GB2331897A

GB2331897A - Decoding data with two different error protection levels

Info

Publication number: GB2331897A
Application number: GB9725358A
Authority: GB
Inventors: Soren Sorensen; Ken Jackobsen; Timothy Simon Klein
Original assignee: Motorola AS
Current assignee: Motorola Solutions Danmark AS
Priority date: 1997-12-01
Filing date: 1997-12-01
Publication date: 1999-06-02
Anticipated expiration: 2017-12-01
Also published as: GB2331897B; GB2331897A9; GB9725358D0

Abstract

A method of decoding data in a radio communication system includes receiving a first data block (54) having a first error protection level and a second data block (56) having a second error protection level lower than the first level. The first data block (54) is then decoded and a decision made as to whether to decode and/or error check the second block of data (56) is made based on the determination of the first data block (54). If the first block is determined to be in error the second block is likely to also be in error and it is therefore not decoded or error checked. Preferably, the error protection involves cyclic redundancy checking (CRC) schemes, with a 16 bit CRC being used on the first block and a 4 bit CRC being used on the second block.

Description

1 111 2331897 METHOD AND APPAIRATUS FOR DECODING DATA

Field of the Inven

This invention relates generally to the encoding and decoding of data. The invention is applicable to, but not limited to, the encoding and decoding of data blocks having differing levels of error protection.

Baround of the Invention Many communications systems support data transfer between two points, perhaps through a communication link comprised of smaller links (hops). The data is often segmented into frames for transmission and may include voice and/or "true" data transmissions; it being noted that in digital communications systems speech transmissions use binary formatting and are therefore transmitted as data. The frames are protected by error detection or error correction information transmitted with each frame. An error occurring during transmission of data may result in a re-transmission of the affected frame, causing undesirable delay to the users. However, in some cases, re-transmission of data may be impractical, and error mitigation for the affected frames may be preferable. Alternatively, an error in a speech frame may cause popping or distortion effects that are annoying.

For real time applications, such as voice or video transmissions, a bad frame substitution (BFS) technique may be employed to estimate or correct for errors in the received frame, with the errors determined by say, a Cyclic Redundancy Checking (CRC) method. For such real time applications, forward error correction (FEC) schemes are also used, and are developed on a case by case basis. FEC uses redundancy to allow the receiver of a corrupted digital signal to determine the actual signal sent. FEC thus mitigates against data corruption arising due to error prone transmission paths. Furthermore, interleaving of a number of data blocks can be used to further spread out the data blocks for transmission and thereby reduce the effect of interference in a localised data stream. Such intermittent interference can be caused in a radio communication by fading and/or multipath effects or indeed by 2 external equipment such as radio signals from an another communications system.

In the case of speech, the development of an FEC scheme usually involves creating a different FEC scheme for each and every system, taking into account the different impact of bit-errors during transmission on different parameters of the bit stream. In designing an FEC scheme for a particular type of media transmission e.g. speech or video, it is desirable to employ some type of joint source/channel or unequal protection coding technique to allow greater protection for the more important segments of the data stream, whilst some of the bits may not receive any FEC protection at all.

Such differing requirements on the FEC protection of certain parameters or bits are a known characteristic of low bit-rate source coders for a single source, i.e. which act on one particular type of media transmission. For example in the context of video coding, an error in the motion vectors would cause a greater perceived degradation than an error in the DCT (Discrete Cosine Transform) coefficients. This requirement implies that the FEC is applied with detailed knowledge of the source coding algorithm.

Current mobile system FEC schemes for single services; e.g.

speech, almost invariably make use of Rate Compatible Punctured Convolutional (RCPC) Coding in order to provide differential error protection for the different bits of the coder bit stream. See for example the prior art publication Hagenauer, "Rate-Compatible Punctured

Convolutional Codes (RCPC Codes) and their Applications", IEEE Trans Comms, Vol. COM-36, No. 4, April 1988. However, in such single service applications, the number of bits protected is usually fixed, as are the positions in the FEC frames of the different parameters.

In current mixed data communications systems, where speech and data transmissions are inter-woven to a significant degree, providing high error protection for certain important data has a converse effect on data of lesser importance with the provision of relatively poor error protection capabilities. One typical example would be where a signalling message is sent adjacent to a speech message in the same block, or frame, of data. The signalling message may be provided with high levels of error protection, and consequently the speech data block may then be allocated poor levels of error protection.

3 The allocation of poor levels of protection to say, speech frames, increases the possibility of passing speech frames that have been corrupted. Such acceptance of corrupted speech frames causes annoying distortion effects and sometimes loud popping noises in the recovered speech signal, particularly with speech coders in digital communications systems.

This invention seeks to provide an improved radio communication system, radio unit and method of decoding blocks of data to mitigate at least 10 some of the aforementioned disadvantages.

SWiMuMS of the hw-eAnon In a first aspect of the present invention a method of decoding data in a radio communication system is provided. The method includes the steps of receiving a first data block having a first error protection level and a second data block having a second error protection level lower than the first error protection level. The first data block is decoded and a decision made as to whether to error check or decode the second block of data based on said determination of the first data block.

In this manner, subsequent data having low levels of error protection can be assessed for the likelihood that they are in error based on the decoding of the first block of data. As a beneficial consequence of this determination, the second block of data can be rejected if the first, better protected, block of data is found to be in error. This improves the decoding process and helps to prevent erroneous received data to be incorrectly deemed to be valid and error free, due to the low protection level applied to it.

Preferably the step of decoding the first data block determines whether the first data block has been received substantially error-free, and the second data block is error checked and/or decoded primarily only when the determination of the first data block is substantially error-free. The first data block includes substantially better error protection than the second data block such that if the first data block is determined to be in error upon decoding the first data block, the second data block is likely also in error and is subsequently not decoded or error checked. In the preferred embodiment of the invention, the first and second data block use the same 4 error protection scheme, which is a cyclic redundancy checking (CRC) scheme, where there is a 16 bit CRC on the first data block, a stealing block, and a 4 bit CRC on the second data block, a speech block. In this manner, the decision on whether to error cheek and/or decode the speech block substantially prevents falsing of the codec on receiving non-error free blocks. A third synchronisation data block is used, adjacent to the first and second data blocks, for synchronising the timing of the receiver to the data blocks.

A preferred embodiment of the invention will now be described, by way of example only, with reference to the drawings.

Brief d2tion of the Dras FIG. 1 shows a block diagram of a communication system according to a preferred embodiment of the present invention.

FIG. 2 shows a timing diagram applied to the communication system 20 according to the preferred embodiment of the present invention.

FIG. 3 shows a flowchart of a process of determining, and perhaps subsequently discarding, blocks of data in the aforementioned communication system of the preferred embodiment of the present invention.

Detailed DesMpAon of the Drawings In its most general form, the invention concerns the decoding of data blocks and determination of the usability of subsequent data blocks based on such decoding.

In a first aspect of the present invention, FIG. 1 shows a data and 35voice communication system 10 in which a remote unit, for example a mobile or portable radio unit 12 establishes a communication with a second remote unit 14 over a communication link via a base station site 11. In the preferred embodiment of the invention, the remote units 12 and 18 are subscriber units capable of wireless digital voice communications over radio frequency (RF) channels, and the base station 11 is operably coupled to further base stations and/or system controllers, either by wireline or over-the-air transmissions, to form an infrastructure. Communication systems so organised are known in the art.

Generally, the remote (subscriber) unit 12, in a transmitting mode of operation, encodes speech into data frames via a speech encoder 18, and supplements the data frames with encoded error detection information for channel transmission, via a channel encoder 20. The transmitting subscriber unit 12 then modulates the data frames on a carrier signal, via a modulator 22, before communicating with the base station 11. The base station 11 typically receives and demodulates the channel transmission.

The base station 11 then performs channel decoding while checking for transmission errors using the encoded error detection information, before forwarding the transmission to the second remote unit 14.

The second remote (subscriber) unit 14, in the scenario described is in a receiving mode of operation and has the same transmitter and receiver functionality of the first remote unit 12. The second remote unit demodulates the data frames from the received carrier signal, in a demodulator 26 to provide data for speech decoding and channel decoding in speech decoder 30 and channel decoder 28 respectively.

Referring now to FIG. 2, a timing diagram applied to the communication system of the preferred embodiment of the present invention, is shown. The operation of the invention is described with reference to the Trans-European Trunked Radio (TETRA) communication system, being developed by the European Telecommuni cations Standard Institute (ETSI). The timing diagram includes a received data burst 50 having a training synchronisation block 52, together with first and second data blocks 54, 56. Although FIG. 2 shows the training synchronisation block 52, preceding the two data blocks 54, 56, it is within the contemplation of the invention, that a variety of other arrangements with additional data blocks would benefit from implementing this invention, including alternative block positioning.

In TETRA, true data fl-ames in a traffic channel can be "stolen" by users of adjacent frames in circumstances where they are unused and the user of the adjacent frame requires increased capacity. These frames are termed STealing CHannels (STCH). Alternatively, the frame can be used -"v 6 for transmission of speech. In TETRA, a number of training sequences can be provided, depending upon the type of transmissions to be used in the data burst 50:

(i) a first training sequence 58, includes two speech frames. In this situation the two speech frames are interleaved and provided with 8 bits of CRC to provide adequate protection; (ii) a second training sequence 60 may be used which includes a STCH followed by a speech frame. In this situation the STCH is dealt with independently of the speech frame and is afforded 16 bits of CRC protection. 10 The subsequent speech fi-ame is only provided with 4 bits of CRC.

Consequently, in a worst case scenario, a speech frame containing errors will be deemed to be error-free every 1:16 (2A4) frames. Passing speech frames having errors has a significantly degrading impact on a digital speech coder performance, particularly with the Adaptive Codebook Excited Linear Prediction (ACELP) used in TETRA; (iii) a third training sequence 62 may also be used containing two STCH frames. In this situation the two STCH frames are not interleaved and are kept independent. Each STCH frame is provided with 16 bits of CRC.

It is also within the contemplation of the invention, that a weighting algorithm can be applied to adjacent blocks. In this manner, the applicability of adjacent blocks for use in determining whether a particular data block should be decoded can be assessed by the allocated weighting factor. Data blocks closer in the timing to the particular data block to be decoded would have a higher wieghting due to the likelihood that that data block would suffer the same fading or multipath interference effects as the particular data block to be decoded. Furthermore, 'close' data blocks having more protection would be given a higher weighting factor than similarly close data blocks having less protection.

Referring now to FIG. 3, a flowchart is shown of a process of determining, and perhaps subsequently discarding, blocks of data in the communication system of the preferred embodiment of the present invention. The flowchart includes the step of receiving a data burst 50, as in step 100 and decoding the training sequence, as shown in step 102. If the training sequence is not decoded correctly, the process moves to the next received data burst, as in step 108. If the training sequence is decoded correctly, and it is determined that the next data block is a speech data 7 block, the CRC is checked to see if the speech has been corrupted, as in step 104. If the speech has been corrupted, the process moves to the next received data burst, as shown in step 108. If the speech has not been corrupted, the speech data block is decoded, as in step 106, before the process moves to the next received data burst, as shown in step 108.

If the training sequence is decoded correctly, and it is determined that the next data block is a STCH data block, the CRC is checked to see if the STCH data block has been corrupted, as in step 110. If the STCH data block has been corrupted, the process moves to the next received data block, as shown in step 118. If the next received data block is determined to be a speech dat.6 block, the process assumes that this speech data block is also corrupted - it having less protection than the earlier, corrupted STCH frame. In this case, the subsequent speech data frame is not error checked or decoded and the process moves to the next received data burst, as shown in step 108. If the first STCH data block in step 110 has been corrupted, and the second data block in step 118 is a second STCH block, the second data block undergoes a CRC cheek to see if it is worth decoding. If the second data block has also been corrupted in step 120, the process moves to the next received data burst, as shown in step 108. If the second data block has not been corrupted in step 120, the second data block is decoded in step 122, before the process moves to the next received data burst, as shown in step 108.

If, in the above case, the training sequence is decoded correctly, and it is determined that the first data block is a STCH data block, and that the CRC is checked and found not to have been corrupted, as in step 102 and step 110, the first STCH data block is decoded in step 112. The CRC is then checked for the second data block, whether it is a STCH data block or a speech data block. If the second data block has been corrupted, the process moves to the next received data burst, as shown in step 108, otherwise the second data block is decoded, as in step 116, before the process moves to the next received data burst, as shown in step 108.

Hence, in communications systems that provide a mixed voice and data capability, thereby likely offering different levels of protection to the differing data transmissions, a method of discarding voice frames, that are highly likely to be in error but may be determined to be not corrupted due to their low protection level, is provided. This substantially prevents falsing of speech codecs in situations where speech frames are accepted, but are now deemed to be in error.

1 8 Thus, a communications system, a radio unit for operation on the communication system and a method of decoding blocks of data have been provided that mitigate some of the aforementioned disadvantages with 5 prior art decoding arrangements.

1 1

Claims

9 Claims

1 1 1. A method of decoding data in a radio communication system comprising the steps of.

receiving a first data block having a first error protection level and a second data block having a second error protection level lower than the first error protection level; decoding the first data block; deciding whether to error cheek or decode the second block of data based on said determination of the first data block.

2. A method of decoding data in a radio communication system in accordance with claim 1, wherein the step of decoding the first data block is to determine whether the first data block has been received substantially error-free, the method further comprising the step of error checking or decoding the second data block when said determination of the first data block is substantially error-free.

3. A method of decoding data in a radio communication system in accordance with claims 1 or 2, wherein the first data block includes substantially better error protection than the second data block such that if the first data block is determined to be in error upon decoding the first data block, the second data block is likely also in error and is subsequently not decoded or error checked.

4. A method of decoding data in a radio communication system in accordance with claims 1, 2 or 3, wherein the first and second data block use the same error protection scheme.

5. A method of decoding data in a radio communication system in accordance with any one of the preceding claims, wherein the error protection scheme is a cyclic redundancy checking (CRC) scheme.

6. A method of decoding data in a radio communication system in accordance with any one of the preceding claims, wherein there is a 16 bit CRC on the first data block and a 4 bit CRC on the second data block.

7. A method of decoding data in a radio communication system in accordance with any one of the preceding claims, wherein the first data block is a stealing block.

8. A method of decoding data in a radio communication system in accordance with any one of the preceding claims, wherein the second data block is a speech block and the decision on whether to decode or error check the speech block substantially prevents falsing of the codec on receiving non-error free blocks.

9. A method of decoding data in a radio communication system in accordance with any one of the preceding claims, wherein there is a third synchronisation data block adjacent to the first and second data block for synchronising the timing of the receiver to said blocks.

10. A method substantially as hereinbefore described with reference to, or as illustrated by, FIG. 3 of the drawings.

11. Apparatus adapted to perform the method of any one of claims 1-9.

12. A communications system adapted such that an apparatus according to claim 11 can perform the method of any one of claims 1-9.