WO2005024785A1

WO2005024785A1 - Digital transmission system

Info

Publication number: WO2005024785A1
Application number: PCT/GB2004/050002
Authority: WO
Inventors: Kendall Castor-Perry
Original assignee: Toumaz Technology Ltd
Current assignee: Toumaz Technology Ltd
Priority date: 2003-09-11
Filing date: 2004-08-19
Publication date: 2005-03-17
Anticipated expiration: 2006-03-11
Also published as: GB2406019B; GB2406019A; GB0321271D0

Abstract

A method of transmitting information from a transmitter to a receiver over a transmission link. The method comprises decomposing said information into a first signal being an approximation to said information, and at least one further signal providing an approximation to the error between said first signal and said information. Said first signal is sent from the transmitter to the receiver using a first transmission mechanism, and said at least one further signal is sent from the transmitter to the receiver using a second transmission mechanism. Said first transmission mechanism is designed to be more robust to noise and interference than said second transmission mechanism. Said first signal and optionally at least one further signal is received at the receiver, and an approximation to said information recovered.

Description

Digital Transmission System

Summary of the Invention

The present invention relates to digital transmission systems and in particular, though not necessarily, to digital transmission systems for transmitting audio signals over a wireless link.

Background to the Invention

Whilst many data transmission systems still rely upon the use of analogue modulation schemes to transport data over a link, there is an increasing tendency to use digital modulation schemes due in part to the improved transmission quality which such schemes provide. Digital modulation schemes generally prove resistant to noise and other interference at least up to certain limits, whilst analogue schemes will be degraded in direct relation to the levels of noise and interference.

A typical example of a system which might make use of a digital modulation scheme is one which establishes a wireless communication link between a transmitting and a receiving device to carry audio information between the devices. Nearly all such systems will have a limited bandwidth available, limiting the maximum bit rate at which data can be transmitted over the wireless link. In order to optimise the use of that bandwidth, the system may code the audio data using a suitable compression algorithm. For example, an adaptive prediction algorithm may be used in which a prediction of a current sample value is made based upon a small set of preceding sample values (e.g. 160) and a set of prediction coefficients which are defined for each block or segment of speech. A residual value is then calculated by subtracting the predicted value from the actual value. As the dynamic range of the residual values are likely to be significantly less than that of the signal itself, and only one set of prediction coefficients is required per block, significant compression is achieved. (The signal can be compressed still further by coding the residual values, e.g. using an excitation codebook.) The prediction coefficients and coded residual values are multiplexed to form a single coded signal and are modulated onto a carrier signal for transmission. The receiver is able to reconstruct an approximated audio signal by generating predicted values using the prediction coefficients and adding the respective decoded residual values. More complex and efficient coding algorithms using similar principles have been developed.

In the presence of interfering signals or physical restrictions on bandwidth or a reduction of signal level (e.g. due to increasing distance between transmitter and receiver), the ability of a conventional digital system to recover data from a carrier signal may be seriously compromised. Once the degradation of the digital signal exceeds some threshold, it may not be possible to recover any information at the receiver. Of course, increasing the transmission power may raise the threshold at which information loss occurs, although this will place a greater burden on the transmitter and, at least in the case of wireless systems, will increase electromagnetic interference in neighbouring areas. Error correction codes and other techniques may be used to improve reception quality, but these will have limits and also place greater processing requirements and therefore increase power consumption at both the transmitter and receiver. Whilst analogue systems tend to allow a graceful degradation of reception quality, digital systems often offer only high quality reception or no reception at all.

A further drawback of digital systems is that they do not allow the receiver to trade off reception quality against power consumption. The receiver must recover all of the transmitted data in order to be able to generate the original information. There is no option to recover only a part of that information and generate from that information lower quality but still comprehensible information. In many transmission systems, the processing and electrical power available to a receiver component is significantly less than that of a transmitting component. When the power available to the receiver is low, e.g. the battery source is running low, or to extend battery lifetime, it may be acceptable to trade off power consumption against reception quality. Digital systems in general do not offer this trade off.

Summary of the Invention It is an object of the present invention to provide a digital transmission system which allows for a trade off between reception quality and power consumption and which provides for a graded decrease in reception quality with relative increases in noise and interference levels.

According to a first aspect of the present invention there is provided a method of transmitting information from a transmitter to a receiver over a transmission link, the method comprising: decomposing said information into a first signal being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information; transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism; and receiving said first signal and optionally at least one further signal at the receiver, and recovering an approximation to said information.

Embodiments of the invention enable a receiver to combine the first and the or each further signal to recover a high quality version of the original information. The receiver may recover a lower quality version of that information by recovering only the first signal or the first signal and a proportion of the further signals in the case where two or more farther signals are transmitted. The receiver may choose to recover only a lower quality signal, e.g. to reduce power consumption, or this may be essential because the reception quality of the further signal, or one or more of the further signals, is degraded below some threshold level.

Preferably, said first transmission mechanism comprises the modulation of said first signal on a carrier signal, and said second transmission mechanism comprises the modulation of said second signal on that same carrier signal. More preferably, the first and second transmission mechanisms use different modulation schemes such that the first mechanism is more robust to the second mechanism. In certain embodiments of the invention, said transmission link is a wireless link, and said receiver and transmitter comprise wireless reception means and wireless transmission means respectively. Whilst the wireless link may be for example an infrared link or an ultrasonic link, a preferred link is a Radio Frequency (RF) link.

The present invention is applicable in particular to the transmission of audio information, as audio information lends itself to graded reductions in quality whilst at the same time being comprehensible to a listener upon playback. However the invention is also applicable to other forms of information such as video which also lend themselves to graded quality reductions.

It will be appreciated that the first and further signals may be coded at the transmitter and decoded at the receiver, e.g. to achieve compression and decompression.

Said at least one error signal may contain actual error values or components for reconstructing error values, or may contain filter coefficients so that when the approximated signal is passed through a filter an improved approximation is recovered.

In the case where information is coded at the receiver using an adaptive prediction algorithm, said first signal comprises for each sample a set of prediction coefficients, whilst said at least one further signal comprises for each sample a residual value.

According to a second aspect of the present invention there is provided an information transmission system comprising a transmitter and a receiver for transmitting information over a communication link, the transmitter comprising processing means for decomposing said information into a first signal being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information, transmission means for transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and for transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism, and the receiver comprising receiving means for receiving said first signal and optionally at least one further signal at the receiver, and processing means for recovering an approximation to said information.

According to a third aspect of the present invention there is provided a transmitter for transmitting mformation over a transmission link, the transmitter comprising: processing means for decomposing said information into a first signal being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information, transmission means for transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and for transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism.

According to a fourth aspect of the present invention there is provided a receiver for recovering information sent over a transmission link, the receiver comprising: means for receiving one or more of a set of signals sent over said transmission link, one of said set of signals being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information; processing means for selectively defining a number of signals to be processed, and for combining the selected signal or signals to recover an approximation to the sent information.

According to a fifth aspect of the present invention there is provided a hearing aid comprising a receiver according to the fourth aspect of the invention, wherein the information sent over said link is audio information and said processing means is arranged to play the recovered approximate information to the user. Brief Description of the Drawings

Figure 1 illustrates a system for allowing a hearing impaired person to listen to broadcast audio signals;

Figure 2 illustrates schematically a transmitter of the system of Figure 1 ;

Figure 3 illustrates schematically a receiver of the system of Figure 1; and

Figure 4 is a flow diagram illustrating a method of operation of the receiver of Figure

2.

Detailed Description of Certain Embodiments

Figure 1 illustrates an audio signal broadcast system which might for example be used in an auditorium to allow hearing impaired persons to listen to a speaker giving a presentation from a lectern 1 using a microphone 2. The audio signal from the microphone is sent to a broadcast unit 3 which typically amplifies the signal and transmits it to a set of loudspeakers (not shown) distributed around the auditorium. In addition, the broadcast unit 3 converts the audio signal from the analogue to the digital domain and broadcasts the digital signal using modulation of one or more Radio Frequency (RF) carriers. The broadcast is typically of relatively low power so that the broadcast can only be picked up within the auditorium. Specially designed hearing aids 4 (the Figure shows a behind the ear type aid) have receiving and processing circuitry for receiving and processing the broadcast, and for playing the recovered audio signal using conventional means.

Figure 2 illustrates in more detail a transmitter 5 of the broadcast unit 3. The transmitter 5 comprises an input 6 for receiving the analogue audio signal. The received signal is provided to the input of an Analogue to Digital Converter (ADC) 7 which samples the analogue signal at some specified sampling rate and converts each sample into a corresponding n-bit binary value. The binary or digital signal is passed to a prediction unit 8 which codes the digital signal using a predictive coding algorithm. This algorithm will not be explained in detail here, and reference should be made to the brief description given above. What is relevant is that for each block of sampled digital audio signal, the unit 8 generates a set of prediction coefficients which allow a digital value of that block to be predicted from a series of preceding values, and a coded residual value which represents the difference between the predicted value and the actual value. The combination of the coefficients and the residual represents a sampled value in a significantly compressed form.

The output from the prediction unit 8 is provided to a processing block 9 which splits the digital signal into two channels or signals, a first of which 10 contains the prediction coefficients and a second of which 11 contains the residual coefficients. The split signals are provided to respective modulation blocks 12,13.

Considering firstly the modulation block 12, this receives the first split signal 10 carrying the prediction coefficients. The block 12 uses simple Amplitude Modulation (AM) to modulate the first digital signal onto a carrier signal 14 received from an oscillator 15, although it will be appreciated that other modulation schemes other than AM could be used. Error correction codes are also included in the AM signal. The amplitude modulated carrier signal is then provided to the second modulator 13 which receives the second split signal 11 carrying the coded residuals. This second split signal is modulated onto the AM carrier signal using a suitable phase modulation scheme which results in phase modulation of the amplitude modulated carrier signal. The modulated analogue signal 16 output by the second modulator 15 is provided to the input of a broadcast circuit 17 for broadcast over the air interface.

The person of skill in the art will realise that AM will provide only a very low data rate channel for carrying digital data, e.g. of the order of 2kbps. However, this channel is robust against noise and other interference. On the other hand, phase modulation will provide a higher data rate channel, e.g. of the order of 64kbps, which is less robust against noise and interference.

Figure 3 illustrates a receiver 18 provided in the hearing aid 4. The receiver 18 receives the broadcast signal and comprises an AM demodulation block 19 which comprises a phase locked loop (PLL) 20 and which recovers the amplitude modulation component of the signal. The recovery of this component provides the receiver with the prediction coefficients for each block of sample values of the audio signal, and proceeds robustly even at low channel signal to noise ratios (SNR). The PLL 20 also provides a stable clock signal to a block 21, where the clock signal serves as the sampling clock for a phase modulation demodulation process. The output of the demodulation block 21 is provided to a processor 22, together with the output of the AM demodulation block 19.

Assuming that recovery of the AM modulated signal is successful, the processor 22 examines the signal recovered from the phase modulation signal. If the phase modulation signal has not been recovered successfully (this may be detected using some form of error correction coding) due for example to the presence of noise, the processor 22 approximates the audio signal using only the adaptive prediction coefficients. The approximated audio signal is provided to a playback circuit 24. On the other hand, if the phase modulation signal has been recovered successfully, the processor combines the signal approximated with the adaptive prediction coefficients with the residual values to generate an improved approximated signal which is passed to the playback circuit 24. It may be appropriate to incorporate some delay in the processor (preferably prior to decompression to reduce the buffer requirements), e.g. 100ms for a 30m spacing between the speaker and the hearing aid, to synchronise the received direct sound with the audio playback.

The system described here has the advantage that as a user moves through the auditorium, or even moves outside, he or she will not necessarily have to suffer abrupt on/off switching of the signal. For example, whilst he or she remains in the auditorium, the quality may remain high, with both the modulation signals being recovered and processed. When he or she moves out of the auditorium, the signal might degrade to the lower quality as the phase modulation signal is lost. As he or she moves out of the building, the signal is lost completely. This represents a more natural hearing sensation.

The processor 22 receives at an input 25 a selection signal which allows the phase modulation demodulation process to be switched on and off either by user selection or as a result of an automatic control mechanism. Thus, if the selection is set to off, the processor will generate only the low quality audio signal using the adaptive prediction coefficients, even if the phase modulation signal is available for recovery. This allows battery power to be conserved. The general method of operation of the receiver of Figure 3 is illustrated in the flow diagram of Figure 4.

It may be appropriate in certain circumstances to broadcast the same information on multiple carrier frequencies to avoid local interference which affects only certain of the carriers. Information identifying the various available carriers, as well as synchronisation information (e.g. timestamps) to allow information on different carrier carries to be synchronised, may be carried on the low data rate AM signal multiplexed with the adaptive coding coefficients. The receiver circuit of the hearing aid is able to hop between bands to identify the carrier frequency offering the best reception.

The system described above makes use of a primary signal carrying the adaptive prediction coefficients and one secondary or further signal carrying residual values. The performance of the system may be enhanced by using a first secondary signal to carry residual values which are an approximation to the true residuals, and another secondary channel to carry residual values which represent the error between the approximated residuals and the actual residuals. Both secondary signals are carried by the carrier signal, but different modulation mechanisms are used to ensure that the first signal is more robust to noise and interference than the second, with the second secondary signal providing a higher data rate than the first (which in turn provides a higher data rate than the AM signal).

By way of example, the moderate data rate and high data rate channels may be carried by a two stage hierarchical DQAM coding applied directly to the carrier signal. Differential coding ensures freedom from low frequency sidebands (and some compensation can be applied at the transmission coding stage if the natural rejection is insufficient). The high data rate channel (e.g. 190kbps) is modulated onto individual quadruplet clusters of data points in the QAM constellation, and the moderate data rate channel (64kbps) is modulated onto the positions of these clusters in a QAM16 arrangement. The composite modulation form is thus QAM16 for the moderate data rate channel and QAM4 at QAM64 resolution level for the high data rate channel. It is noted that the secondary signals are arranged so that they do not produce low frequency modulation sidebands, allowing the low data rate AM modulation to be extracted by simple filtering of the downconverted signal.

The QAM demodulation process begins with an analogue to digital converter (ADC) which samples the down-converted signal. The sampling frequency is derived from the input PLL (whose jitter must therefore not unduly degrade the effective conversion dynamic range). The presence of the low data rate channel amplitude- modulating the entire signal is a complication and is resolved either by using the recovered AM modulation to drive a gain control circuit which removes low frequency gain variations in the channel, or by modulating the reference of the ADC so that constellation positions in the I and Q channels are sampled correctly regardless of the long-term through gain of the carrier channel. If both the moderate and the high data rate channel need to be recovered, the ADC is set to provide enough resolution, noise and settling time margin to cope with the whole constellation resolution. If only the moderate data rate channel is required, or if signal strength detectors and error rate detectors detect an uncorrectable error rate in the high data rate channel, the ADC performance can be significantly reduced, along with other elements of the front end processing, saving considerable power.

The transmitter is responsible for deciding what signals to broadcast. This may depend upon local spectrum and channel characteristics, or the particular usage environment. Signal information is broadcast on the low data rate signal. The intrinsic programmability of the channel data rate plan in this system allows this allocation with no involvement from the hearing aid wearer.

Although the present invention has been illustrated here with reference to a hearing aid for hearing impaired persons, it will be appreciated that its applications are not limited to such devices. Many other applications will be readily apparent to those of skill in the art.

Claims

1. A method of transmitting information from a transmitter to a receiver over a transmission link, the method comprising: decomposing said information into a first signal being an approximation to said information, and at least one further signal providing an approximation to the error between said first signal and said information; transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism; and receiving said first signal and optionally at least one further signal at the receiver, and recovering an approximation to said information.

2. A method according to claim 1, wherein said first transmission mechanism comprises the modulation of said first signal on a carrier signal, and said second transmission mechanism comprises the modulation of said second signal on that same carrier signal.

3. A method according to claim 1 or 2, wherein the first and second transmission mechanisms use different modulation schemes such that the first mechanism is more robust to the second mechanism.

4. A method according to any one of the preceding claims, wherein said transmission link is a wireless link, and said receiver and transmitter comprise wireless reception means and wireless transmission means respectively.

5. A method according to any one of the preceding claims, wherein said information is audio information.

6. A method according to any one of the preceding claims, wherein said step of decomposing the information comprises using an adaptive prediction algorithm, said first signal comprising for each sample of the information signal a set of prediction coefficients, whilst said at least one further signal comprises for each sample a residual value.

7. A method according to any one of the preceding claims, wherein said information is decomposed into a first signal being an approximation to said information, and at least two further signals, a first providing an approximation to the error between said first signal and said information and a second providing a residual signal which when added to said error improves the error approximation.

8. An information transmission system comprising a transmitter and a receiver for transmitting information over a communication link, the transmitter comprising processing means for decomposing said information into a first signal being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information, transmission means for transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and for transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism, and the receiver comprising receiving means for receiving said first signal and optionally at least one further signal at the receiver, and processing means for recovering an approximation to said information.

9. A transmitter for transmitting information over a transmission link, the transmitter comprising: processing means for decomposing said information into a first signal being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information, transmission means for transmitting said first signal from the transmitter to the receiver using a first transmission mechanism, and for transmitting said at least one further signal from the transmitter to the receiver using a second transmission mechanism, said first transmission mechanism being more robust to noise and interference than said second transmission mechanism.

10. A receiver for recovering information sent over a transmission link, the receiver comprising: means for receiving one or more of a set of signals sent over said transmission link, one of said set of signals being an approximation to said information, and at least one further signal being an approximation to the error between said first signal and said information; processing means for selectively defining a number of signals to be processed, and for combining the selected signal or signals to recover an approximation to the sent information.

11. A hearing aid comprising a receiver according to the fourth aspect of the invention, wherein the information sent over said link is audio information and said processing means is arranged to play the recovered approximate information to the user.