WO1991003894A1

WO1991003894A1 - Signalling information multiplexer

Info

Publication number: WO1991003894A1
Application number: PCT/CA1990/000269
Authority: WO
Inventors: Douglas James Millar; John F. Michaelides
Original assignee: Northern Telecom Limited
Priority date: 1989-09-11
Filing date: 1990-08-22
Publication date: 1991-03-21
Also published as: AU6172890A

Abstract

A method of multiplexing blocks of signalling information along with normal voice and data information, on a single data channel is disclosed. In particular, an error correction code is added to each block containing the signalling information. The resulting multiplexed blocks are then sent through the data channel (20) to a receiver (25). At the receiver (25), an error correction is performed on each incoming block. If the error correction was successful, then signalling information was identified and decoded. However, if the error correction was unsuccessful, then normal information was sent and that block will be normally processed.

Description

SIGNALLING INFORMATION MULTIPLEXER

This invention relates to the multiplexing of communications signals, but more particularly, to the multiplexing of supervisory signalling information into the normal voice and data information bit stream.

Background of the Invention

In a cellular radio channel communication system, it is common to send a supervisory signalling stream of information multiplexed onto the normal voice or data stream of information. The supervisory information is sometimes used to identify the mobile unit during cell handoff or is used for timing and power level adjustments. The problem associated with sending supervisory signalling information, is that in order to differentiate between the two types of signals being transmitted, marker bits are used to separate the information. However, these marker bits are also subject to transmission errors, with the result, that the decoder may wrongly interpret the incoming data as a normal stream of information, as opposed to signalling information. The use of marker bits not only increases the required capacity of the system, but also, is not always reliable when prone to errors. There accordingly exists a need for a new method of multiplexing a signalling stream of information with the normal voice and data stream of information and which can be used when error rates on the communication channel is high. It is therefore an object of the present invention to provide a method of multiplexing signalling and normal streams of information onto a single data channel, such that supervisory signalling information is detected correctly, even with large error rates in the transmission.

Another object of the present invention is to provide a method of multiplexing signalling and normal streams of information onto a single data channel, wherein no extra bit(s) is required in the supervisory signalling information block to indicate the type of information that is carried and wherein the normal voice and data information can be transmitted unmodified.

Summary of the Invention

The present invention provides a method of multiplexing blocks of signalling information along with normal voice and data information, on a single data channel. In particular, an error correction code is added to each block containing the signalling information. The coded blocks of signalling information are then multiplexed with blocks of normal information. The resulting multiplexed blocks are then sent through the data channel to a receiver. At the receiver, an error correction is performed on each incoming block. If the error correction was successful, then signalling information was identified and decoded. However, if the error correction was unsuccessful, then normal information was sent and that block will be normally processed.

Brief Description of the Drawings

Figures la and lb are illustrations of the frame coding arrangement according to the present invention; and Figure 2 is an illustrative block diagram for use in describing the multiplexing method of the present invention.

Description of the Preferred Embodiment In the present embodiment, it can be assumed that the information is carried on the channel in blocks of bits of constant length. In addition, it can be assumed that the receiver is able to extract the information one block at a time. As shown in Figure la, the encoded speech signal occupies two slots of a TDMA frame in the case of a 16K bits/s rate and one slot in the case of an 8K bits^'/s rate. In both cases encoded speech is divided into two fields. The first field contains the most important information and thus is protected from the high error rate of a fading radio channel by forward error correction (Reed-Solomon) . The second field contains the less important information and is not protected. The two fields are time multiplexed into one or two slots per slot of TDMA frame, as shown in Figure la, for the one slot case. When a supervisory message is to be transmitted it replaces the unprotected speech field of the slot, as is shown in Figure lb. In this case, all the supervisory message blocks are concatenated to form a complete message.

Figure 2, shows for illustration purposes, a general block diagram describing the operation of the present invention. At the transmitter 10, each block to be transmitted contains either normal information 11 (i.e. voice and/or data) or, alternately and on demand, supervisory signalling information 12. The transmission rate is the rate required to transmit normal information. When signalling information 12 is to be transmitted, it is injected in or multiplexed with the normal information 11 by multiplexer 13. The signalling block will appear in the bit stream 14, as a separate block. The insertion of signalling information can be viewed as "stealing bits" from the normal information. In the preferred embodiment, data channels should tolerate corruption of some information blocks or at least a temporary loss of throughput. Because of the potential for errors on channel 20, an error protection code such as the Reed-Solomon forward error correction code is added, at coder 15, to each block of signalling information before transmission over a channel 20. The message to be transmitted consists of two parts (a) the information bits, which compose the actual message given by the higher layer and (b) the parity bits, that are added by the error correction code. All information bits are protected since they are all of equal importance.

The speech coder may indicate the best time at which a supervisory message could be transmitted. The best time is determined from collected statistics. This minimizes the degradation of the speech quality resulting from the replacement of the unprotected speech field by supervisory messages. The normal information 11 can either be sent unprotected or protected with another coding scheme.

Upon the arrival of a frame at receiver 25, the incoming stream is demultiplexed at demultiplexer 30 and each incoming block 31 is treated as if it contained signalling information. Reed-Solomon decoding is performed on each block at decoder 32. If the error correction was successful, then signalling information was identified and the message is extracted and the remainder of the frame is used by the speech decoder. However, if the error correction was unsuccessful, then this indicates that the incoming frame contained normal information and the speech coder takes over to perform its normal decoding procedure. A property of some error correction codes, is that they can indicate whether they were able to correct all errors in the transmission or, on the contrary, that too many errors occurred and the block should be rejected. Normal information can be mistakenly interpreted as signalling information, but it has been demonstrated that this will occur very rarely with blocks of sufficient length. A (13,6) block code can be used for a supervisory message with 6 bits for each Reed-Solomon symbol. The 13 Reed- Solomon symbols are inserted in one slot only (for compatibility with the 8K bits/s scheme as well as to minimize the impact on speech quality) in the field normally occupied by the unprotected speech. There are in total 36 data bits which compose the supervisory message. The Reed-Solomon decoder locates and corrects up to two symbol errors. No erasure information is used by the decoder. Three parity symbols are left for error detection. This provides 18 bits of detection parity against corrupted supervisory messages.

In a practical environment, a very high error rate on the channel is possible. The error correction scheme may occasionally be unable to correct all the errors and a signalling block will be wrongly treated as normal data.

Accordingly, in the preferred embodiment, the normal data channel should be able to tolerate the occasional reception of such erroneous data. For fail-safe operation of the signalling channel, an Automatic Retransmission Request (ARQ) protocol can be implemented. The ARQ protocol has a simple Stop and Wait scheme. When the receiver decodes a supervisory message successfully, it sends an acknowledgement (ACK) on the reverse channel 33. The acknowledgement (ACK) is sent the same way as a supervisory message. If it does not receive an ACK within a specified timeout period, it will retransmit the previous message. The proposed method combines forward error correction with an Automatic Repeat Request (ARQ) protocol to provide for an error free transmission of the supervisory messages. It also provides for the distinction of speech from supervisory messages without using any extra bit(s) by exploiting error correction code properties.

Claims

CLAIMS :

1. A method of multiplexing blocks of signalling information along with normal voice and data information, on a single data channel, comprising the steps of: adding an error protection code to each of said signalling blocks; multiplexing said blocks of signalling information with a block of normal information; and coding the resulting blocks of information for transmitting over a data channel to receiver means.

2. A method of multiplexing as defined in claim 1, further comprising the steps of: performing an error correction at said receiving means, wherein said blocks of signalling information are identified and decoded if said error correction was successful and normal information is identified if said error correction was unsuccessful.

3. A method of multiplexing blocks of signalling information along with normal voice and data information, on a single data channel, comprising the steps of: adding a first error correction code to said blocks of signalling information; multiplexing said coded blocks of signalling information with uncoded blocks of normal information; sending said multiplexed signal through said data channel to receiving means; and performing an error correction at said receiving means, wherein said blocks of signalling information are identified and decoded if said error correction was successful and normal information is identified if said error correction was unsuccessful.

4. A method as defined in claim 3, wherein said normal information is multiplexed with said signalling information without an error protection code.

5. A method as defined in claim 3, wherein said normal information is multiplexed with said signalling information with a second error protection code.

6. A method as defined in claim 3, wherein said multiplexing of information is done in the time domain.

7. A method as defined in claim 3, wherein an Automatic Repeat Request (ARQ) protocol is used to ensure successful transmission of said signalling information.