|Publication number||USRE41775 E1|
|Application number||US 11/345,693|
|Publication date||Sep 28, 2010|
|Filing date||Feb 1, 2006|
|Priority date||Jul 14, 2000|
|Also published as||DE60114856D1, DE60114856T2, EP1316240A1, EP1316240B1, US6839447, US20040037442, WO2002007479A1|
|Publication number||11345693, 345693, US RE41775 E1, US RE41775E1, US-E1-RE41775, USRE41775 E1, USRE41775E1|
|Inventors||Peter Ostergaard Nielsen, John L Melanson|
|Original Assignee||Gn Resound A/S|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (4), Classifications (11), Legal Events (1)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is a re-issue application of U.S. patent application Ser. No. 10/342,625, issued as U.S. Pat. No. 6,839,447, which is a continuation of International Application No. PCT/DK01/00493 filed on Jul. 13, 2001, which claims priority to and the benefit of Danish Patent Application No. PA 2000 01094 filed on Jul. 14, 2000.
The present invention relates to a binaural hearing system that comprises two fully or partly synchronously operating hearing prostheses capable of performing bi-directional data communication over a wireless communication channel. Fully synchronous operation between the hearing prostheses is preferably maintained by utilising direct sequence spread spectrum technology to lock all clock signals of a slave hearing prosthesis to a coding clock signal provided by a clock oscillator in the master hearing prosthesis during the bi-directional data communication. Thus, simultaneous sampling of respective microphone signals of the hearing prostheses is obtained so as to provide a wireless binaural hearing system that supports binaural signal processing techniques and algorithms.
Hearing aid systems with bi-directional communication capability are well known in the art. U.S. Pat. No. 5,991,419 discloses a so-called bilateral hearing instrument that comprises two units for placement in a hearing aid user's left and right ears, respectively. Each instrument comprises an associated transceiver circuit so as to provide bi-directional wireless communication between the instruments. WO 99/43185 discloses a resembling binaural digital hearing aid system adapted to exchange raw or processed digital signals between two hearing aids to allow each aid to perform a processing of its own input signal as well as a simulated processing of the processing performed in the other aid, i.e. the hearing aid that is arranged on the user's reverse side. The simulated processing of reverse side signals is performed to provide a binaural signal processing technique that can restore binaural sound perception by taking into account differences in hearing loss and compensation between the user's two ears. U.S. Pat. No. 5,751,820 discloses an integrated circuit design for bi-directional communication utilising reflective communication technology to obtain low power consumption, thereby making the design suitable for battery operated personal communication systems, such as binaural digital hearing aid systems.
However, while it has been noted in the above-mentioned prior art that a practical binaural hearing aid system must have control of the synchronisation between the ear units, and that U.S. Pat. No. 5,991,419 states that the phase error between the units should correspond to time errors less than 10 μS, there has not been disclosed an adequate wireless synchronisation technology that would actually be capable of providing the required synchronisation between the units or aids.
To perform correct binaural processing of the respective signals of such binaural hearing aid systems it is mandatory to assure that the individual hearing aids or instruments are operating synchronously with respect to each other. In particular, the respective microphone signals must be sampled substantially synchronously to enable e.g. binaural beamforming and off-axis noise cancellation. Time shifts as small as 20-30 μS between sampling instants of the respective microphone signals in the two hearing aids may lead to a perceivable shift in the beam direction. Furthermore, a slowly time varying time shift between the sampling instants of the respective microphone signals, which inevitably will occur if the hearing aids are operated asynchronously, will result in an acoustic beam that appears to drift and focus in alternating directions. An undesirable effect, which certainly will be very annoying for the hearing aid user.
Consequently, in order to provide a practical binaural hearing system it is highly desirable to provide a wireless communication technique that assures synchronised operation between the individual hearing prostheses and which, at the same time, is practical for miniature and low-power battery operated equipment such as hearing prostheses.
A first aspect of the invention relates to a binaural hearing system comprising a first and a second hearing prosthesis adapted for wireless bi-directional communication of digital data signals; the first hearing prosthesis comprises a first microphone adapted to generate a first input signal in response to receiving acoustic signals,
The second hearing prosthesis comprises a second microphone adapted to generate a second input signal in response to receiving acoustic signals,
According to the invention, the first clock generator operates as a master clock circuit for both hearing prostheses of the binaural hearing system during bi-directional communication of the first and second digital signals or data signals to ensure synchronous sampling of the respective microphone input signals. By locking the second clock and data retrieval means onto the received first modulated data signal, it is ensured that the retrieved clock signal and the second sampling clock signal in the second hearing prosthesis are synchronous to the coding clock signal generated by the first clock generator in the first hearing prosthesis. The microphone signal in the second hearing prosthesis is therefore sampled synchronously to the sampling of the microphone signal in the first hearing prosthesis. Thus, a binaural beam-forming algorithm, or other types of binaural processing algorithms, executed in the binaural hearing system are capable of correctly determine directions to acoustic sources by examining inter-device differences between the digital input signals, such as phase or group delay differences.
Frequencies of the synchronous coding and data rate clock signals may be selected to about 9600 kHz and 600 kHz, respectively. The coding clock signal is used to clock the first sequence generator and the data rate clock signal is preferably used to control a timing of the first data signal in order to synchronise the repetitive coding sequence to the first data signal. The first sampling clock signal is finally also derived synchronously to the coding clock signal (and therefore to the data rate clock signal) to allow the first or master clock generator to control the timing of the sampling of the first input signal. The sampling clock signal and the data rate clock signal may be derived from the coding clock signal by well-known clock division and/or multiplication methodologies e.g. using D-Flip-Flops, PLLs, etc.
The first and second analogue-to-digital converters are preferably both of an oversampled sigma-delta type with a sampling frequency of about 1 MHz, thus making it possible to avoid analogue lowpass filters to bandwidth limit the first and second input signals provided by the respective microphones before sampling. The first and second digital input signals may be represented by respective non-decimated, e.g. single bit format signals, or by corresponding decimated signals having a sampling rate in or close to the audio-frequeny range, e.g. about 16 kHz with a resolution of 1-20 bits such as 16 bits.
The first and second data signals, provided by the respective data generating means, may be constituted by the first and second digital input signals, respectively, so that substantially unprocessed or “raw” time discrete microphone input signals are transmitted to the other hearing prosthesis. In this situation, the data rate of each of the first and second data signals, during transmission, may be selected to about 512 Kbit/s. Such a data rate corresponds to representing each of the first and second data signals with a sequence of 16 bits samples at a sample rate of 16 kHz during bi-directional communication in a time-multiplexed mode with a transmission duty cycle of 50%.
Alternatively, the first and/or second data signal(s) may be pre-processed digital signals which has or have been derived by their respective data generating means that, for the purpose of processing the data signals, may comprise one or more DSPs. This pre-processing may modify audio characteristics of the digital input signals such, e.g. filtering and/or compressing one or several frequency bands of the respective data signals.
Preferably, the data generating means are adapted to encode their respective data signals prior to transmission in accordance with a predetermined error detection and/or correction scheme. The encoding allows data errors, typically caused by electromagnetic interference from other RF-sources, introduced into the data signals during transmission to be detected and/or corrected. The encoding may also be adapted to reduce the data rates of the data signals and/or to remove a DC content of the data signals. A large number of suitable encoding schemes has been disclosed in the relevant literature and will as such be well known to the skilled person. Accordingly, this issue will not be discussed further here. Finally, encoding of the first and/or second data signal may implemented by inserting control data in one or both of the data signals in order to communicate control data from the first to the second hearing prosthesis and/or vice versa. The control data may be utilised to support e.g. co-ordination in operation mode between the first and second prostheses, e.g. co-ordinate automatic or user controlled switching between a number of pre-set listening programs and/or between different audio input sources such microphone input, dual-microphone input, telecoil input, direct audio input etc.
The first and second sequence generators are preferably both adapted to generate respective versions of an identical pseudorandom noise (PN) sequence. The two PN sequences will be phase-aligned, and synchronous to the coding clock signal, when the second clock retrieval and generating means have locked onto the first modulated data signal. Sequence generators for generating PN sequences are particularly well suited for implementation in digital circuits where a number of low-power and die-area efficient implementations are possible. The modulation of the first and second data signals with their respective repetitive coding sequences can furthermore be implemented by simple sign encoding or modulation e.g. by switching the data signals to +1/−1 volt. Sign modulation is particularly convenient to implement in CMOS technology since CMOS transistors are relatively good switch elements. By applying the above-mentioned modulation scheme, the resulting modulation of the digital signals is commonly referred to as direct sequence spread spectrum modulation (DS-SS). Alternatively, the first and second sequence generators may be adapted to control respective frequency synthesisers controllable to transmit signals on anyone of a plurality of carrier frequencies. Values of the PN sequence is utilised to randomly select a particular carrier frequency of the plurality of carrier frequencies and thus modulate the data signal. Thereby, the repetitive coding sequences will comprise a carrier signal that hops between different carrier frequencies in a pseudorandom manner. This latter modulation scheme is commonly referred to as frequency hopped spread spectrum modulation (FH-SS).
In order to process the first and second input signals with advanced binaural signal processing algorithms, one of the first or second hearing prosthesis or both of them may comprise a Digital Signal Processor. Accordingly, the binaural hearing system may operate in either a symmetric or in an asymmetric mode. In the asymmetric mode, the data generating means of the first hearing prosthesis comprise a Digital Signal Processor(DSP) adapted to process the first digital input signal and the second data signal in accordance with a predetermined signal processing algorithm to provide the first processed data signal or vice versa if the DSP is located in the second hearing prosthesis. In this asymmetric mode, the DSP is preferably adapted to also generate a first or second data signal that has been binaurally processed and which therefore may be passed directly to the output means on the reverse side hearing prosthesis. Thereby, the asymmetric binaural hearing system may operate with a single DSP that processes the digital input signals from both hearing prostheses and generate binaurally processed data signals for both aids. Naturally, such an asymmetric binaural hearing system may contain DSPs in both hearing prostheses so that the asymmetric operation is obtained by programming one of the devices as a master device during the initial fitting of the binaural hearing system. The master device, in this situation, is programmed to execute the predetermined signal processing algorithm to generate and provide respective binaurally processed signals for both hearing prostheses. An advantageous property of this latter embodiment of the invention is that the hearing prostheses in a binaural pair can be identical units which may simplify the distribution and repair handling procedures.
In the symmetric operating mode, the data generating means of the first hearing prosthesis comprise a first Digital Signal Processor adapted to process the first digital input signal and the second data signal in accordance with a predetermined first signal processing algorithm to provide the first processed data signal to the first output means. The data generating means of the second hearing prosthesis comprise a second Digital Signal Processor adapted to process the second digital input signal and the first data signal in accordance with a predetermined second signal processing algorithm to provide the second processed data signal to the second output means.
According to a preferred embodiment of the invention, the first Digital Signal Processor and the first output means operate synchronously to the coding clock signal, and the second Digital Signal Processor and the second output means operate synchronously to the retrieved clock signal. Thereby, the acoustical or electrical output signals of the respective hearing prostheses are synchronised in time so as to provide a hearing system capable of delivering phase aligned acoustic or electrical output signals to the user's eardrums. All clock signals within the second hearing prosthesis are preferably locked to the retrieved clock signal (and thereby to the coding clock signal) while all clock signals within the first hearing prosthesis are synchronised to the coding clock signal. This embodiment of the invention provides a simple and efficient method of synchronising all clock signals within the entire binaural hearing system, i.e. also across the wireless communication channel. Such a completely synchronised hearing system supports binaural processing algorithms that are capable of retaining naturally occurring binaural signal cues, such as interaural phase and level differences, in the acoustic or electrical output signals provided to the user.
For some applications of the present binaural hearing system, it may be advantageous to make the second hearing prosthesis capable of operating as a stand-alone device, independently of whether or not the first hearing prosthesis transmits the first modulated data signal. This has been accomplished by a binaural hearing system wherein the second hearing prosthesis comprises a second clock oscillator adapted to generate a second coding clock signal and the second sampling clock signal. The second hearing prosthesis further comprises clock mode selection means operatively connected to the second clock and data retrieval means and the second clock oscillator and adapted to selectively use the second clock and data retrieval means or the second clock oscillator as a source for clock signals in the second hearing prosthesis. Thereby, a mono-aural operation mode is supported by both hearing prostheses during time periods with interruptions in the first modulated data signal.
According to this embodiment of the invention, the second hearing prosthesis is adapted to automatically operate in mono-aural mode if the clock mode selection means detect that the first modulated data signal and/or the first data signal is/are absent or contain(s) too many errors to be used.
Since it may be impractical to sell and distribute binaural hearing systems where only one of the pair of hearing prostheses is capable of operating as a master device during bi-directional communication, a preferred embodiment of the invention is one wherein the first hearing prosthesis further comprises first clock and data retrieval means allowing the prosthesis to lock onto the second modulated data signal to synchronise clock signals of the first prosthesis to the second clock oscillator. In such a binaural hearing system, operation as a master device is supported for both the first and the second hearing prosthesis. In a particularly preferred embodiment of the invention, the selection of which of the hearing prostheses that should operate as the master(and the other as a slave device) during binaural operation can be selected during the initial fitting session by programming the devices from a fitting system. Each of the hearing prostheses comprises a programming interface for exchanging programming data between a host programming system and the hearing prosthesis, and a configuration register programmable through the programming interface and operatively connected to the clock mode selection means to control their operation.
According to yet another embodiment of the invention, the first and second modulated data signals are transmitted by their respective wireless transceivers without having any further RF modulation applied than the modulation provided by the repetitive coding sequence. This embodiment of the invention has as a particularly attractive feature that commonly employed RF modulators and demodulators can be dispensed with to minimise current and area consumption and reduce design complexity of the first and second wireless transceivers.
However, for other applications it may be more effective, in particular in terms of minimising power consumption, to include within the first wireless transceiver a first RF modulator adapted to further modulate the first modulated data signal to generate and transmit a first RF modulated data signal to the second hearing prosthesis and a first RF demodulator adapted to recover the second modulated data signal from a second RF modulated data signal. The second wireless transceiver further comprises a second RF modulator adapted to further modulate the second modulated data signal to generate and transmit the second RF modulated data signal to the first hearing prosthesis and a second RF demodulator adapted to recover the first modulated data signal from the first RF modulated data signal from the first wireless transceiver. This embodiment may be more power efficient than the direct transmission of the first and second modulated data signals since a carrier frequency of the RF modulators may be selected so as to provide an optimum match to a particular type of transmission/reception antennas. Accordingly, in the present specification and claims the term “modulated data signal” may designate a data or digital signal which solely has been modulated with the coding sequence prior to transmission. Or the term may designate a data signal that has been modulated with the coding sequence to form a composite signal and thereafter further modulated or up-converted with a RF carrier signal so as to provide e.g. a FSK modulated RF composite signal.
The first and second wireless transceivers must comprise some form of antenna means to transmit/receive the modulated data signals. For hearing aid applications, it may be difficult to provide sufficient housing space for an effective RF antenna. This is particularly true if it is desired to transmit the modulated data signals in the RF range below about 1 GHz due to relatively large wavelengths, in comparison to typical dimensions of hearing aids, of such RF signals.
According to an embodiment of the invention, each of the first and second wireless transceivers comprises an inductive coil where the inductive coils are adapted to transmit and receive the modulated data signals, or the RF modulated data signals, by utilising near-field magnetic coupling between said inductive coils. Each of the inductive coils may be tuned to a target transmission frequency by arranging a suitable tuning capacitor across the coil so as to provide a Q for each of the inductive antennas of about 4, preferably between 3 and 10 to optimise the received/transmitted power at the antennas. The communication frequency is preferably selected to a frequency somewhere between 50-100 MHz for such a magnetically coupled system.
The above-described binaural hearing system is adapted to communicate bi-directional data signals to support binaural signal processing algorithms and thereby allow the hearing system to restore or enhance binaural signal cues in the acoustic input signals.
However, it may also be advantageous to provide a hearing aid system where spread spectrum techniques are employed for the purpose of synchronising the signal processing between the hearing aids to secure e.g. identical sampling frequencies between the aids. A signal delay or group delay through a DSP based hearing prostheses is commonly dominated by a group delay associated with the digital processing of the input signal. This group delay is furthermore substantially proportional to the inverse of each individual hearing prosthesis' own master clock frequency. Since a common tolerance on the latter value is about +/−5-10%, the group delay difference between two randomly selected hearing prostheses may be quite large. Consider a case where a particular hearing prosthesis has a nominal group delay value of 5 ms. Individual prostheses of the same type may exhibit a group delay anywhere from 4.5 ms to 5.5 ms. The group delay difference between these values is more than the maximum interaural time delay of 600-700 μS that occurs in natural, i.e. unaided, human hearing. By providing matching of the signal delays through the hearing prostheses, binaural signal cues in the input acoustic signals can better be preserved.
A second aspect of the invention therefore relates to a wireless synchronised hearing aid system comprising a first and a second hearing prosthesis wherein the first hearing prosthesis comprises:
According to this second aspect of the invention, spread spectrum technology is employed to synchronise the signal processing of the hearing prostheses by the transmitted synchronisation signal and based on the repetitive coding sequence. By not transmitting bi-directional data signals during operation, power consumption within the wireless transceivers may be significantly reduced in both hearing aids.
The first DSP may, furthermore, be adapted to generate a digital control data signal for controlling an operation mode of the second hearing prosthesis and the first wireless transmitter may be adapted to modulate the digital control data with the repetitive coding sequence and use the digital control data as the synchronisation signal. The control data are thus modulated with the repetitive coding sequence and transmitted to the second hearing prosthesis where they are retrieved in a manner corresponding to the retrieval of the first and second data signals described in connection with the first aspect of the invention.
The repetitive coding sequence provided by the first and second sequence generators of the binaural hearing system or by the sequence generators of the synchronised hearing aid system may comprise, or be constituted by, a pseudo-random noise (PN) sequence. Alternatively, each sequence generator may be adapted to select a carrier frequency provided by a frequency synthesiser based on values of a pseudorandom noise (PN) sequence to generate a frequency-hopped repetitive coding sequence.
In the following, a specific embodiment of a DSP based hearing aid system according to the invention is described and discussed in greater detail. The present description discusses in detail only a wireless DS-SS bi-directional communication system and its utilisation to synchronise corresponding clock signals between two individual hearing aids of the system.
To support low power and low voltage operation of the present wireless DS-SS communication system and associated DSPs, logic gates and other digital circuits are preferably implemented in a low threshold voltage CMOS process. Preferred processes are 0.5-0.18 μm CMOS processes with threshold voltages located in the range from about 0.5 to 0.8 Volt.
In the overall system diagram of the binaural hearing aid system shown on
In the simplified block diagram of
In the second hearing prosthesis, a second antenna 30 receives the composite RF signal transmitted by the first hearing aid. A RF demodulator 35 downconverts the received composite RF signal to a baseband frequency range and extracts the first modulated data signal. Thereafter, a clock and data retrieval and generating circuit 40 multiplies the first modulated data signal with an synchronous version of that PN code that was used to encode the first data signal in the first hearing aid.
Since the product of two versions of a predetermined repetitive pseudorandom noise sequence or PN sequence is one only if the two versions are exactly in phase, the clock and data retrieval and generating circuit 40, within the second hearing aid, is able to acquire and maintain lock to the transmitter by continuously evaluating an autocorrelation function between the two versions of the PN code and adjust a relative phase between the PN sequences to obtain a maximum correlation value. This issue will be addressed further in connection with the description of
In an alternative embodiment of the above-described integrated DS-SS transceiver system, the (traditional) RF modulator 15 and demodulator 35 circuits have been designed to operate at communication frequency which is very low compared to typical RF communication frequencies, e.g. lower than the above-mentioned 200 MHz-1 GHz RF communication frequency range. Such a low RF carrier frequency may be as low as only about 4-8 times higher than the chipped-rate of the modulated data signals, to further save power and reduce complexity of the transceivers. RF antennas 20 and 30 has also been replaced by respective inductive coils adapted to communicate the first and second data signals between the first and second hearing aids by utilising near-field magnetic coupling between the inductive coils. The requirement to transmission distance of a binaural hearing aid system is in the order of 15-25 cm. The above-described wireless magnetic coupling technique is practical because of the short transmission distance. Furthermore, magnetically coupled system, has as another attractive, a limited far-field emission of electro-magnetic signals compared to the emission of traditional far-field coupled system which are obtained at higher communication frequencies and communicated over antennas designed to operate at such higher communication frequencies.
Consequently, instead of using traditional antennas, it may prove more power efficient to transfer the digital data signals for hearings aid applications, and other very short-range applications, by way of magnetic induction. Crucial issues are that the distance between the hearing aids is not much larger than physical dimensions of the coils, and that the physical dimensions of the coils are very small (at least about 10 times smaller) compared to the wavelength of the RF carrier. Under such conditions, the transmitter power required to transmit a desired bandwidth and at a sufficiently low bit error rate (BER) may be transferred by near-field magnetic coupling, or mutual induction, while at the same time minimising far-field coupling. Minimising the far-field coupling helps improving the interference immunity and compliance to EMC regulations in general.
The first and second data signals may be coded versions of digital audio signals processed within the respective hearing aids, such as coded versions of the first and second digital input signals obtained from the respective microphone signals. The first and second data signals may also be constituted by digital signals that has been processed by the DSPs or the first and second data signals may represent unencoded digital input signals. The coding may be provided to support error detection and/or correction of the received digital signals according to a number of methods well known in the art, e.g. Reed Solomon coding. Encoding may further be applied for the purpose of removing any DC content of the digital signals prior to their transmission in order to simply the design of the receiving part of the transceivers. Finally, the coding of the digital data signals may comprise the step of inserting control data or information into the first and/or second data signal(s) and extract these control data at the receiving side to communicate control information between the hearing aids.
The transmission frequency for the present near-field magnetically coupled communication system is preferably selected in the range 50-100 MHz and each inductive coil may have a inductance of between 200 nH and 2 μH. The data or symbol rate of each of the first and second data signals is preferably about 600 Kbit/s in order to support an audio rate of about 256 Kbit/s of each of the first and second data signals in combination with an effective transmission duty cycle of about 50% plus overhead data for a forward error correction scheme. Accordingly, if these 600 Kbit/s first and second data signals are modulated with 16 codes of the PN code sequence per data bit, the resulting chip rate of each of the modulated data signals will be about 9600 Kbit/s. If an even higher transmission frequency is desired, further RF modulation or up-conversion may be applied to the “chipped” modulated data signal in order to further raise its transmission frequency to a desired, or target, range, as explained above. For the near-field magnetic coupled communication system, the further RF carrier frequency is preferably selected to be only about 4-8 times higher than the chipped rate of the modulated data signals. An important advantage of operating the integrated DS-SS transceiver system by near-field magnetic coupling is that it may be possible to reduce the required transmission power to a level that is below RF spurious emission requirements according to national and/or international EMC norms. These spurious emission requirements are in practice measured in the far-field of the device under consideration.
However, a near-field magnetic coupled communication system is capable of coupling more of the transmitter's emitted electromagnetic power to the receiving antenna than a corresponding traditional RF based communication system is capable of for any fixed level of far-field electromagnetic power. Consequently, for the purpose of suppressing RF spurious emission power from the transceivers, as measured in the far-field, the near-filed magnetic coupled system has superior characteristics.
According to the European EMC norm EN55022 all radio transmitting devices must have an emitted spurious power density of less than −54 dBm in most of the frequency range below 230 MHz and below −54 dBm from 230 MHz-1 GHz. Consequently, if the emitted power density of the integrated DS-SS transceiver system is kept below −54 dBm everywhere in the 0 Hz-1 GHz transmission frequency band, the transceiver system will be able to meet these requirements.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5434924 *||Mar 6, 1991||Jul 18, 1995||Jay Management Trust||Hearing aid employing adjustment of the intensity and the arrival time of sound by electronic or acoustic, passive devices to improve interaural perceptual balance and binaural processing|
|US5751820 *||Apr 2, 1997||May 12, 1998||Resound Corporation||Integrated circuit design for a personal use wireless communication system utilizing reflection|
|US5991419 *||Apr 29, 1997||Nov 23, 1999||Beltone Electronics Corporation||Bilateral signal processing prosthesis|
|US6549633 *||Feb 18, 1998||Apr 15, 2003||Widex A/S||Binaural digital hearing aid system|
|U.S. Classification||381/312, 381/315|
|International Classification||H04R5/033, H04S1/00, H04L27/10, H04L7/00, H04R25/00|
|Cooperative Classification||H04R25/552, H04R25/554|
|European Classification||H04R25/55D, H04R25/55B|