GB2358562A

GB2358562A - Radio communications system for use with two different standards/protocols

Info

Publication number: GB2358562A
Application number: GB0100887A
Authority: GB
Inventors: Peter William Stephens; Alister John Bailey
Original assignee: Simoco International Ltd
Current assignee: Simoco International Ltd
Priority date: 2000-01-12
Filing date: 2001-01-12
Publication date: 2001-07-25
Also published as: GB0000634D0; WO2001052563A2; WO2001052563A3; GB0100887D0; AU2862201A

Abstract

The application is generally concerned with a communications system that can accommodate radio standards or protocols having different timeslot structures, preferably the TETRA and APCO standards. In one embodiment, APCO frames are generated, coded for transmission, and appended into groups before being placed into TETRA timeslots. In another embodiment, APCO generated frames are coded using a TETRA coding process, the coding process being modified so that resulting transmission frames contain less bits than usual and fit into the TETRA timeslot structure. Various techniques for modifying the coding are disclosed, including changing the coding rate or changing the relative priority associated with the data bits. Alternatively the frame generation process may be modified. Other embodiments are concerned with timeslot stealing techniques for the APCO/TETRA system (figure 1). Further embodiments are concerned with implementing encryption, e.g. APCO DES encryption, in the APCO/TETRA system.

Description

2358562 Radio Communications System The present invention relates to radio

communications systems and in particular to digital radio communications systems which transmit data in discrete timeslots, such as the TETRA (TErrestrial Trunked RAdio) system.

In such digital radio systems, the structure and data capacity of the timeslots, which are usually of constant length, are usually predefined, typically in the radio communications air-interface Standard with which the radio system complies, and all data to be is transmitted in the system must be arranged in accordance with the defined data transmission (timeslot) structure.

Where data, particularly data relating to a continuous data source (such as a speech or video signal), is to be transmitted in such a digital radio system, the initial input signal (e.g. speech or video signal) is usually converted for transmission into a series of discrete frames or portions of digital bits, usually of constant length, so as to, for example, facilitate processing of the signal.

The raw digital data frames generated from the input signal are usually then further processed or "coded" prior to transmission, as is known in the art, to, for example, add additional data bits to the raw digital data to provide protection against errors that may occur during transmission (which errors may be more likely given that a radio transmission is inherently more susceptible to transmission induced errors than some other forms of data transmission). Such coding is generally referred to as "channel coding", and the overall channel coding which is applied is usually made up of a number of different coding techniques.

one such form of coding is "Forward Error 2 Correction" (FEC) which codes the user data by addin(31 "redundant" data in algorithmic fashion such that if errors occur in the coded data, then their presence ca be detected and in some cases subsequently corrected for. Such coding schemes can be block or stream based and can additionally include techniques such as the interleaving of bits in the data stream. Examples of, suitable coding schemes include Hamming codes, Reed-Solomon codes and Golay codes. A Cyclic RedundEL y Check (CRC) is another form of coding which provides 1 good error detection, but does not assist in error correction. Channel coding will usually include, but: Is not limited to, forward error correction.

Since it is the "coded" data which is actually is transmitted and must therefore fit into the transmissj timeslot structure, both the raw data frame generaticii process and the channel coding process used in a give] radio system must be such that they provide coded datil for transmission in a form compatible with the predefined data (timeslot) structure to be used for transmission. For this reason the raw data frame generation (e.g. speech coding) process and channel coding process used by a radio system will therefore, typically, also be predefined and fixed for the radio system, usually by the radio Standard with which the radio system complies. Thus, for example, the TETRA radio Standard has a defined data transmission timeslct structure, and channel coding and speech encoding processes.

The channel coding and raw data generation processes intended for use with one data transmission timeslot structure will usually be optimised to that timeslot structure. They will therefore not normally be, and not be expected to be, readily suitable for u e with a second, different air-interface data transmissilc timeslot structure, because they would generate data Ec transmission at a rate and in frames of a size which iir 3 not appropriate for the other transmission data structure.

This may not be a problem where radio systems are intended to be selfcontained and not intended to interoperate with other systems. However, it is becoming increasingly desirable for radio units to be able to operate in more than one radio system, which systems may well use different air-interface data transmission structures. For example, it may be desired to introduce a newer radio system operating in accordance with a new radio standard into an area where there is a pre-existing radio system which uses a different radio standard. It could in such circumstances be useful if radio units could use both the new and the old systems, and if communication between the radio units provided for one system (e.g. the new one) and existing, unmodified radio units of the other (e.g. older) system could be achieved without the need to modify at least one of the systems, i.e. if the two radio systems could be interconnected.

One example of where this situation has arisen is in North America, where it is desired to introduce TETRA compliant radio systems into areas where radio systems that operate according to the APCO Project 25 Standard (the Project 25 Standard agreed by Associated Public Safety Communications Officers, Inc. (APCO), National Association of State Telecommunications Directors (NASTD), and agencies of the US Federal Government (FED) and Telecommunications Industry Association (TIA), (see, for example, the 102 Series of Documents by TIA, such as the Project 25 System and Standards Definition document (TIA/EIA TSB102-A)) - herein referred to as 11APC011), already exist, and to provide for dual-mode TETRA/APCO operation of radio units, and to allow new TETRA compliant radio units to communicate with existing, unmodified APCO radio units, i.e. to interconnect the TETRA and APCO systems.

4 The most straightforward way to provide such dual-mode operation would be to equip radio units wit.-I the data generation (e.g. speech coders/decoders) and channel coding processes of the different radio systenz (e.g. TETRA and APCO), which could then be used selectively as desired. However, the Applicants beliAe that in some circumstances it may be preferable, for example as regards data processing requirements, to reduce the number of raw data generation (e.g. speech coding) processes and/or channel coding processes provided in a given radio unit. However, there is thei the problem of how to enable a radio unit which does n:t have all the defined data generation processes and/or all the defined channel coding processes (e.g. only haE a single data generation process and/or only has a single channel coding process), to use two different air-interface data transmission structures. It is ar object of the present invention to provide solutions t this problem.

According to a first aspect of the present invention, there is provided a method of operating a.

radio unit for use in a digital radio communications system, in which system data can be transmitted in twc different pre-defined data structures each comprising a succession of timeslots, there being defined for each data structure a process for generating digital data Jor transmission from an input signal and a process for channel coding the generated data prior to its transmission, which data generation and channel codimc processes provide data for transmission in the form ()J discrete frames, the method comprising:

generating digital data for transmission from an input signal in accordance with the data generation process defined for a first one of said transmission data structures; coding the data generated from the input signal prior to its transmission using the channel coding process defined for said first one of said transmission data structures; when it is desired to transmit the coded data using the first of the two different predefined data structures, transmitting the data as defined for that transmission data structure; and when it is desired to transmit the coded data using the second of the two different predefined data structures, apportioning the coded data frames generated according to the data generation and channel coding processes for the first defined data structure into timeslots of the second defined data structure in such a way that each timeslot of the second defined data structure contains only entire coded data frames.

According to a second aspect of the present invention, there is provided a radio unit for use in a digital radio communications system, in which system data can be transmitted in two different pre-defined data structures each comprising a succession of timeslots, there being defined for each predefined data transmission structure a process for generating digital data for transmission from an input signal and a process for channel coding the generated data prior to its transmission, which data generation and channel coding processes provide the data for transmission in the form of discrete frames, the radio unit comprising:

means for generating digital data for transmission from an input signal in accordance with said data generation process defined for a first one of said transmission data structures; means for coding the data generated from the continuous input signal prior to its transmission using said channel coding process defined for said first one of said transmission data structures; and means for, when it is desired to transmit the coded data using the second of the two different predefined data structures, apportioning the coded data frames 6 generated according to the data generation and chann(Il coding processes for the first defined data structurel into timeslots of the second defined data structure in such a way that each timeslot of the second defined clata structure contains only entire coded data frames.

These aspects of the present invention relate to the situation where a radio system has two different predefined transmission data timeslot structures (wh:-c-.

would typically each be in accordance with a different defined radio standard), and each transmission data structure has predefined data generation and channel coding processes, such as might be the case in a Joint TETRA/APCO system. However, instead of using the dat-3.

generation and channel coding process intended for ezic, transmission data structure as appropriate, in these aspects of the present invention the same data generation and channel coding processes are always useL and, furthermore, the processes used are those definei for the same data transmission structure. Thus, the riw data frames for transmission are generated according A the process defined for one of the defined transmissiA data structures, and raw data frames are then channel coded using the channel coding protocols defined for that same transmission data structure. Thus, for example, in an embodiment of an APCO/TETRA system whl71 is in accordance with these aspects of the present inven tion, the APCO data generation and channel coding processes would always be used, regardless of whethe:t.-, n APCO or TETRA transmission was to be made.

This arrangement has the advantage that the raw data generation and subsequent channel coding (e-g.

forward error correction) processes used are both intended for use with the same, first defined transmission data structure (e.g. radio standard), ani will therefore, typically be optimised to each other i d for transmission with the first data structure. It ma also permit easier retrospective modification of existing equipment for use with both data structures, and permit easier interoperability between a new and an existing radio system, since, for example, no modifications to the data generation and channel coding processes in existing equipment configured for the first data structure (radio system) would be required.

The data generated for transmission in this arrangement can, of course, be directly transmitted in the form of the defined transmission data structure for which the data generation and channel coding processes are defined. However, where it is desired to transmit the data in the other (i.e. second) defined data structure, the coded data frames (which are at that point entirely in accordance with the first defined data structure) must be fitted into that other, second defined data transmission structure.

In these aspects of the present invention, this is achieved by apportioning the coded data frames generated as for the first defined transmission data structure into the timeslots of the second defined data structure in such a way that each timeslot of the second structure contains only entire, complete coded frames (and thus no partial, or incomplete, coded frames). This arrangement effectively means that there is no spreading of a coded 2S data frame across timeslots of the second data structure. This is believed to be advantageous, since it helps to ensure that any successfully received timeslot can be fully processed by the receiver without the need to wait for another timeslot to be received, and helps to ensure that the improper reception or loss of a timeslot does not affect successful reception of data in other timeslots (as usually an entire coded data frame is needed for successful decoding).

Preferably, the arrangement is such that the same integer number, which will typically be two or more, of complete coded data frames produced using the data generation and channel coding processes defined for the 8 first transmission data structure are placed into ea# timeslot of the second defined data transmission structure.

It is not essential to use all of the data capaci y in each timeslot of the second data structure, altho,c preferably as much of the timeslot data capacity is Is d as possible. Equally, it would be possible to omit crie of the coded data frames from the transmission if notl all of those frames will fit into the second transmission data structure timeslots, although this li less desirable as then some data would be lost.

However, it could be possible in some cases to satisfactorily interpolate the missing data frames friop the data frames that are transmitted.

The arrangement of these aspects of the inventio n is particularly applicable to a combined APCO and TETR system, where one generates and channel codes the dat:a for transmission according to the processes defined f the APCO standard. In that case, a convenient way to apportion the APCO generated coded data frames into t-h normal sized TETRA timeslot structure for transmission over a TETRA air-interface is to place three APCO code:

frames into every TETRA timeslot (the Applicants have found that three APCO coded data frames will fit into one TETRA timeslot).

As will be appreciated from the above, the abovel aspects of the present invention are particularly suitable when the data generation and channel coding processes defined for the one transmission data structure produce data frames of a size and at a ratel that is less than or equal to the data frame capacityl f the other, second transmission timeslot structure.

However, it may be that some combinations of raw data frame generation and channel coding processes will result in more data in a given time interval than can e carried in that time interval by one of the defined (13.-a structures to be used for the data transmission. Th].3 could, for example, be the case where it is desired to use the channel coding process defined for one defined transmission data structure on raw data frames generated according to the process defined for another, different defined transmission data structure. For example, the Applicants have found that coding normal APCO generated raw speech frames using normal TETRA channel coding produces too much coded data (i.e. at too high a data rate) to fit into the TETRA transmission timeslot structure. In such a case the arrangement of the above aspects of the invention could be less useful.

It may be therefore that in some circumstances a different arrangement for allowing a radio unit to use two different air-interface data transmission structures would be desirable.

According to a third aspect of the present invention, there is provided a method of operating a radio unit for use in a digital radio system which can use two predetermined data transmission structures, each comprising timeslots of a defined size, and in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, each transmission data structure having its own defined data generation and channel coding processes, the method comprising:

generating from an input signal frames of digital data using one of the defined data generating processes; processing the generated frames of digital data using one of the defined channel coding processes to produce a series of coded data frames for transmission; and modifying the data generating process and/or modifying the channel coding process when it is intended to transmit the data in accordance with at least one of the defined data structures, so as to alter the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be presentili each frame when using the unmodified as-defined processes.

Thus according to a fourth aspect of the presenz.

invention, there is provided an apparatus for use in a radio unit for use in a digital radio system which car use two predetermined data transmission structures, (-:,a h comprising timeslots of a defined size, and in which system data to be transmitted is generated from an ii t signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, each transmission data structure having its own defined data generation and channel coding processes, the apparatus comprising:

means for generating from an input signal frames f digital data using one of the defined data generating processes; means for processing the generated frames of digital data using one of the defined channel coding processes to produce a series of coded data frames fo transmission; and means for modifying the data generating process and/or modifying the channel coding process when it IS intended to transmit the data in accordance with at least one of the defined data structures, so as to al4b the amount of data in each coded data frame to be transmitted as compared to the amount of data that wc,.

be present in each frame when using the unmodified as defined processes.

In these aspects of the invention the data generation and/or channel coding process used are modified so as to alter the size of the coded data frames to be transmitted. This provides a technique J-'C altering, for example, reducing, the amount of overall data (i.e. the total of both the wanted user data and any added data such as "coding" or error correcting "redundancy" data) to be transmitted, but while avoid the need to, for example, omit entire data frames from the transmission. It is believed that this technique may also be preferable to, for example, providing additional, alternative channel coding and/or data generation processes in the radio unit.

In these aspects of the invention, the processes could be modified so as to reduce the amount of overall data in each coded data frame (i.e. the size of each coded data frame) to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as-defined processes, for.example where the unmodified processes produce too much data to fit into the timeslot structure to be used. The arrangement could also be so as to increase the amount of overall data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as defined processes. This may be useful where the unmodified processes leave spare capacity in each timeslot.

The way that the processes are modified to alter the size of the coded data frames can be selected as desired. For example, the raw data frames could be generated at a different, e.g. lower, data rate.

2S However, preferably the raw data frames are generated at the same rate as normal, so as to ensure that the amount of raw, or wanted user data is not altered, and instead the processes are modified such that a different amount, e.g. less, coded data is generated from that amount of 'raw' data. The Applicants believe that this is a preferable way to proceed, since it ensures that the amount of 'raw' data transmitted remains the same.

The way that this is achieved can be selected as desired.

In many radio system data generation techniques, the raw data for transmission generated from an input signal includes data bits of different relative 12 importance classifications. For example, many speec-. coding techniques (including those used by TETRA and.] APCO) generate digital data from the raw speech whic' includes bits of differing levels of relative importance, for example according to how important t.-e bits are for accurate reproduction of the speech.

The channel coding techniques in such radio syste s are then often also arranged to code the raw data in. accordance with the relative importance levels of the raw data bits, and to, for example, apply greater eric protection to the more important bits. This means t.tatthe number of additional "redundant" bits added by ttE channel coding per bit of the raw data typically var,jE depending on the relative importance levels of the r_=Y is data bits.

The Applicants have recognised that where the r.-=v data is classified into bits of differing levels of relative importance and the subsequent channel codinc introduces different numbers of additional bits into the coded data depending on the relative importance level. of the given bits to be coded, then one way therefore to adjust the overall amount of coded data generated fro a given amount of raw data is to modify the processes, :;i.h that the channel coding process is effectively appliec to a different classification of the importance levelE of the generated raw data bits, such that the channe-.4 coding produces a different, e.g. reduced, number of: coded bits from the raw data. In other words, the importance levels of bits in the raw data as effectiVEly seen by the channel coding process are modified such that the coded data has a different, e.g. reduced, number of bits in it than it would have had had the modification not taken place. This technique of effectively changing the number of raw data bits at Ole or more relative importance levels thus allows the overall number of bits in the coded data to be altere but while still using (and without needing to modify the predefined channel coding process.

The effective modification of the number of bits at each relative importance level could be carried out in the initial bit generation process, by modifying that process to generate different numbers of bits at each relative importance level to usual. Alternatively, or additionally, the processes could be modified such that the channel coding process interprets at least some of the raw data bits as having a different relative importance level to the level effectively allocated by the data generation process (e.g. to treat some or all high importance bits as having only medium importance for the purposes of the channel coding). The relative importance levels of one or more of the bits in the raw data could also be changed prior to channel coding of the data, if desired. These latter techniques would allow the raw data to be generated using the defined data generation process (e.g. speech coding process) for the system.

Thus, according to a fifth aspect of the present invention, there is provided a method of preparing data for transmission in a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames which contain bits of Iifferent levels of relative importance, which raw data frames are then further processed prior to transmission by a channel coding process in which the number of bits in the coded data generated from a given bit in the raw data varies depending on the relative level of importance of the raw data bit, and which system has a pre-defined data generation process, the method comprising:

causing the channel coding process to be effectively carried out on raw data in which the relative importance levels of one or more of the raw data bits differs to the relative importance level that the bits would have had under the defined data 14 generation process, so as to alter the amount of data each coded data frame to be transmitted as compared tc the amount of data that would be present in each frantE if the changing of the effective relative importance, levels had not taken place.

According to a sixth aspect of the present invention, there is provided an apparatus for preparirg data for transmission in a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames which contain bits of different levels of relative importance, which raw data frames are then further processed pri-di to transmission by a channel coding process in which de number of bits in the coded data generated from a gi.yEn bit in the raw data varies depending on the relative level of importance of the raw data bit, and which system has a pre-defined data generation process, thE apparatus comprising:

means for causing the channel coding process to 1 effectively carried out on raw data in which the relative importance levels of one or more of the raw data bits differs to the relative importance level t.I.C-the bits would have had under the defined data generation process, so as to alter the amount of data in each coded data frame to be transmitted as compared C the amount of data that would be present in each framE if the changing of the effective relative importance levels had not taken place.

The effective change in the importance levels of the bits can be selected as desired. The changing aj14 reclassification will primarily be governed by two factors.

The first is to ensure that the coded bits fit 11L C the transmission data structure. One way to do this might be to reclassify the raw data bits such that ti, number of bits at each level of relative importance,. 1 similar to the number of bits at each level of relative - is - importance which would be provided by a data generation process defined for the transmission data structure. A straightforward approach to reduce the amount of data might be simply to reduce the number of bits at the importance level which receives the most error protection, i.e. for which the largest number of additional bits are generated by the channel coding.

The second factor that should be borne in mind when reclassifying the importance levels of the data bits is to try to avoid too much degradation of the quality of the raw data (e.g. speech data) being transmitted. Thus, for example, the arrangement is preferably such that any quality critical parameters are less detrimentally affected by the importance level is reclassification than other parameters in the raw data. For example, it could be arranged that highest importance level bits do not have their importance levels re-classified.

Where speech data is being transmitted, the fundamental frequency of the speech can be an important quality affecting parameter, as it may not only define the fundamental frequency but also the structure of the rest of the speech frame, and must in that case be error-free to allow the frame to be decoded correctly.

Such a parameter should therefore preferably still be heavily error protected even after the importance level recla ssification. On the other hand, some parameters in the raw data, such as certain synchronisation bits, may be redundant or unnecessary, and so can be reclassified to lower levels of importance (error protection) without any significant effect on data quality.

An alternative way of altering, e.g. reducing, the size of (number of bits in) the coded data frames that the Applicants have recognised, would be to modify the channel coding process so as to alter, e.g. reduce, the number of coded output bits produced as compared to when using the unmodified channel coding process. Channel 16 coding processes typically generateadditional "redundant" data bits which are included in the coded 1 data. Therefore by adjusting the channel coding to 1.

alter, e.g. reduce, the number of additional data bits generated, i.e. the extent to which the channel codiiig process expands the raw data, one can alter, e.g.

reduce, the size of the coded data frames generated J'r:m a given amount of raw, uncoded data.

According to a seventh aspect of the present i invention, there is provided a method of preparing dats for transmission in a digital radio system, in which system data to be transmitted is generated from an i:tipt signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, the method comprising:

generating from an input signal frames of digitall data; processing the generated frames of digital data 1.

using a channel coding process to produce a series of:] coded data frames for transmission; and selectively modifying the channel coding procesS as to reduce the amount of data in each coded data f:i.-aite to be transmitted as compared to the amount of data t-hit would be present in each frame when using the unmodifi d process.

According to an eighth aspect of the present invention, there is provided an apparatus for prepari data for transmission in a digital radio system, in which system data to be transmitted is generated from 3n input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, the radio unit comprising:

means for generating from an input signal frailLt--t3 1..L digital data; means for processing the generated frames of digital data using a channel coding process to produc- a series of coded data frames for transmission; and means for selectively modifying the channel coding process so as to reduce the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified process.

Typically, a channel coding process will generate additional data bits at a given rate per 'raw' data bit (e.g. one additional data bit per two raw data bits: a channel coding rate of 0.5). Thus one way to modify the channel coding process to alter, e.g. reduce, the number of coded bits would be to alter, e.g. reduce, the "coding" rate. Where the raw data has bits of different levels of relative importance, different coding rates are often used for each different level of importance.

In such a case, the coding rates for one or more of the different levels of importance could be altered. one relatively straightforward way to achieve this might be to use the same coding rate for bits of two or more relative levels of importance (e.g. the same coding rate for both high and medium importance bits). This could be a new, e.g. reduced, rate, or one could, for example, simply use one, e.g. the lower, of the already defined coding rates for both levels of relative importance.

Preferably the coding process is modified such that the gross number of coded bits will fit into the defined transmission data structure, but whilst maximising the level, or retaining the same level, of error protection and/or detection (e.g. forward error correction code strength) for at least one, and preferably each, level of bit importance in the raw data.

It would, of course, be possible both to modify the channel coding process used and to re-classify importance levels of bits in the uncoded data to reduce, alter, e.g. the number of bits in the coded data for transmission, if desired. However, as this requires two levels of modification to existing processes, it may be 18 preferable to use only one or other technique where q13.t can acceptably be done.

Both of the above techniques, i.e. effectively YE classifying the relative importance levels of the 1 generated data and modifying the channel coding, havE. been found by the Applicants to be particularly applicable to an arrangement where it is desired to channel code APCO generated speech frames using TETR forward error correction protocols and to transmit ttE coded frames over a TETRA air-interface. For examplg,, by using a 2/3 coding rate for the TETRA forward errci correction coding protocols on both the high and medi-cm importance bits produced by the APCO speech coding, it is possible to fit in such an arrangement three raw tF 0 speech frames into one TETRA timeslot.

The above examples have talked particularly aboLt reducing the amount of coded data for transmission f:rcri a given amount of raw data. However, as discussed above, it may in some circumstances be desirable to increase the amount of coded data generated from the r w data. For example, where there is spare data capacity in the transmission data structure, additional data 1--its could be added to pad out the spare data capacity. Alternatively, the channel coding rate could be increased and/or data bits treated as having higher relative importance levels, so as to, for example, increase the error protection where there is the capacity to do so.

The above aspects of the present invention are concerned with the fitting of data into predefined transmission data (timeslot) structures. However, the Applicants have recognised that this may not be the cnly problem to be encountered when attempting to transmit over an air-interface using data generating and channel coding processes which are not specifically designed f:)r that air- interface's transmission data structure.

In particular, many radio communications system particularly mobile radio communications systems, support, as is known in the art, the "stealing" of user data to allow other data such as control signalling to be transmitted in a stream of user data. In such an arrangement user data is replaced by the other data and the receiver usually attempts to interpolate the missing (stolen) user data from the data that it does receive.

Such stealing techniques are usually arranged to steal' an entire timeslot or part of a timeslot which matches exactly a number of complete coded data frames, as this avoids the stealing leaving partial coded data frames which would then typically result in the loss of all the data in the partial frames from the transmission as well (such that more user data than might strictly be is necessary to provide space for the other, "stealing", data to be transmitted, is lost).

This can be relatively easily achieved where the raw data frame generation process, channel coding process, transmission data (timeslot) structure, and stealing process are defined for use with each other, such as would be the case for a given radio Standard (such as TETRA). However, where coded data frames generated using a raw data frame generation process and channel coding process are to be transmitted using a transmission data structure (i.e. air-interface protocol) for which the data generation and channel coding processes used are not optimised, it may be much more unlikely that the stealing process will steal an exact number of the coded data frames.

For example, as discussed above three APCO speech frames will fit or can be fitted into a single TETRA transmission timeslot. Thus in such an arrangement, if an entire timeslot is to be stolen from the TETRA signal, the stealing process can proceed essentially as normal, as three entire APCO coded speech frames would be stolen and no partial coded speech frames would be left over. However, TETRA also supports the stealing of a half timeslot (Ihalfslot stealing'). Stealing half a TETRA timeslot in such an arrangement would require the space occupied by one and a half of the APCO coded speech frames. Thus the TETRA half-slot stealing arrangement would typically result in the loss of twe f the three APCO coded speech frames to be carried in the timeslot which is being stolen from, even though the stealing only requires the timeslot space occupied b,i one and a half APCO frames.

The Applicants have therefore developed a numbe-c f techniques for improving user data "stealing" operat:icris in radio communications systems where the size of the stolen portion of the transmitted signal (e.g. timeslct.

or part timeslot) does not match exactly a whole numtei is of complete coded data frames.

Thus, according to a ninth aspect of the present invention, there is provided a method of transmitting data in a radio communications system, which system supports stealing of data to allow the transmission cf other data such as control signalling in a stream of data to be transmitted, and in which system data to he transmitted is arranged in a succession of discrete data frames, and is transmitted in a succession of discrete timeslots, each timeslot usually containing three or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the method comprising:

when stealing of a portion of a timeslot is to take place, selecting from the data frames which would ha-sk been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot in such a manner that immediately adjacent --'ata frames in the data to be transmitted are not both stolen.

According to a tenth aspect of the present invention, there is provided an apparatus for transmitting data in a radio communications system, which system supports stealing of data to allow the transmission of other data such as control signalling in a stream of data to be transmitted, and in which system data to be transmitted is arranged in a succession of discrete data frames, and is transmitted in a succession of discrete timeslots, each timeslot usually containing three or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the apparatus comprising:

means for, when stealing of a portion of a timeslot is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot in such a manner that immediately adjacent is data frames in the data to be transmitted are not both stolen.

In these aspects of the present invention, where three or more data frames are included in a given transmission timeslot and a portion of the timeslot is to be stolen, the data frame or frames that are still transmitted in the remaining, lunstolen' part of the timeslot are selected such that immediately consecutive data frames are not both stolen (i.e. immediately consecutive data frames are not both omitted from the transmission due to the stealing). This produces relatively smaller gaps of missing data in the transmitted data, and although there may be more of these smaller gaps, the Applicants believe that such smaller <gaps of missing data may result in, for example, less distortion of speech and thus allow more accurate interpolation of speech, than a single but much larger gap. Thus it is believed that these aspects of the invention should help to increase data recoverability when such stealing is occurring.

This aspect of the present invention is applicable to the above example of a TETRA/APCO system where three APCO coded data frames are inserted in each TETRA 22 timeslot and half-slot stealing is occurring over thE TETRA air-interface. In such an arrangement, the ce:rtl-e frame (in time) of the three APCO frames that were tc be transmitted in the TETRA timeslot would be selected -::E the data frame to transmit in its entirety in the remaining, unstolen TETRA half-slot.

According to an eleventh aspect of the present invention, there is provided a method of transmittinc data in a radio communications system, which system 1 supports stealing of data to allow the transmission cf other data such as control signalling in a stream of data to be transmitted, and in which system data to tE transmitted is arranged in a succession of discrete c;s-1-a frames, and is transmitted in a succession of discretE is timeslots, each timeslot usually containing two or mcie of the data frames, and wherein stealing of a portior f a timeslot can take place, the method comprising:

when stealing of a portion of a timeslot is to tF-ke place, selecting from the data frames which would ha--sE been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot on the basis of an assessment of the relatiSE importance levels of the data frames that were to be transmitted.

According to a twelfth aspect of the present invention, there is provided an apparatus for transmitting data in a radio communications system, which system supports stealing of data to allow the transmission of other data such as control signalling In a stream of data to be transmitted, and in which syst.Em data to be transmitted is arranged in a succession of discrete data frames, and is transmitted in a successiOn of discrete timeslots, each timeslot usually containiiS two or more of the data frames, and wherein stealing 4E a portion of a timeslot can take place, the apparatu comprising:

means for, when stealing of a portion of a tiMet>.Dt is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot on the basis of an assessment of the relative importance levels of the data frames that were to be transmitted.

In these aspects of the invention, when stealing of part of a timeslot takes place, a selection is again made as to which data frame to transmit in the remaining portion of the timeslot, but the selection is based on the relative importance to the data being transmitted of the data frames that were to be transmitted. The Applicants believe that such an "intelligent" selection of the data frame to still transmit would again offer is advantages in terms of compensation for and interpolation of the missing data over a more arbitrary frame selecting process.

The frame or frames still transmitted should be that frame or those frames assessed to have the highest relative importance according to the assessment criteria used. The importance assessment could be carried out as desired. It could, for example, be based on an assessment of how useful each frame might be for the interpolation process or for preserving data intelligibility. Where the data is speech frames, the importance assessment could be based on any suitable speech importance assessment techniques, such as the speech energy of each frame, or a more sophisticated measure of speech importance.

In these arrangements, an indication of which of the possible data frames has actually been transmitted is preferably also transmitted by the receiver. This could be useful where a selection based on frame importance is being made since in such an arrangement it will not necessarily always be the frame or frames in the same relative timing position which is still transmitted at each stealing event.

24 In both of the above stealing arrangements, it is possible that there will be spare unused data capaci-t, in the remaining unstolen part of the timeslot once -t. data frame or frames selected for transmission are inserted therein. This would be the case, for examp.1% in the above discussed TETRA/APCO system where a sincl complete APCO frame is to be transmitted in the remaining TETRA half-slot. In such a case, the sparE data capacity remaining in the timeslot is preferabl-y used to transmit additional information that will he.Irto interpolate and/or compensate for the missing (stolen) data. This additional information could, fcx example, be information which will aid interpolation across the data gaps and/or enhance intelligibility, such as information relating to or from the missing cata (e.g. speech) frames. It could also or instead be information indicating which of the possible data franes has actually been transmitted.

The above stealing arrangements ensure that one cr more entire data frames are still transmitted, but will typically mean that other data frames are omitted frcu the transmission in their entirety. The Applicants believe that in some circumstances it may be desirablE to try to transmit at least some data from plural data frames, or data from as many data frames as is possikle, even if the data frames are not all transmitted in t.Eir original, complete state.

Thus, according to a thirteenth aspect of the present invention, there is provided a method of transmitting data in radio communications system, in. which system data to be transmitted is arranged in a. succession of data frames, which frames are transmittEd in timeslots, each timeslot usually containing two o2 more data frames, and in which system stealing of pat of timeslot can occur to allow other data such as control signalling to be transmitted in a stream of dEta frames, the method comprising:

when it is determined that stealing of part of a timeslot is to take place, generating at least one data frame having a reduced size from one of the data frames that was to be transmitted in the timeslot to be stolen from, so as to allow data from that frame to fit into and be transmitted in the remaining unstolen part of the timeslot.

According to a fourteenth aspect of the present invention, there is provided an apparatus for transmitting data in radio communications system, in which system data to be transmitted is arranged in a succession of data frames, which frames are transmitted in timeslots, each timeslot usually containing two or more data frames, and in which system stealing of part is of timeslot can occur to allow other data such as control signalling to be transmitted in a stream of data frames, the apparatus comprising:

means for, when it is determined that stealing of part of a timeslot is to take place, generating at least one data frame having a reduced size from one of the data frames that was to be transmitted in the timeslot to be stolen from, so as to allow data from that frame to fit into and be transmitted in the remaining unstolen part of the timeslot.

In these aspects of the present invention, when stealing is to take place, one or more data frames having a reduced size are generated from a data frame or data frames that would have been transmitted.

Generating a reduced size data frame can allow the data frame to fit into the remaining part of the timeslot and thus allow at least some of the data from the original data frame to still be transmitted, rather than the entire data frame simply being lost.

The way that the size of the data frame is reduced can be selected as desired. In one preferred arrangement, data from the original data frame is deleted to reduce the size of the frame. This could be 26 done by, for example, deleting the necessary number 'If bits from the data frame according to a systematic, predetermined process. In a preferred such arrangem4lri, where bits in the data frame have differing levels of relative importance (e.g. where the raw data is generated having differing levels of relative importance), then the bits of the lowest importance 1 level are preferably preferentially deleted first, followed by bits of the next lowest importance level and so on. A further or alternative selection criteria,,--uch as the impact of the bits on the data quality could ilSo be used to select those bits to omit in the reduced::-i,,e data frame.

An additional or alternative way to reduce the::-::L e of the data frames in a system where raw data frames are channel coded to produce coded data frames for transmission, would be to modify the raw data frames c the data frame channel coding process to produce smaller coded data frames from a given amount of raw data. '[his could be done using, for example, any of the techniqijeE for this discussed previously, such as changing the effective number of raw data bits at each level of relative importance as seen by the channel coding process, and/or by adjusting the coding process, e.g.

the forward error correction coding rate, so as to produce less coded data from a given amount of raw dat;L.

only one data frame in the original data could he reduced in size when stealing is to take place. Alternatively, if desired, corresponding smaller, reduced data frames for transmission could be generatel from two or more of the data frames originally to be transmitted in the timeslot.

The way that the data frames, including the redijcEd size data frame or frames, are arranged for transmission in the remaining part of the timeslot can be selected ELS desired. For example, one or more complete unmodified data frames could be transmitted and any remaining sr.,ace in the remaining part of the timeslot filled with one or more reduced size data frames. Alternatively only a number of reduced size data frames could be transmitted in the unstolen part of the timeslot (i.e. such that no unreduced data frames are transmitted).

Thus, for example, in the above discussed example of transmitting three APCO coded data frames in a TETRA timeslot, when half-slot stealing is to occur, the remaining TETRA half-slot could be used, for example, to transmit two reduced size data frames from the three intended for that timeslot, or to transmit one complete, unreduced data frame and one or two reduced size data frames.

Where, in the above arrangements, not all of the original data frames are to be transmitted and/or some are to be reduced and others still transmitted in their entirety, the frames to be omitted, reduced, etc, are preferably selected on the basis of an assessment of their relative importance to the data being transmitted.

This selection can be done along the lines discussed above, for example where speech frames are being transmitted, on the basis of a speech importance assessment, e.g. a speech energy assessment.

As a further refinement to the stealing process, it would be possible to modify the stealing process such that a reduced size portion of the timeslot is stolen so as to allow more room for the user data frames to be transmitted. This would provide less capacity for the stealing data, e.g. control signalling, but may be acceptable in some circumstances as it allows more user data to be transmitted. This arrangement is preferably used in conjunction with transmitting one or more reduced size data frames in the remaining unstolen part of the timeslot, but could, if sufficient space can be left by using a reduced size stolen timeslot portion, be used with unmodified data frames being transmitted in the unstolen part of the timeslot.

28 Thus according to a fifteenth aspect of the prese invention, there is provided a method of transmitting data in a radio communications system, in which syst(:,,rr data to be transmitted is generated as a succession (f discrete data frames and the data frames are transmitt in a succession of discrete timeslots, and in which system there is a predefined data stealing process whereby a portion of a timeslot of a predetermined siz is replaced with other data to allow transmission of other data such as control signalling in a stream of data frames, the method comprising:

when it is determined that such stealing should occur selectively modifying the stealing process so i3s to steal a portion of the timeslot which is smaller th the size of said predetermined stealing portion so az allow more space for the data frames that were to be transmitted to be fitted into the remaining unstolen,l I part of the timeslot when such stealing occurs.

According to a sixteenth aspect of the present invention, there is provided an apparatus for transmitting data in a radio communications system, jr which system data to be transmitted is generated as succession of discrete data frames and the data framglE are transmitted in a succession of discrete timeslotE, and in which system there is a predefined data stealir process whereby a portion of a timeslot of a predetermined size is replaced with other data to all( transmission of other data such as control signalling a stream of data frames, the apparatus comprising:

means for, when it is determined that such stea - should occur, selectively modifying the stealing proo so as to steal a portion of the timeslot which is smaller than the size of said predetermined stealing portion so as to allow more space for the data fraMeffl) that were to be transmitted to be fitted into the remaining unstolen part of the timeslot when such stealing occurs.

In these aspects of the present invent ion, the stealing process is modified so as to steal a smaller portion of the timeslot. This allows more space for the data frames that were to be transmitted to be fitted into the remaining unstolen part of the timeslot.

This technique would preferably be used where it is recognised that the data frames being transmittedare not of the normal size for the radio system (i.e. not of a size that is optimised to the stealing process(e.g. of the size that would match appropriately the predetermined portion of the timeslot to be stolen)), such that a modification to the stealing process may be desirable. This situation could be recognised by, for example, determining whether the data frames to be is transmitted are of a predetermined size, and if it is determined that the data frames are not of said predetermined size, modifying the stealing process.

Thus, for example, in the above discussed example of transmitting three APCOcoded frames in a TETRA timeslot, the TETRA half-slot stealing process could be modified so as to steal less than half a slot (e.g. to steal one third of a slot so as to allow room for two of the three APCO frames in the remaining unstolen part of the TETRA timeslot), preferably upon recognition that the data frames did not match the TETRA data frame size.

In these arrangements an indication that a reduced size timeslot portion has been stolen is preferably transmitted to the receiver to allow the receiver to recognise that the stealing process is not in accordance with the usual, defined stealing process. This indication could comprise, for example, a stealing portion length indicating word, or a marker, e.g. escape sequence, at the end of the stolen portion of data. In an APCO/TETRA system operating in this way, each stolen timeslot portion could carry an indication as to whether the next timeslot portion is stolen, as is the case for normal TETRA stealing (in which the first half-slot indicates whether the second half-slot is also stole-r, i.e. full-slot stealing is occurring).

In all of the above arrangements, where two or ticre data frames (whether both reduced in size or otherwiEe) are to be transmitted in the remaining portion of a partially stolen timeslot, then where a data interleaving technique is used for the data transmission, the data transmission preferably uses modified interleaving structure across the entire remaining timeslot portion, rather than maintaining (.ata frame boundaries as would be more usual, so as to increase the channel coding (e.g. forward error correction) immunity to burst errors.

Encryption is a further factor that the Applicant have recognised needs to be considered when combining data generation and channel coding processes defined f.)r different data transmission structures, or when attempting to transmit data generated and channel codei according to processes defined for one form of transmission data structure using another different transmission data structure.

As is known in the art, many encryption processes even if they do not generate additional data from the original unencrypted data, still require the transmission of encryption signalling, such as encryption algorithm and key identity data, encrypticn synchronisation information, etc, to the receiver tol allow successful decrypting of the transmitted data. Where this data is normally sent in the stream of encrypted data, it becomes additional data which must be fitted into the timeslot structure being used. one %a5, to achieve this would be to use the various data reduction and apportionment techniques already discussed to reduce the spaceoccupied as a result of the user data (e.g. the size of the (coded) data frames) to be transmitted to allow the encryption signalling to fitj into the timeslot structure.

Thus, according to a seventeenth aspect of the present invention, there is provided a method of transmitting data in an encrypted form in a radio communications system, in which system data can be transmitted in one of two different predefined data structures, and in which system at least one of the data structures has defined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver, and in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, the method comprising:

is when it is desired to transmit encrypted data generated and encrypted according to the processes defined for said at least one data structure over the other data structure, including said encryption control information in the data transmission, and modifying the data generating process and/or modifying the channel coding process so as to alter the amount of data in each co ded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as-defined processes.

According to an eighteenth aspect of the present invention, there is provided an apparatus for transmitting data in an encrypted form in a radio communications system, in which system data can be transmitted in one of two different predefined data structures, and in which system at least one of the data structures has defined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver, and in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process 32 prior to transmission, the apparatus comprising:

means for, when it is desired to transmit encrylted data generated and encrypted according to the proces:Ees defined for said at least one data structure over thE other data structure, including said encryption contxcl information in the data transmission, and for modify.inj the data generating process and/or modifying the charnel coding process so as to alter the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as-defined processes.

In the application of this arrangement to a TETRA/APCO system, one could always define the use o:E the APCO codec as being in partnership with DES encryption and therefore embed 4 bits per speech frame of re-synchronisation information into the (start) o:E each speech frame. The bits could optionally be classified for forward error correction coding. In this arrangement the coded bit rate could be modified to reduce the coded frame size according to one of the techniques discussed above. The four bits could be set to an easily recognisable pattern, e.g. all one's or all zerols, when there is no encryption.

Thus according to a further aspect of the present invention, there is provided a method of transmittin encrypted data frames, comprising:

allocating a predetermined number of bits of spac for encryption synchronisation information in each datEL frame; and setting the predetermined bits to a predetermined pattern when there is no encryption.

The Applicants believe that in some circumstances other techniques could be advantageous.

Thus, according to a nineteenth aspect of the present invention, there is provided a method of transmitting data in an encrypted form in a radio communications system, which system includes both a control channel for control signal transmission and one or more traffic radio channels for user data transmissions, and in which system data can be transmitted in one of two different predefined data structures and in which system at least one of the data structures has defined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver and which encryption control information is in the defined process effectively embedded in the user data to be transmitted, the method comprising:

when it is desired to transmit encrypted data generated and encrypted according to the processes defined for said at least one data structure over the other data structure, removing said embedded encryption control information from the data to be transmitted over the traffic channel and transmitting it instead over a control channel.

According to a twentieth aspect of the present invention, there is provided an apparatus for transmitting data in an encrypted form in a radio communications system, which system includes both a control channel for control signal transmission and one 2S or more traffic radio channels for user data transmissions, and in which system data can be transmitted in one of two different predefined data structures and in which system at least one of the data structures has defined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver and which encryption control information is in the defined process effectively embedded in the user data to be transmitted, the apparatus comprising:

means for, when it is desired to transmit encrypted data generated and encrypted according to the processes 34 defined for said at least one data structure over tli& other data structure, removing said embedded encrypt.-.'I control information from the data to be transmitted _r the traffic channel and transmitting it instead over control channel.

In these aspects of the present invention, when Z different timeslot structure is to be used for encryptUd data transmission, the additional encryption control. (synchronisation) information which is part of the dELM3.

to be transmitted is removed from the data to be transmitted on the traffic channel and instead sent cr a control channel of the radio system (as is known in ti art many radio systems use a channel arrangement whicl includes both control channels for control signallin:

and traffic channels for the transmission of user datz e.g. speech data). This means that capacity on the control channel, rather than traffic channel capacit, is used for the encryption control information transmission, thereby helping the fitting of the remaining user data onto the traffic channel. Also, hy transmitting the encryption control information on t1E control channel, the receiver can more readily recog.r I e it as being control signalling, whereas it may need specially marking to distinguish it from user data if it were transmitted on a traffic channel.

The encryption information can conveniently be Ent on the control channel as part of the call set-up I signalling, where appropriate.

This arrangement is applicable to an APCO/TETRA system where APCO coded data frames are to be end-tO-E d encrypted according to the APCO standard but then transmitted over a TETRA air- interface. APCO uses DEIc encryption (Data Encryption Standard from the National Institute of Standard and Technology (NIST) - see, fcx example, NIST, Data Encryption Standard, FIPS Publication 46-2, and ANSI, Data Encryption Algorithn, ANSI X3.92 - 1981), which is applied as a stream cipt E The encryption cipher is initially synchronised with a key identity, algorithm identity and a message indicator, which initial encryption synchronisation information is included in the data stream to be transmitted to the receiver.

In accordance with the above aspects of the present invention, when it is desired to transmit such APCO encrypted data over a TETRA air-interface, the initial encryption synchronisation information would be stripped out of the data stream and transmitted as part of the call set-up signalling over a TETRA control channel, rather than as part of the data transmitted over the traffic channel. This avoids using traffic data capacity for the encryption synchronisation information transmission, and also allows TETRA to recognise that the data is control signalling without the need to mark it to distinguish it from e.g. speech data.

Many encryption arrangements also require the periodic transmission of encryption information during the encrypted data transmission, to permit, for example, so-called "late entry" to the transmission and to allow a receiver to remain synchronised to the encrypted signal. The APCO DES system, for example, requires the message indicator to be repeated every 18 data frames to facilitate late entry.

In the present invention, such late entry and periodic encryption signalling is also preferably transmitted on a control channel, preferably together with the encryption synchronisation information, at appropriate intervals, rather than including it in the traffic channel transmission.

An alternative way to send the late entry signalling in the present invention would be to send it by using a traffic channel data stealing process. This is believed to be preferable to allowing the late entry signalling to remain in the user data to be transmitted in the form that it is generated, since transmitting it 36 by a stealing process can allow the encryption 1 i synchronisation data to be recognised as not being, (A speech data, without the need to transmit additional data or to modify existing processes. Where such traffic channel stealing techniques are used to transir the late entry signalling, the frequency of the latel entry signalling is preferably reduced, where this cali acceptably be done (e.g. the Applicants have recogni:-E that this would be possible in a TETRA/APCO system, aj correct synchronisation at the receiver can be i maintained over relatively longer time periods wherell receive circuits are correctly clocked).

A further factor that the Applicants have recognised needs to be considered where one is makin.- is encrypted transmissions in a system that can operate accordance with the present invention, is that where deliberately reduced or truncated data frame is to bE sent, for example as in the case of the data stealinc mechanisms discussed above, the missing data bits fi-c the frame or frames will typically have to be allowec for in the encryption process so as to maintain cori.E i encryption synchronisation. i In one preferred embodiment, this is achieved b removing the necessary number of data bits (to reduce the frame size) after encryption, with a correspondin number of bits being reinserted into the data in the receiver prior to decryption (or prior to passing th( data into another part of the system, e.g. if the ra( unit is acting as a gateway), to maintain correct decryption synchronisation (i.e. to compensate for t.lli "missing bits,'). This provides a relatively straightforward way of maintaining encryption synchronisation while transmitting a reduced number 6 data bits. The bits added (reinserted) by the receiv are preferably added to the reduced size data frame according to a predetermined pattern, such as all be! zero or other values selected so as to have a minimiAlr impact upon the data, so as not to affect too adversely any subsequent data processing steps carried out on the decrypted data, such as speech decoding. The predetermined pattern of bits added to the partial data frame to replace the missing bits can be fixed, or vary in a predetermined manner (e.g. change frame by frame), if desired.

As will be appreciated by those skilled in the art, all of the above aspects and arrangements of the present invention can be used singly, and/or in appropriate combinations, in a radio unit and a radio system. The invention is particularly, although not exclusively, applicable to a combined TETRA/APCO system and in - particular to permitting transmission (in particular of speech) over the TETRA air-interface by a radio unit which uses APCO speech coding with APCO or TETRA channel coding. Such a radio unit could include any one or more of the above aspects and arrangements of the present invention in appropriate combinations.

Thus, according to a further aspect of the present invention, there is provided a method of transmitting speech frames generated in accordance with the APCO Project 25 radio Standard speech coding protocols over a TETRA radio Standard air-interface, comprising:

generating from an input speech signal speech frames for transmission in accordance with the speech frame generation process defined for the APCO Project 25 radio Standard; channel coding the generated speech frames using the channel coding processes defined for the APCO Project 25 radio Standard; and transmitting the coded speech frames over a TETRA air-interface by inserting three coded speech frames into each TETRA timeslot.

According to a yet further aspect of the present invention, there is provided a method of transmitting speech frames generated from a speech signal in 38 accordance with the APCO Project 25 radio Standard a TETRA radio Standard air-interface, the method comprising:

generating from a speech signal speech frames i:- accordance with the speech frame generation process defined for the APCO Project 25 radio Standard; coding the generated speech data frames using tl-,e forward error correction process defined for the TETRA radio Standard, but causing the forward error correctiDn processes to be effectively carried out on data frames in which the relative importance levels of at least scte of the data bits differ from the relative importancel levels that they would be given by an unmodified APCC, Project 25 speech frame data generation process, so as to reduce the amount of data in each coded data frame o be transmitted as compared to the amount of data that would be present in each frame if the changing of the relative importance levels had not taken place; and transmitting the coded data frames over a TETRA air-interface.

According to a yet further aspect of the present invention, there is provided a method of transmitting speech frames generated in accordance with the APCO Project 25 radio Standard over a TETRA radio Standard air-interface, the method comprising:

generating from a speech signal frames of speech data in accordance with the APCO Project 25 radio Standard; channel coding the generated speech frames using the channel coding processes defined for the TETRA radLo Standard, but modified so as to reduce the coding rate for bits of at least one level of relative importance; and transmitting the so-coded frames over a radio channel structured in accordance with the TETRA radia Standard.

According to another aspect of the present invention, there is provided a method of transmitting speech data frames generated in accordance with the APCO Project 25 radio Standard over a TETRA radio Standard air-interface, the method comprising:

transmitting data corresponding to three APCO generated speech frames in each TETRA timeslot; and when TETRA half-slot stealing is to occur, transmitting in the remaining half-slot of the timeslot which is to be stolen from, data corresponding to the centre APCO speech frame of the three speech frames which were to be transmitted in the timeslot.

According to another aspect of the present invention, there is provided a method of transmitting speech data frames generated in accordance with the APCO is Project 25 radio Standard over a TETRA radio Standard air-interface, the method comprising:

transmitting data corresponding to three APCO generated speech frames in each TETRA timeslot; and when TETRA half-slot stealing is to occur, transmitting in the remaining half-slot of the timeslot which is to be stolen from, one of the three APCO speech frames which were to have been transmitted in the timeslot on the basis of an assessment of the relative speech importance levels of the speech frames that were to be transmitted in the timeslot.

transmitting data corresponding to three APCO generated speech frames in each TETRA timeslot; and when TETRA half-slot stealing is to occur, generating from one or more of the APCO speech frames that were to have been transmitted in the timeslot a speech frame having a reduced size and transmitting that reduced size speech frame in the remaining unstolen part of the timeslot.

According to yet another aspect of the present invention, there is provided a method of transmittina_ speech frames generated according to the APCO Projeci_- 15 radio Standard over a TETRA radio Standard airinterface, the method comprising:

when such frames are being transmitted over a THT air-interface and TETRA half-slot stealing is to take place, modifying the TETRA half-slot stealing process So as to steal a smaller portion of the timeslot than half of the timeslot, and, preferably, so as to steal one third of the timeslot.

According to a further aspect of the present I invention, there is provided a method of transmitting encrypted speech frames generated and encrypted in accordance with the APCO Project 25 radio Standard o-ver a TETRA radio Standard air-interface, the method comprising:

allocating 4 bits of space for encryption I synchronisation information in each speech frame; and setting the four bits to a predetermined patterti when there is no encryption.

According to another aspect of the present invention, there is provided a method of transmitting encrypted speech frames generated and encrypted in accordance with the APCO Project 25 radio Standard o-ver a TETRA radio Standard air-interface, the method comprising:

when it is desired to transmit such encrypted speech frames over the TETRA air-interface, transmittilg the speech data over a TETRA traffic channel and transmitting the initial encryption synchronisation, information over a TETRA control channel.

In this aspect of the invention, any periodic encryption information to be sent during the encrypted transmission is also preferably sent over a TETRA control channel or by means of TETRA stealing mechanisms.

The invention also extends to apparatus for carrying out, and radio units capable of carrying out, any or all of the above methods. Similarly, all of the above aspects of the invention can include any or all of the preferred features of the invention discussed.

Although the invention is applicable in many of its aspects, where appropriate, to transmission of data in radio systems in general, it is particularly applicable to the transmission of user traffic, such as in particular speech data and frames.

The methods in accordance with the present invention may be implemented at least partially using software e.g. computer programs. It will thus be seen is that when viewed from further aspects the present invention provides computer software specifically adapted to carry out the methods hereinabove described when installed on data processing means, and a computer program element comprising computer software code portions for performing the methods hereinabove described when the program element is run on a data processing means. The invention also extends to a computer software carrier comprising such software which when used to operate a radio unit or system comprising a digital computer causes in conjunction with said computer said unit or system to carry out the steps of the method of the present invention. Such a computer software carrier could be a physical storage medium such as a ROM chip, CD ROM or disk, or could be a signal such as an electronic signal over wires, an optical signal or a radio signal such as to a satellite or the like.

It will further be appreciated that not all steps of the method of the invention need be carried out by computer software and thus from a further broad aspect the present invention provides computer software and such software installed on a computer software carrier for carrying out at least one of the steps of the - 42 methods set out hereinabove.

A number of preferred embodiments of the present invention will now be described by way of example on-Ly and with reference to the accompanying figure, Figure which shows an arrangement for half-slot stealing ov(:,TETRA air-interface in a combined APCO/TETRA system _[Ti accordance with one aspect of the present invention.

The present invention will be described with reference to the operation of a radio system which allows data transmission over an APCO Project 25 radi Standard air-interface ("APCO") and a TETRA radio Standard normal (i.e. not scaled in any way) air interface. However, as will be appreciated, the pre,,,-,[ invention is not exclusively limited to such an arrangement and will be applicable to the combination other radio Standards and systems.

The preferred embodiments of the invention will also be described with reference to the transmission 6 speech in such an APCO/TETRA system, since in most rad systems speech transmission will be of particular importance. However, as will be appreciated by those skilled in the art, the invention is applicable to transmissions other than of speech.

Both TETRA and APCO have defined transmission dat structures which comprise plural discrete timeslots of defined size. A given radio transmission will normal consist of periodic timeslots of a constant length an( at a constant rate which can therefore carry a given, fixed, period of speech in a given time period.

In both systems, digital data for transmission fr an input speech signal is generated in the form of discrete speech frames by a speech coder/decoder (codec). The speech frames are generated at a constan rate and have a constant size, such that a given numbe of speech frames will then carry a particular period b speech. The raw speech frames are then channel coded using, inter alia, forward error correction coding.

both systems the speech coding, and related channel coding are optimised to ensure that the best fit into the defined transmission data timeslot structure is achieved.

However, the two standards generate speech frames of differing sizes and have different channel coding processes. The various speech-related parameters of the two systems are summarised in the following table:

Parameter description TETRA APCO-25

Gross bit rate 7.2 kbps 7.2 kbps 216 bits per speech frame 144 bits per speech frame 2 speech frames/timeslot 432 bits Speech frame duration 30 ms 20 ms Uncoded speech bit-rate 4.567 kbps (173 bits/frame) 4.4 bps (88 bits/frame) FIigh importance bits/frame 30 48 Medium importance 56 33 bits/frame Low importance bits/frame 51 7 Coding rate for Mgh R7-8/18 (info bitstcoded bits) R=0.53 (23,12) Importance bits R=8/17 for half slots Coding rate for Medium R=2/3 R=0.73 (15, 11) Importance bits Coding rate for Low R7-1 (Unprotected) R-- 1 (Unprotected) Importance bits Cyclic redundancy check 8 + 4 tail bits (1/2 slot: 4+4) None bits (coded as Mgh Importance) To illustrate the present invention, an arrangement where it is desired to transmit speech encoded using the APCO speech coding arrangement using the normal TETRA transmission data structure (i.e. over a standard TETRA air-interface) will be considered.

44 It can be seen from the above table that a single TETRA timeslot is intended to transmit 60 milliseconcl,, of speech, and that that period corresponds to the speech encoded in 3 APCO speech frames. Thus a commoil multiple unit of speech between the two standards is a speech period of 60 ms. This corresponds to two TETRi speech frames which is 274 bits of speech as output t, the speech codec to be transmitted in each timeslot. Three APCO speech frames would contain 264 bits, so in principle, the three APCO speech frames could fit int.0 single TETRA timeslot.

However, there then arises the question of having to channel code the raw speech frames before they are transmitted. Applying the TETRA channel coding (forwa d is error correction) to an APCO speech frame gives per coded speech frame:

(48 x 18/8) + (33 x 3/2) + 7 165 bits j when rounded up. This gives 495 (i.e. 3 x 165) bits D r ms of speech, i.e. per TETRA timeslot. However, e3. h TETRA timeslot only has the capacity for 432 bits. T-i s it is not directly possible to encode the raw APCO speech frames using unmodified TETRA channel coding, 3Ad then transmit them in the TETRA timeslot structure.

However, the Applicants have recognised that the APCO channel coding, which uses two basic types of coding for voice frames, Golay (23, 12) and Hamming (15, 11), which block codes are applied to the most important and medium importance bits, in fact produce total of 432 bits for the three APCO speech frames (bi s per frame = (48 x 23/15) + (33 x 15/11) + 7 = 144).

Therefore the APCO data generating and channel coding processes generate 432 bits of data for transmission per 60 ms of speech, which number of bit will fit into a single TETRA timeslot which is also intended to carry 60 ms of speech. Thus a first preferred embodiment of the present invention for transmitting APCO generated speech frames over a TETRA air-interface is to both generate the raw speech frames and code them in accordance with the APCO Standard, and to then place three APCO coded speech frames into each TETRA timeslot. (In TETRA the forward error correction coded speech is subsequently re-ordered, interleaved and scrambled, but as this results in no change to the number of bits it can be ignored for the purposes of fitting APCO coded speech frames into the TETRA timeslot structure.) one potential disadvantage that the Applicants have recognised with the above arrangement which uses the APCO speech frame generation and channel coding systems is is that new channel coding (forward error correction) encoding and decoding schemes would have to be employed within TETRA terminals designed to operate in this system, which could be disadvantageous. Such an arrangement may also no longer perform as required by the, e.g. TETRA radio Standard, and so may not be TETRA 'compliant, and/or require extensive testing to adapt it to meet the standard. It may therefore in'som( circumstances be preferable to use the APCO speech frame generation process, but to use TETRA channel coding processes on those speech frames. However, if TETRA forward error correction is applied directly to an APCO single speech frame, that yields:

(48 x 18/8) + (33 x 3/2) + 7 = 165 bits when rounded up. Three such speech frames (i.e. 60 ms speech) would therefore result in 495 bits, which is greater than the capacity of single TETRA timeslot. It is not therefore possible to apply directly the TETRA forward error correction coding to APCO raw speech frames.

In alternative preferred embodiments of the present 46 invention therefore, the Applicants propose to modify aspects of the speech coding process so as to produce coded frames using basic TETRA channel coding processe: on APCO generated speech frames, but which contain les- data bits and so will fit into the TETRA timeslot structure.

In a first preferred embodiment of this arrangement, the reduction in the amount of coded data generated from a given raw APCO speech frame is achiev:d by effectively reclassifying the importance levels of the output data bits from the APCO speech codec at lea:t as regards their treatment by the TETRA forward error correction coding, while still using the TETRA forwar error correction codes in an otherwise un-modified manner. In this embodiment the bits output in the APC-. speech frames are effectively reassigned amongst the different relative importance classes so as to change the number of bits at each level of relative importan so as toreduce the size of the frames after the chann 1 coding is applied. The reclassification of the relative importance levels of the bits should be Such that the number of coded bits generated per raw speech frame is reduced sufficiently that the output of thre error coded APCO speech frames will fit, preferably exactly, into a single TETRA timeslot.

The way that the relative importance levels of th. bits are reclassified can be selected as desired. One suitable reclassification would be a reclassification which results in an almost identical number of bits per relative importance class as for a standard TETRA sp(:,.e:!h encoding arrangement. The reclassification of the relative importance levels of the bits in the raw AP:0 speech frame should be done so as to try to minimise ariy impact on speech quality. Thus, for example, the fundamental frequency, which is a critical parameter, should still be heavily error protected, as it not onl. defines the frequency, but also the structure of the 47 - rest of the speech frame and must be error free in order to decode the frame correctly. On the other hand, the APCO sync bit is not needed in TETRA as TETRA time can be used instead for maintenance of synchronisation and so that bit can be reclassified to a lower importance level or be omitted, if desired.

It is believed that this arrangement may be a particularly convenient arrangement, since it allows standard TETRA forward error correction coding to be used and there is only some re-classification of the importance levels of the raw speech bits, which should involve relatively little change to existing equipment and processors.

An alternative technique for reducing the size of a is coded data frame from a given APCO raw speech frame is to modify the TETRA forward error correction schemes to reduce the coding "rate", i.e. such that less coded data is generated for a given raw data bit. Again the rate should be appropriately adjusted such that three raw APCO speech frames will fit into a single TETRA timeslot.

Thus taking R2 as being the rate of the channel coding applied to the most important bits, and R1 as the rate applied to the medium importance bits (with no coding being required for the least important bits), then for three APCO speech frames the total number of coded bits will be:

[l/R2 x ((48 x 3) + 12)l + [l/R1 x (33 x 3)l + [7 X 31 and this must be less than or, preferably, equal to, 432 bits to fit into a single TETRA timeslot when R1 and R2 are chosen. (The 12 bits added to the most important bits is for the 8 bit cyclic redundancy check and 4 tail bits needed for the TETRA forward error correction decoder.) If the standard 2/3 TETRA forward error correction - 48 coding rate is applied to the medium importance bits each APCO speech frame (i.e. R1 = 2/3), then the numb of bits needed per TETRA timeslot for the coded medii-tlrL importance bits is:

(3 x 33) x 3/2 149 bits when rounded up. The least important APCO speech bit. are not channel coded and so require (3 x 7) = 21 bit13 capacity per TETRA timeslot.

This leaves 432 - 149 - 21 = 262 is spare bits per TETRA timeslot for assignment to the mD important bits in the three APCO frames to be containe in a timeslot. A suitable forward error correction coding rate, R2, for the highest importance bits in L arrangement to use this available capacity is therefoi H48 x 3) + 12)/262 = 0.6.

Thus one possibility would be for Ri z 0.67 and R2 = 0.6. In that arrangement the protection applied the most important bits would be lower than is normal used in APCO, but the medium importance bits would be better protected than they normally are in an APCO system.

A particularly preferred embodiment of this arrangement is to code both the high and the medium importance bits at the same rate of 2/3 using the TE11 forward error correction coding. This would result li H48 + 33) x 3/2) + 7 = 129 bits per APCO speech fram( when rounded up, i.e. 387 bits per TETRA timeslot. 1 arrangement would use less capacity than is available 49 a TETRA timeslot, and the remaining capacity could be ignored or used for further error detection e.g. by use of a cyclic redundancy check. Alternatively, the coding rate for the high and medium importance bits could be increased from 2/3 to 81/137 to fill completely the available timeslot capacity, if desired.

The TETRA coding rate can be varied by modifying the "puncturing" scheme used. (As is known in the art, in TETRA the convolutional coding is produced from a mother code of rate 1/3 which is then "punctured" (i.e.

bits are removed) to produce the lower rate code (11RCPC11 code) actually used. In the receiver, the punctured bits are replaced by zeros into appropriate locations in the received data stream.) is The above arrangements permit the combination of the speech codec defined for the APCO radio standard with TETRA forward error correction coding schemes such that data corresponding to three APCO speech frames will fit into a single TETRA speech traffic timeslot. This is achieved by effectively reducing the size of the coded data frames produced from the APCO raw speech frames, either by re-classification of the importance of the output data bits from the APCO speech codec such that even when using the TETRA forward error correction codes in an otherwise un-modified manner, the number of coded output bits generated is reduced, or by the modification of the TETRA forward error correction coding streams to reduce the "coding rate" so that the gross number of coded bits generated from the three raw APCO speech frames is reduced.

The above embodiments of the present invention transmit three APCO speech frames in each TETRA timeslot. This arrangement will operate satisfactorily during normal, continuous user data transmission.

However, TETRA supports the "stealing" of user data, to, for example, permit the transmission of control signalling in a stream of user data. When such stealing so occurs, the timeslot capacity is lost to signalling a c the missing user data (e.g. speech) must be llrecovere 1 at the receiver by some mechanism. Such user data stealing will also need to be taken account of in a combined APCO/TETRA system.

TETRA supports both full timeslot and half-timeE;-.cstealing. When fullslot stealing occurs in the arrangements of the present invention where three APC0 speech frames are transmitted in each TETRA timeslot, then the situation would be the same as for standard TETRA, i.e. 60 ms of speech would be lost, and therefoie no further mitigating action is required.

However, half-slot stealing is a different mattEl- As discussed above, in standard TETRA, half a ti-TI slot carries 216 bits which corresponds to 30 ms of speech, which is lost when stealing occurs. However, the TETRA/APCO system of the present embodiment where three APCO speech frames are carried in one TETRA timeslot, the loss of half a TETRA timeslot to stealing would in an unmodified TETRA stealing arrangement causf the loss of two of the three APCO speech frames (i.e., 40 ms of speech) since it would not normally be expec-re for a partially complete speech frame to be of any use (as a complete frame is usually needed for decoding)..

The two APCO speech frames lost would be the first two of the three APCO speech frames, i.e. only the third frame is transmitted, since it is the first half-slot which is stolen.

In a first preferred stealing arrangement of the present invention, where three APCO speech frames are transmitted in each TETRA timeslot, upon recognition D half-slot stealing over the TETRA air- interface, the second (centre) APCO speech frame (in time) rather tha the third (last) frame of the three frames which were o be transmitted in the TETRA timeslot is transmitted. This provides two breaks of 20 ms in the transmitted speech data, which it is believed would be easier to interpolate compared to a single 40 ms break in the speech data. Figure 1 illustrates this arrangement.

(In Figure 1, the speech frames from the slot previous to the slot to be stolen from are labelled, in increasing time order, A01, A02, A03, the speech frames in the slot to be stolen from are then All, A12, A13, and the speech frames in the next slot are in time order, A31, A32, A33.) As can be seen in Figure 1, when stealing occurs, the centre speech frame, A12, which was to be transmitted in the slot, is transmitted in the remaining unstolen TETRA half-slot.

In this arrangement there are some spare bits in the remaining TETRA half timeslot (since a single coded APCO speech frame will not fill the entire half-slot).

is The spare bits are preferably used to transmit partial information about the missing speech frames to help the interpolation process and enhance intelligibility.

In an alternative arrangement, the one of the three speech frames to be transmitted could be selected based on an assessment of the relative speech "importance" of the respective speech frames, and/or how useful each frame might be for the interpolation process or for preserving intelligibility. The "speech importance" evaluation could be based on for example the speech energy in each frame, or a more sophisticated measure of speech importance. It is believed that this arrangement could offer a better quality of reconstructed speech, since the selection of the speech frame still to transmit is made in an inteligent way. In this arrangement, the spare capacity in the half-slot could be used to indicate to the receiver which of the three frames has been transmitted and also information regarding how to interpolate across the gaps.

In another preferred embodiment, when TETRA half slot stealing is to occur two speech frames both with deliberately reduced speech data content (i.e. of a reduced size) are transmitted in the unstolen half-slot.

The data reduction is achieved by removing,low 1 importance" bits output from the speech coding from t-..,i speech frames and, if necessary, by also removing some medium importance bits (which would preferably be i selected so as to minimise the impact of the loss of those bits upon speech quality).

To further reduce the size of the coded speech frame generated from a given raw speech frame to allc), two coded speech frames to fit into the remaining halfslot, the forward error correction coding rate o.

the convolution code could be adjusted as discussed above to reduce the size of the coded frames generate from a given amount of uncoded data. This would redti:

the error protection but allow more raw data to be transmitted. This could mean, for example, that sufficient reduction of the size of the speech frames, could be achieved simply by removing "low importance" bits from the speech, without the need to remove medii importance bits.

An alternative preferred arrangement is to transr in the remaining half-slot one speech frame in an un-modified fashion and one or more speech frames wit} deliberately reduced data (with that data loss again preferably being primarily from the low importance anc then medium importance bits output from the speech coding).

A further preferred arrange ment for facilitating half-slot stealing is to modify the stealing process such that less timeslot capacity is stolen for the control signalling (i.e. provided for the control signalling) and more capacity is provided for user traffic. Thus, for example, one third of a TETRA slot could be stolen, so as to provide sufficient capacity.

for transmitting two out of three of the APCO speech frames (although the data in the speech frames could also be reduced as discussed above if desired). An indication that such modified stealing has occurred should be transmitted to the receiver.

In each of the above arrangements, the speech frames to be transmitted (whether with reduced data or otherwise) are preferably selected on the basis of their relative "speech importance", as discussed above.

Each of the above stealing arrangements in which two speech frames are effectively transmitted in the remaining half-slot could additionally use a modified TETRA interleaving structure across the entire remaining half-slot, rather than maintaining speech frame boundaries, in order to provide the forward error correction coding with maximum immunity to burst errors.

A further consideration with the above arrangements is where end-to-end encryption is required. APCO uses is DES encryption which in APCO is applied as a stream cipher which is exclusive ORed with the data in the speech frames. This DES encryption should be applied before any channel (forward error correction) coding which relates to the air-interface in order to maintain compatibility with existing APCO systems.

APCO DES encryption does not modify the quantity of data, nor does it propagate errors. However, the encryption is initially synchronised with a key identity (KID = 16 bits), an algorithm ID (ALGID = 8 bits), and a message indicator (MI = 72 bits). The message indicator is then repeated in APCO systems every 18 speech frames to provide "late entry". This initial encryption synchronisation data and the subsequent late entry signalling must also be transmitted to a receiver to allow the receiver to decrypt the transmitted signal.

This encryption signalling could still be included in the user traffic (and fitted into the TETRA timeslot accordingly). The TETRA D-INFO and UINFO Packet Data Units (PDUs) could, for example, be used for carrying data embedded into the user traffic (e.g. synchronisation (Message Indicator) information) on the downlink (i.e. base station to mobile station) and 54 uplink (i.e. mobile station to base station, respectively. Such transmission would ensure that tle encryption synchronisation data maintains its relatiie synchronisation to the user data. 1 Another way of sending the encryption synchronisation data would be to always define the use of the APCO codec as being in partnership with DES encryption and therefore embed 4 bits per speech frame of re-synchronisation information into the (start) of each speech frame. The bits could optionally be classified for forward error correction coding. In t is arrangement the coded bit rate would need modificatioi to reduce the coded frame size according to one of ti techniques discussed above, and the encryption synchronisation should be put back together for 1 1 forwarding to any existing APCO systems, so that they can use the Message Indicator in a block, to provide backwards compatibility. The four bits could be set an easily recognisable pattern, e.g. all one's or all zerols, when an APCO codec is being used but there is APCO DES encryption.

However, as will be appreciated from the above, when using the arrangement where three APCO generated and channel coded speech frames are being transmitted n each TETRA timeslot, there may be no remaining room:1i the TETRA timeslot for any of the encryption synchronisation information to be transmitted. i Therefore, in another preferred embodiment of the present invention, the initial synchronisation information is sent as part of the call set-up signalling on a TETRA control channel, since that will not require capacity on the traffic channel. It is also preferred that late entry is effected by repeating the synchronisation information and the late entry signalling on the TETRA control channel. The TETRA D-SETUP and U-SETUP Packet Data Units (PDUs) could bEl used to carry the initial downlink and uplink, respectively, call-set up information and late entry signalling on the control channel. (The inclusion of an encryption synchronisation message into a D-SETUP PDU for either call initiation or late-entry will require a reference to the TETRA time so that the call synchronises correctly.) An alternative way to provide the late entry signalling would be to use TETRA half-slot stealing mechanisms. However, because, as will be discussed below, half-slot stealing in such an arrangement could imply the loss of a relatively larger amount of speech frames, it is preferred in such an arrangement that the late entry signalling is intercepted and reduced to a maximum of once per second in the TETRA system (the periodic repeat of the message indicator every 18 APCO speech frames represents a rate of 2.7 Hz). The "missing" message indicators in such an arrangement would be compensated for by continued correct synchronisation at the receiver, which should be readily achievable if the received circuits are correctly clocked.

Thus, where encryption is being used, all the extra encryption signalling is preferably mapped to the call set-up control signalling and stealing channels available within TETRA, rather than embedded in the user data to be transmitted as normal on the TETRA traffic channels. This allows the TETRA system to, for example, more readily distinguish the encryption control signalling from the user data. It in effect uses the existing processes available for transmitting control signalling to transmit the encryption control information, rather than trying to modify traffic transmissions to do so.

It should be noted that the message indicator used in APCO contains 64 bits of data and an undefined 8 bit reserved field. The 8 bit reserved field could be used as a cyclic redundancy check or other error check in

56 order to check the validity of the message indicator.1 I If the message indicator is then found to be corrupt,i is preferably not used to update the receiver's synchronisation.

In a number of the above half-slot stealing arrangements a deliberately truncated speech frame is sent. This could create difficulties when end-to-end APCO DES encryption is being applied, since the missin speech bits will have to be allowed for in the encryption coding so as to maintain correct synchronisation. one preferred way of achieving this to remove the bits to reduce the size of the speech 1 frame after encryption and then on reception to insert bits into the data stream to compensate for the "missing" bits so as to maintain correct decryption.

The inserted bits to be added to the partial speech frame are preferably chosen according to a predetermii4 pattern (e.g. all zeros), and/or are selected for minimum impact upon speech quality, since the insertE( bits will be presented to the speech codec. The pre-, determined pattern added to the partial speech frame c be fixed or rolling (i.e. changing frame by frame), ai desired.

It is preferred in all of the above embodiments that the APCO speech codec is flagged as a non-TETRA speech codec by asserting the "speech identifier" available in the CMCE (Circuit Mode Control Entity) protocol of TETRA, so as to allow compatibility and inter-working with pure TETRA systems.

Similarly, when using TETRA direct mode operatick (i.e. in which communication takes place not via the fixed radio network), the layer 3 circuit mode type definitions are preferably used to indicate the use of an alternative speech coding mechanism when APCO speec coding is being used. This could be advantageous whe:

for example, the direct mode terminal is directly I passing data on to an infrastructure, i.e. acting as ik direct mode gateway.

The end-to-end encryption and use of the APCO speech codec are preferably defined as supplementary services for the TETRA air-interface. This would mean, for example, that the control data required could be sent using the "facility" field in TETRA. If they are not defined as supplementary services, they can be defined as proprietary services, with the "proprietary" field in TETRA then being used for control data transmission.

It can be seen from the above that the present invention provides techniques for in particular mapping a constant-rate data-stream generated in frames (e.g. from a voice or video coder) on to a telecommunications is bearer which can carry data in a series of discrete timeslots at a normally constant data rate (except for occasional "stealing" of capacity for system control use), and also provides a communications terminal which can be used in such a communications system which in effect contains a low-bit digital speech encoder/decoder function and an error encoding/decoding function and a mechanism for mapping the output data into "frames" for transmission/reception. It in particular applies to the situation where the data frame rate and size does not match the timeslots of the radio bearer transmission system exactly on a one-to-one basis, i.e. the basic coded data frame size and timeslot size are not equal.

* In the particular preferred embodiment described, such a communications system is provided which also maintains backwards compatibility to two already defined radio systems (TETRA and APCO Project 25). In particular, techniques are provided for mapping speech generated by the APCO Digital Voice System Inc's Improved Multi-Band Excitation digital voice coder, which produces speech at a different frame-rate to the defined TETRA speech coding, onto TETRA's basic timeslot structure, by fitting three frames of coded APCO speech - 58 into each single TETRA timeslot (both of which correspond to 60 ms of speech). In these arrangemenLE the APCO error correction scheme, or modified version of the raw APCO speech data and/or of TETRA's forwaral error correction schemes, are used to allow mapping olf the APCO speech frames into the TETRA timeslot structure.

The invention also provides techniques for allow i for TETRA's timeslot and half-slot stealing protocolsi the combined APCO/TETRA system. In particular, modifications are made to the error correction scheme and/or the timeslot (bursts) building protocol when half-slot stealing is to occur, to mitigate the probl-, of TETRA half-slot stealing resulting, because the AP, speech frames do not exactly fit into a TETRA half-sl-. in more speech loss than with TETRA's normal speech encoding.

Finally, arrangements and modifications are proposed to allow the implementation of APCO DES encrypted frames into TETRA timeslots in a way which remains backwards compatible with APCO phase 1 and wjj can cope with TETRA half-slot stealing mechanisms.

All of these techniques are preferably used in appropriate combinations to provide a combined APCO/TETRA system.

Claims

1. A method of operating a radio unit for use in a digital radio system which can use two predetermined data transmission structures, each comprising timeslots of a defined size, and in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, each transmission data structure having its own defined data generation and channel coding processes, the method comprising:

generating from an input signal frames of digital data using one of the defined data generating processes; processing the generated frames of digital data using one of the defined channel coding processes to produce a series of coded data frames for transmission; and modifying the data generating process and/or modifying the channel coding process when it is intended to transmit the data in accordance with at least one of the defined data structures, so as to alter the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as-defined processes.

2. The method of claim 1, wherein the data generating process and/or channel coding process is or are modified so as to reduce the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified as-defined processes.

3. The method of claim 1 or 2, wherein the data bits in the generated data frames are classified into differing levels of relative importance and the subsequent channel coding process introduces differeri numbers of additional bits into the coded data dependli on the relative importance level of the given bits tol coded, further comprising when the data generating process and/or the channel coding process is or are be modified, modifying that process or those procesSE such that the channel coding process is effectively applied to a different classification of the importan I P levels of the generated raw data bits.

4. The method of claim 3, wherein the data generatir process and/or channel coding process is or are modifi such that the channel coding process interprets at 1E-_some of the raw data bits as having a different relaft is importance level to the level effectively allocated h the data generation process.

5. The method of claim 3 or 4, wherein the relative importance levels of one or more of the bits in the rE data are changed prior to channel coding of the data., 6. A method of preparing data for transmission in a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames which contain bits of differei levels of relative importance, which raw data frames z then further processed prior to transmission by a channel coding process in which the number of bits iii the coded data generated from a given bit in the raw data varies depending on the relative level of importance of the raw data bit, and which system has E pre-defined data generation process, the method comprising:

causing the channel coding process to be effectively carried out on raw data in which the relative importance levels of one or more of the raw data bits differs to the relative importance level tI the bits would have had under the defined data generation process, so as to alter the amount of data.in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame if the changing of the effective relative importance levels had not taken place.

7. The method of any one of the preceding claims, comprising modifying the channel coding process so as to alter the number of coded output bits produced as compared to when using the unmodified channel coding process.

8. A method of preparing data for transmission in a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are then further processed by a channel coding process prior to transmission, the method comprising:

generating from an input signal frames of digital data; processing the generated frames of digital data using a channel coding process to produce a series of coded data frames for transmission; and selectively modifying the channel coding process so as to reduce the amount of data in each coded data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified process.

9. The method of claim 7 or 8, wherein the channel coding process generates additional data bits at a given rate per original data bit, comprising modifying the channel coding process so as to alter the coding rate.

10. The method of claim 9, comprising, where the original data has bits of different levels of relative 62 - importance, altering the coding rates for one or morE-' but not all of the different levels of importance.

11. An apparatus for use in a radio unit for use in digital radio system which can use two predetermined data transmission structures, each comprising timesloit of a defined size, and in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are thf further processed by a channel coding process prior t-, transmission, each transmission data structure havinc its own defined data generation and channel coding processes, the apparatus comprising:

means for generating from an input signal frames- digital data using one of the defined data generatin--f processes; means for processing the generated frames of digital data using one of the defined channel coding processes to produce a series of coded data frames fo transmission; and means for modifying the data generating process and/or modifying the channel coding process when it j intended to transmit the data in accordance with at least one of the defined data structures, so as to al the amount of data in each coded data frame to be transmitted as compared to the amount of data that WOL be present in each frame when using the unmodified aSdefined processes.

12. The apparatus of claim 11, wherein the data bits the generated data frames are classified into differii levels of relative importance and the subsequent cha i coding process introduces different numbers of I additional bits into the coded data depending on the relative importance level of the given bits to be cod( further comprising means for, when the data generatir 'I process and/or the channel coding process is or are t I be modified, modifying that process or those processes such that the channel coding process is effectively applied to a different c lassification of the importance levels of the generated raw data bits.

13. The apparatus of claim 12, wherein the means for modifying the data generating process and/or channel coding process modifies that process or those processes such that the channel coding process interprets at least some of the raw data bits as having a different relative importance level to the level effectively allocated by the data generation process.

14. The apparatus of claim 12 or 13, comprising means for changing the relative importance levels of one or more of the bits in the raw data prior to channel coding of the data.

15. An apparatus for preparing data for transmission in a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames which contain bits of different levels of relative importance, which raw data frames are then further processed prior to transmission by a channel coding process in which the number of bits in the coded data generated from a given bit in the raw data varies depending on the relative level of importance of the raw data bit, and which system has a pre-defined data generation process, the apparatus comprising:

means for causing the channel coding process to be effectively carried out on raw data in which the relative importance levels of one or more of the raw data bits differs to the relative importance level that the bits would have had under the defined data generation process, so as to alter the amount of data in each coded data frame to be transmitted as compared to 64 the amount of data that would be present in each fram if the changing of the effective relative importance levels had not taken place.

16. An apparatus for preparing data for transmissioi-.i n a digital radio system, in which system data to be transmitted is generated from an input signal in the form of discrete frames, which raw data frames are they further processed by a channel coding process prior to transmission, the radio unit comprising:

means for generating from an input signal frame,,:. ()f digital data; means for processing the generated frames of digital data using a channel coding process to produc:e series of coded data frames for transmission; and means for selectively modifying the channel coding process so as to reduce the amount of data in each cc.-,( data frame to be transmitted as compared to the amount of data that would be present in each frame when using the unmodified process.

17. The apparatus of claim 16, wherein the channel coding process generates additional data bits at a given rate per original data bit, comprising means for modifying the channel coding process so as to alter ti coding rate.

18. A method of transmitting data in a radio communications system, which system supports stealing M data to allow the transmission of other data such as control signalling in a stream of data to be transmitted, and in which system data to be transmitte is arranged in a succession of discrete data frames, is transmitted in a succession of discrete timeslots, each timeslot usually containing three or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the method comprising:

when stealing of a portion of a timeslot is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot in such a manner that immediately adjacent data frames in the data to have been transmitted in the timeslot are not both stolen.

19. A method of transmitting data in a radio communications system, which system supports stealing of data to allow the transmission of other data such as control signalling in a stream of data to be transmitted, and in which system data to be transmitted is arranged in a succession of discrete data frames, and is transmitted in a succession of discrete timeslots, each timeslot usually containing two or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the method comprising:

when stealing of a portion of a timeslot is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot on the basis of an assessment of the relative importance levels of the data frames that were to be transmitted.

20. The method of claim 18 or 19, further comprising transmitting to the receiver an indication of which of the possible data frames has actually been transmitted.

21. The method of claim 18, 19 or 20, further comprising transmitting in any spare data capacity remaining in the timeslot additional information that will help to interpolate and/or compensate for the stolen data.

22. A method of transmitting data in radio - 66 communications system, in which system data to be I transmitted is arranged in a succession of data frame 1 3 which frames are transmitted in timeslots, each time,--z;L usually containing two or more data frames, and in WM system stealing of part of timeslot can occur to all< other data such as control signalling to be transmitl:.

in a stream of data frames, the method comprising:

when it is determined that stealing of part of =i timeslot is to take place, generating at least one da-- frame having a reduced size from one of the data fram that was to be transmitted in the timeslot to be stoll from, so as to allow data from that frame to fit into and be transmitted in the remaining unstolen part of timeslot.

23. The method of claim 22, wherein in the radio sys-raw data frames are channel coded to produce coded da: frames for transmission, further comprising reducing: size of the data frames by modifying the raw data fraT or the data frame channel coding process to produce smaller coded data frames from a given amount of raw data.

24. The method of any one of claims 18 to 23, furthe comprising modifying the stealing process such that a reduced size portion of the timeslot is stolen so as allow more room for the user data frames to be transmitted.

25. A method of transmitting data in a radio communications system, in which system data to be transmitted is generated as a succession of discrete I data frames and the data frames are transmitted in a succession of discrete timeslots, and in which system there is a predefined data stealing process whereby a portion of a timeslot of a predetermined size is replaced with other data to allow transmission of otl if data such as control signalling in a stream of data frames, the method comprising: when it.is determined that such stealing should occur selectively modifying the stealing process so as to steal a portion of the timeslot which is smaller than the size of said predetermined stealing portion so as to allow more space for the data frames that were to be transmitted to be fitted into the remaining unstolen part of the timeslot when such stealing occurs.

26. The method of claim 24 or 25, further comprising transmitting to the receiver an indication that a reduced size timeslot portion has been stolen.

27. The method of any one of claims 18 to 26, further comprising determining whether the data frames to be transmitted are of a predetermined size, and if it is determined that the data frames are not of said predetermined size, modifying the stealing process, but if it is determined that the data frames are of said predetermined size, not modifying the stealing process.

28. The method of any one of claims 18 to 27, wherein when two or more data frames are to be transmitted in the remaining portion of a partially stolen timeslot, and a data interleaving technique is used for the data transmission, further comprising using a modified interleaving structure across the entire remaining timeslot portion, rather than maintaining data frame boundaries.

29. A method of transmitting data in an encrypted form in a radio communications system, in which system data can be transmitted in one of two different predefined data structures, and in which system at least one of the data structures has defined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver, and ix which system data to be transmitted is generated from input signal in the form of discrete frames, which r,=LA data frames are then further processed by a channel 1 coding process prior to transmission, the method comprising:

when it is desired to transmit encrypted data generated and encrypted according to the processes defined for said at least one data structure over thE_. other data structure, including said encryption contj----:) information in the data transmission, and modifying th data generating process and/or modifying the channel coding process so as to alter the amount of data in e coded data frame to be transmitted as compared to the amount of data that would be present in each frame wh( using the unmodified as-defined processes.

30. A method of transmitting encrypted data frames, comprising:

allocating a predetermined number of bits of spac for encryption synchronisation information in each dat frame; and I setting the predetermined bits to a predetermine( pattern when there is no encryption.

31. A method of transmitting data in an encrypted for in a radio communications system, which system include both a control channel for control signal transmisslovi and one or more traffic radio channels for user data transmissions, and in which system data can be transmitted in one of two different predefined data structures and in which system at least one of the dat structures has defined for it a data generation proCeS and a data encryption process, which encryption proCEII requires for its use the transmission of encryption control information to the receiver and which encryption control information is in the defined process effectively embedded in the user data to be transmitted, the method comprising:

32. The method of claim 31, further comprising transmitting late entry and periodic encryption signalling on a control channel, rather than including it in the traffic channel transmission.

33. The method of claim 31, further comprising transmitting late entry and periodic encryption signalling by using a traffic channel data stealing process.

34. The method of any of any one of claims 18 to 33 further comprising, when data is to be encrypted and reduced size data frames are to be transmitted, removing the necessary number of data bits to reduce the frame size after encryption, and the receiver reinserting a corresponding number of bits into the data prior to decryption or prior to passing the data into another part of the system, to maintain correct decryption synchronisation.

35. The method of claim 34, wherein the bits reinserted by the receiver are added to the reduced size data frame according to a predetermined pattern.

36. A method of operating a radio unit for use in a digital radio communications system, in which system I data can be transmitted in two different pre-defined data structures each comprising a succession of timeslots, there being defined for each data structure a process for generating digital data for transmission from an input signal and a process for channel coding the generated data prior to its transmission, which da.a generation and channel coding processes provide data Ur transmission in the form of discrete frames, the methD3 comprising:

generating digital data for transmission from an input signal in accordance with the data generation process defined for a first one of said transmission data structures; is coding the data generated from the input signal prior to its transmission using the channel coding process defined for said first one of said transmiSSiDI data structures; when it is desired to transmit the coded data usUg the first of the two different predefined data I structures, transmitting the data as defined for thatd, transmission data structure; and when it is desired to transmit the coded data usig the second of the two different predefined data structures, apportioning the coded data frames genera id according to the data generation and channel coding processes for the first defined data structure into timeslots of the second defined data structure in sucAi a way that each timeslot of the second defined data structure contains only entire coded data frames. J 37. The method of claim 36, wherein the data for transmission is generated and channel coded according -o the processes defined for the APCO standard, and the APCO generated and coded data frames are apportioned into the TETRA timeslot structure for transmission over a TETRA air-interface by placing three APCO coded framis into every TETRA timeslot.

38. An apparatus for transmitting data in a radio communications system, which system supports stealing of data to allow the transmission of other data such as control signalling in a stream of data to be transmitted, and in which system data to be transmitted is arranged in a succession of discrete data frames, and is transmitted in a succession of discrete timeslots, each timeslot usually containing three or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the apparatus comprising:

means for, when stealing of a portion of a timeslot is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot in such a manner that immediately adjacent data frames in the data to have been transmitted in the timeslot are not both stolen.

39. An apparatus for transmitting data in a radio communications system, which system supports stealing of data to allow the transmission of other data such as control signalling in a stream of data to be transmitted, and in which system data to be transmitted is arranged in a succession of discrete data frames, and is transmitted in a succession of discrete timeslots, each timeslot usually containing two or more of the data frames, and wherein stealing of a portion of a timeslot can take place, the apparatus comprising:

means for, when stealing of a portion of a timeslot is to take place, selecting from the data frames which would have been transmitted in the timeslot, the data frame or frames still to transmit in the remainder of the timeslot on the basis of an assessment of the relative importance levels of the data frames that were to be transmitted.

72 40. The apparatus of claim 38 or 39, further comprisAg means for transmitting to the receiver an indication A which of the possible data frames has actually been transmitted.

41. An apparatus for transmitting data in radio communications system, in which system data to be transmitted is arranged in a succession of data frame3 which frames are transmitted in timeslots, each timesL'lt usually containing two or more data frames, and in whAh system stealing of part of timeslot can occur to allog other data such as control signalling to be transmitt.el in a stream of data frames, the apparatus comprising:

means for, when it is determined that stealing oE is part of a timeslot is to take place, generating at leAt one data frame having a reduced size from one of the data frames that was to be transmitted in the timesloL to be stolen from, so as to allow data from that fram to fit into and be transmitted in the remaining unstol-n part of the timeslot.

42. The apparatus of claim 41, wherein in the radio system raw data frames are channel coded to produce coded data frames for transmission, further comprising means for reducing the size of the data frames by modifying the raw data frames or the data frame channe coding process to produce smaller coded data frames fr m a given amount of raw data.

43. The method of any one of claims 38 to 42, further comprising means for modifying the stealing process su:h that a reduced size portion of the timeslot is stolen -0 as to allow more room for the user data frames to be i q transmitted.

44. An apparatus for transmitting data in a radio communications system, in which system data to be 73 - transmitted is generated as a succession of discrete data frames and the data frames are transmitted in a succession of discrete timeslots, and in which system there is a predefined data stealing process whereby a portion of a timeslot of a predetermined size is replaced with other data to allow transmission of other data such as control signalling in a stream of data frames, the apparatus comprising:

means for, when it is determined that such stealing should occur, selectively modifying the stealing process so as to steal a portion of the timeslot which is smaller than the size of said predetermined stealing portion so as to allow more space for the data frames that were to be transmitted to be fitted into the is remaining unstolen part of the timeslot when such stealing occurs.

45. The apparatus of claim 43 or 44, further comprising means for transmitting to the receiver an indication that a reduced size timeslot portion has been stolen.

46. The method of any one of claims 38 to 45, further comprising means for determining whether the data frames to be transmitted are of a predetermined size, and means for, if it is determined that the data frames are not of said predetermined size, modifying the stealing process, and if it is determined that the data frames are of said predetermined size, not modifying the stealing process.

47. An apparatus for transmitting data in an encrypted form in a radio communications system, in which system data can be transmitted in one of two different predefined data structures, and in which system at least one of the data structures hasdefined for it a data generation process and a data encryption process, which encryption process requires for its use the transmission of encryption control information to the receiver, and 74 in which system data to be transmitted is generated f-rI)m an input signal in the form of discrete frames, which raw data frames are then further processed by a chaniiel coding process prior to transmission, the apparatus comprising:

means for, when it is desired to transmit encryptZd data generated and encrypted according to the processe:

defined for said at least one data structure over the other data structure, including said encryption contr I information in the data transmission, and for modifyi: the data generating process and/or modifying the chariaml coding process so as to alter the amount of data in ea-h coded data frame to be transmitted as compared to the amount of data that would be present in each frame whai using the unmodified as-defined processes.

48. An apparatus for transmitting data in an encrypt! form in a radio communications system, which system includes both a control channel for control signal transmission and one or more traffic radio channels ft) user data transmissions, and in which system data can e transmitted in one of two different predefined data structures and in which system at least one of the da: structures has defined for it a data generation procel and a data encryption process, which encryption procE!3 requires for its use the transmission of encryption control information to the receiver and which encrypi-A n control information is in the defined process effectively embedded in the user data to be transmitt:

the apparatus comprising:

means for, when it is desired to transmit encryp-A data generated and encrypted according to the proces.s, defined for said at least one data structure over the, other data structure, removing said embedded encryptio control information from the data to be transmitted over the traffic channel and transmitting it instead over a control channel.

49. A radio unit for use in a digital radio communications system, in which system data can be transmitted in two different pre-defined data structures each comprising a succession of timeslots, there being defined for each predefined data transmission structure a process for generating digital data for transmission from an input signal and a process for channel coding the generated data prior to its transmission, which data generation and channel coding processes provide the data for transmission in the form of discrete frames, the radio unit comprising:

means for generating digital data for transmission from an input signal in accordance with said data generation process defined for a first one of said transmission data structures; means for coding the data generated from the continuous input signal prior to its transmission using said channel coding process defined for said first one of said transmission data structures; and means for, when it is desired to transmit the coded data using the second of the two different predefined data structures, apportioning the coded data frames generated according to the datageneration and channel coding processes for the first defined data structure 2S into timeslots of the second defined data structure in such a way that each timeslot of the second defined data structure contains only entire coded data frames.

SO. The radio unit of claim 49, wherein the data for transmission is generated and channel coded according to the processes defined for the APCO standard, further comprising means for apportioning the APCO generated and coded data frames into the TETRA timeslot structure for transmission over a TETRA air-interface by placing three APCO coded frames into every TETRA timeslot.

51. A computer program element comprising computer 76 software code portions for performing the methods of y one of claims 1 to 10 or of any one of claims 18 to when the program element is run on a data processing means. 5 52. A method of operating a radio unit substantiallyl s hereinbefore described.

53. A method of preparing data for transmission substantially as hereinbefore described.

54. A method of transmitting data substantially as 11 hereinbefore described. q 55. An apparatus for use in a radio unit substantial-:1-:, as hereinbefore described.

56. An apparatus for preparing data for transmission, substantially as hereinbefore described.

57. An apparatus for transmitting data substantially hereinbefore described.

58. A radio unit substantially as hereinbefore described.