference sequence. A control circuit produces the training sequence as a reference and when the correlation is unsatisfactory it produces the identification sequence as the reference.

Description

P/00/011 28/5/91 Regulation 3.2

AUSTRALIA

Patents Act 1990 e

ORIGINAL

COMPLETE SPECIFICATION STANDARD PATENT Invention Title: p p "TRANSMISSION BURST FOR DISCONTINUOUS TRANSMISSION" The following statement is a full description of this invention, including the best method of performing it known to us:- This invention relates to a transmission burst that is specially designed for discontinuous transmission. The invention can be applied to radiocommunications systems in particular and, to facilitate its presentation, reference will be made to the GSM system, the reputation of which is now well established.

In the GSM, discontinuous transmission is provided for conveying speech signals for instance. A speech signal is normally conveyed by successive bursts, each transmitted in the same time slot of successive frames. However, when there is no speech to be conveyed, it has become clear that it was unnecessary to transmit all ."lGg these bursts and that it was sufficient to transmit a silence descriptor, from time to time and over a few bursts, in order to restore the noise present in the absence of speech.

Thus, the receiver, which continuously listens to the time slot assigned thereto, receives either continuous speech or a silence descriptor and then nothing until the next descriptor. More information is given in the GSM recommendation 05.08 version 3.6.1.

Everything happens satisfactorily so long as the transmitter does not transmit anything between two silence descriptors. But the GSM further imposes that a base station continuously transmits on the frequency which carries the main broadcast channel, i.e. the BCCH channel. Thus, when there is no information to be transmitted, S"0 the base station transmits dummy bursts which do not have any specific meaning.

*00oo: 0 If the speech signal is carried by the BCCH frequency, either entirely or partly, when frequency hopping is used, the receiver will receive dummy bursts.

The speech bursts and the dummy bursts have the same structure, i.e. a synchronization field placed substantially in the center of the burst and a data field placed on either side of the synchronization field. In both cases, a training sequence known to the receiver is placed in the synchronization field.

The receiver includes, in particular, correlation means for correlating the sequence received in the synchronization field with a replica of the locally generated training sequence. The result of this correlation is generally used to measure the quality of the received burst which, in the case of a dummy burst, is very likely to be deemed satisfactory. If, furthermore, the contents of the data field of a dummy burst can be assimilated to speech data given the various controls performed on that field, the receiver will then interpret the dummy burst as being a speech burst. This situation, although rare, may still occur and will then generate at the level of the receiver an unpleasant noise know by those skilled in the art as "pling-plong" noise. The production of such a noise is not desirable.

It must be specified at this stage that the training sequence is identified from several synchronization sequences. In the case of the GSM, eight distinct sequences are provided, each being identified by a Training Sequence Code or TSC. The sequences corresponding to the different TSCs have good autocorrelation and intercorrelation properties.

Those skilled in the art who are concerned with the distinction between a speech burst and a dummy burst will thus place in the synchronization field an identification sequence which is different from the eight synchronization sequences that can be encountered in a GSM network, this identification sequence having good intercorrelation properties with each of the eight synchronization sequences. Although satisfactory, this solution requires the introduction into the GSM of a new sequence, which is already a constraint in itself, and may further require a substantial modification of the equipment already in service.

It is thus an objective of the invention to define and process a dummy burst which does not require any identification information not already known to the system.

A method for implementing a frequency hopping communication system producing, in a communication system having a predetermined set of synchronisation sequences, a transmission burst signal having a data field and a synchronisation field, the method including the steps of: making a selection of one of the synchronisation sequences for inclusion in the synchronisation field to identify speech data; and making a selection of a second synchronisation sequence for inclusion in the synchronisation field to identify dummy data. It is thus not necessary to provide for an identification sequence that is specific to the dummy bursts. A sequence already known to the base station is used instead.

The desired result is effectively achieved since, generally, the different synchronization sequences are chosen so that they have good intercorrelation properties.

w ~f 0NJ 'V ~~FS Thus, if the identification sequence selected from the synchronization sequences is not really optimal, it makes it possible to substantially decrease the number of dummy bursts detected as being speech bursts and such an event, which is already rare without implementing the invention, is extremely unlikely to occur with the transmission burst proposed above.

All the elements required for the implementation o the invention are already provided for in the system and its application will only require minor equipment modification.

Appropriately, the transmission burst is defined so that the identification sequence 0000 is predetermined.

0o *o Thus, either the system itself or the knowledge of the training sequence determines unequivocally the knowledge of the identification sequence.

•A transmitter for producing a transmission burst signal having a data field and a synchronisation field, the transmitter including: a burst forming means for forming a transmission burst and having a defined set of synchronisation sequences; and 0000 selection means for selecting one of the sequences for inclusion in the °•co synchronisation field when the data field contains speech data and for selecting a second sequence for inclusion in the synchronisation field when the data field contains dummy data.

Thus, the terminal, which implicitly or explicitly knows the identification sequence, can search for the correlation between this sequence and the sequence it receives in the synchronization field. If the result of this correlation is better than that of the correlation performed with the training sequence, the burst will almost certainly be a dummy burst. This increases the quality of the decision made in the receiver as to the type of burst being dealt with, ie. whether it is a speech burst or a dummy burst.

According to a still further aspect of the invention, there is provided a transmitter including a burst-forming circuit for inserting a synchronization sequence in a synchronization field of a transmission burst, the transmitter further including a selection K circuit for producing the synchronization sequence respectively as a training sequence or 4A as an identification sequence depending on whether the burst is respectively a speech burst or a dummy burst.

A preferred embodiment of the invention, provided for the GSM system, will *c* cc e e e :t et cf *f c f C c

PA

*ft. O4 I3 0** now be described, by way of example only, with reference to the accompanying drawings, in which: Figure 1, is a simplified diagram of a transmission burst, Figure 2, shows means for implementing the invention in a transmitter, and Figure 3, shows means for implementing the invention in a receiver.

Elements common to the various figures are identically referenced.

It must be recalled at this stage that the GSM provides for eight distinct synchronization sequences that are identified by a Training Sequence Code or TSC.

•0I14 Before setting up a call between a terminal and a base station, the base station specifies to the terminal which of the synchronization sequences will be used as the training sequence for the duration of the call.

Figure 1 recalls the simplified structure of a transmission burst transmitted in a frame time slot. The burst includes a synchronization field S here placed substantially in the center of the burst. This field is preceded by a first data sub-field D and followed by a second data sub-field The combination of the first and the second S data sub-fields will be referred to as data field since they both correspond to the same entity. Systems other than the GSM may provide for the synchronization field S to be placed in any other location of the burst.

Referring to Figure 2, the base station, which acts as the transmitter, classically S" includes a burst-forming circuit BU. This circuit BU includes a memory M in which the different synchronization sequences S1 to S8 are stored. It also includes a multiplexer MUX which communicates one of the synchronization sequences to a register R in response to a control signal C. The register includes the sequence which will then be placed in the synchronization field S of the burst.

Prior to the invention, this control signal C corresponded to the sequence code

TSC.

According to the invention, there is now provided a selection circuit SC which produces the control signal C which is a function of the sequence code TSC and a burst identifier BI. The identifier BI specifies whether the burst to be transmitted is a speech burst or a dummy burst. When the burst is a speech burst, the control signal C corresponds to the sequence code TSC, as in the prior art, and the training sequence is placed in the register R. Conversely, when the burst is a dummy burst, the control signal does not correspond to the sequence code TSC and an identification sequence is placed in the register R. Several solutions are possible at this stage.

A first solution consists in deciding that a specific synchronization sequence will never be used as a training sequence. This specific sequence will thus be used as an identification sequence and will systematically be placed by the selection circuit SC in the synchronization field S of all the dummy bursts. Appropriately, this identification sequence will be the synchronization sequence which has the best intercorrelation properties with all the other synchronization sequences.

I. A second solution consists in establishing a relation between the training S sequence used for a speech burst and the identification sequence placed in a dummy burst. As an example, the identification sequence will be the sequence Si+ 1 which immediately follows the training sequence Si. The selection circuit is also particularly simple in this case: if the burst identifier BI is given the value 0 if it is a speech burst and the value 1 if it is a dummy burst, the control signal will correspond to the sum of the values of TSC and BI modulo 8.

A third solution consists in arbitrarily choosing an identification sequence which see.

is distinct from the training sequence. This is not necessarily the easiest solution to :it implement.

It is thus obvious that a plurality of solutions are within the scope of those skilled in the art.

Moreover, it can be noted that according to the first or the second solution, the identification sequence is predetermined. According to the first solution, the identification sequence is always the same sequence and, according to the second solution, it is the knowledge of the training sequence which determines the knowledge of the identification sequence.

The determination of the identification sequence will appropriately be optimized so as to improve the performances of the receiver.

Referring to Figure 3, the receiver classically includes a correlation circuit CORR for producing the result of the correlation RC between the sequence of the synchronization field S found in the received burst and a reference sequence S

R

This reference sequence S R can be produced by a burst-forming circuit BU which is identical to that of the transmitter.

Prior to the invention, the control signal C of this circuit BU corresponded to the sequence code TSC that the receiver had previously been given.

According to the invention, there is provided a control circuit CC which receives the correlation result RC and accordingly produces the control signal C.

When the control circuit CC is advised that a sequence is available in the synchronization field, it produces in a first stage the control signal C so that the reference sequence SR corresponds to the training sequence. It then receives the corresponding correlation result RC A from the correlation circuit CORR. If the result A is satisfactory, i.e. if it exceeds a first threshold T1 the received burst is a speech burst. Conversely, if the result A is not satisfactory, the control circuit CC 0:00 produces, in a second stage, the control signal C so that the reference sequence SR :.105 corresponds to the identification sequence. If the corresponding correlation result RC I is satisfactory, it is then almost certain that the received burst is a dummy burst.

In summary, and to adopt a risk-free procedure, the following method may be oo o S.0. used: if A is more than TI, the burst is a speech burst; if A is less than T1 and I is more than T1, the burst is a dummy burst; and if both A and I are less than T1, the burst is a speech burst.

A second threshold T2, lower than T1, may also be defined so that: if A is more than T1, the burst is a speech burst; if A is less than T2, the burst is a dummy burst; if A lies between T1 and T2 and if I is more than T1, the burst is a dummy burst; and if A lies between T1 and T2 and if I is less than T1, the burst is a speech burst.

It can be seen that several means may be used to significantly increase the safety of the decision made depending on the nature of the burst received in the receiver.

Claims

1. A method for implementing a frequency hopping communication system producing, in a communication system having a predetermined set of synchronisation sequences, a transmission burst signal having a data field and a synchronisation field, the method including the steps of: making a selection of one of the synchronisation sequences for inclusion in the synchronisation field to identify speech data; and making a selection of a second synchronisation sequence for inclusion in the synchronisation field to identify dummy data.

2. A method according to claim 1, wherein the second sequence is predetermined from the set of synchronisation sequences.

3. A method according to claim 1, wherein the second sequence is selected according to a predefined relationship with the first sequence. o**c S

4. A method according to claim 1, wherein the second sequence is arbitrarily selected from the set of synchronisation sequences, after the first sequence has been selected.

5. A receiver for receiving a transmission signal as claimed in claim 1, including: correlation means for correlating a reference synchronisation sequence with a set of synchronisation sequences so as to produce a correlation output; and S.. selection means for initially selecting a first synchronisation sequence as the reference sequence and, if and when the correlation output is not satisfactory, selecting a second synchronisation sequence as the reference sequence.

6. A transmitter for producing a transmission burst signal having a data field and a synchronisation field, the transmitter including: a burst forming means for forming a transmission burst and having a defined set of synchronisation sequences; and selection means for selecting one of the sequences for inclusion in the synchronisation field when the data field contains speech data and for selecting a second sequence for inclusion in the synchronisation field when the data field contains dummy data.

7. A transmitter according to claim 6, wherein the second sequence is predetermined from the set of synchronisation sequences.

8. A transmitter according to claim 6, wherein the second sequence is selected according to a predefined relationship with the first sequence.

9. A transmitter according to claim 6, wherein the second sequence is arbitrarily selected from the set of synchronisation sequences, after the first sequence has been selected.

A transmitter according to any one of claims 6 to 9, wherein the set of synchronisation sequences consists of eight distinct sequences.

11 A frequency hopping communication system adapted to utilise a transmission burst structure of the type including a data field and a synchronisation field, the system :comprising: *o a predetermined set of training sequences for selective allocation to said synchronisation field, wherein, when the burst contains information signals in the data field, the system allocates a first training sequence in the synchronisation field from said predetermined set, and wherein, when the burst does not contain information signals in the data field, the system S• •allocates a second training sequence in the data field from said predetermined set.

12 A frequency hopping communication system as claimed in claim 11, wherein the second sequence is selected according to a predefined relationship with the first sequence.

13. A method according to claim 11, wherein the second sequence is arbitrarily selected from the set of synchronisation sequences, after the first sequence has been selected.

14. A transmitter and a receiver substantially as herein described with reference to figures 1 to 3 of the accompanying drawings. A method of producing a transmission burst signal substantially as herein described with reference to the examples described herein.