HK1059694A - A modulation technique providing high data rate through band limited channels - Google Patents

Description

Modulation techniques for providing high data rates through band limited channels

Technical Field

The present invention relates to a modulation technique that provides high data rates through band limited channels.

Background

It is always desirable to provide data with higher data rates over channels of limited bandwidth. Many modulation techniques have been developed for increasing the data rate over one channel. For example, M-ary Phase Shift Keying (PSK) and Quadrature Amplitude Modulation (QAM) techniques allow data to be compressed by encoding multiple data bits in each transmit symbol. These systems have associated limitations. First, the hardware associated with these systems is expensive. This is because these techniques require a high level of channel linearity for proper operation. Accordingly, a large amount of signal processing for carrier tracking, symbol recovery, interpolation and signal shaping must be performed. Second, these techniques are sensitive to multipath effects. These effects need to be compensated for in the receiver. Third, these systems typically require a bandwidth that is greater than the bandwidth available in some applications (e.g., in-band channel broadcast FM subcarrier service) in order to achieve the desired data rate.

Disclosure of Invention

In accordance with the principles of the present invention, a digital data modulator is coupled to a source of digital data signals. The encoder encodes the digital data using a variable pulse width code. The pulse signal generator generates pulses representing edges of the encoded digital data signal. The carrier signal generator generates a carrier signal whose carrier pulses correspond to the pulses from the pulse signal generator. A corresponding digital data demodulator is connected to a source of a modulated signal having carrier pulses spaced from each other to represent a variable pulse width encoded digital data signal. The detector generates a variable pulse width encoded signal in response to receiving the carrier pulse. The decoder decodes the variable pulse width encoded signal to produce a digital data signal.

Techniques in accordance with the principles of the present invention may be implemented using relatively inexpensive circuitry that is insensitive to multipath interference and provides substantial bandwidth compression.

Drawings

In the drawings:

FIG. 1 is a block diagram of a modulator of the present invention;

FIG. 2 is a waveform diagram that helps to explain the operation of the modulator shown in FIG. 1;

FIG. 3 is a block diagram of a receiver that may receive a signal modulated in accordance with the present invention;

FIG. 4 is a spectral diagram that facilitates an understanding of an application of the modulation technique shown in FIGS. 1 and 2;

FIG. 5 is a block diagram of an FM broadcast transmitter incorporating an in-band channel digital transmission channel implemented using the modulation technique of the present invention;

fig. 6 is a block diagram of an FM broadcast receiver that may receive signals modulated by the FM broadcast transmitter shown in fig. 5;

FIG. 7 is a waveform diagram useful in understanding the operation of a modulator according to the principles of the present invention;

FIG. 8 is a block diagram of another embodiment of the present invention;

fig. 9 is a block diagram of a receiver that may receive signals generated by the system shown in fig. 8.

Detailed Description

Fig. 1 is a block diagram of a modulator for use in the present invention. IN fig. 1, an input terminal IN receives a digital signal. Input IN is coupled to an input of encoder 10. An output of the encoder 10 is coupled to an input of a differentiator 20. An output of the differentiator 20 is coupled to an input of a level detector 25. An output of level detector 25 is coupled to a first input of mixer 30. A local oscillator 40 is coupled to a second input of the mixer 30. An output of mixer 30 is coupled to an input of a Band Pass Filter (BPF) 50. An output terminal of the BPF50 is coupled to an output terminal OUT, the BPF generating a modulated signal representing a digital signal at an input terminal IN.

Fig. 2 is a waveform diagram useful in understanding the operation of the modulator of the embodiment of fig. 1. Fig. 2 is not drawn to scale in order to more clearly illustrate the waveforms. IN the exemplary embodiment, the digital signal at input IN is a bi-level signal IN non-return-to-zero format (NRZ). This signal is illustrated as the top waveform in fig. 2. The NRZ signal carries successive bits, each lasting a predetermined period, called bit period, which is shown by a dashed line in the NRZ signal and has a corresponding frequency, called bit rate. The levels of the NRZ signal all represent the bit values in a known manner. The encoder 10 uses a variable pulse width code to manipulate the encoded NRZ signal. In the illustrated embodiment, the variable pulse width code is a variable aperture code (variable aperture code). Variable pore coding is described in detail in International patent application PCT/US99/05301 by Chandra Mohan, filed on 11/3 1999. In this patent application, the NRZ signal is phase encoded in the following way.

Each bit period in the NRZ signal is encoded as a transition in the encoded signal. An encoding clock running at a multiple M of the bit rate is used to phase encode the NRZ signal. In the above-mentioned patent application, the encoding clock runs at a rate M that is nine times the bit rate. When the NRZ signal transitions from a logic "1" level to a logic "0" level, a transition is made in the eight code clock cycles (M-1) of the code signal by the previous transition. When the NRZ signal transitions from a logic "0" level to a logic "1" level, a transition occurs in the encoding clock period (M +1) of the encoding signal 10 from the previous transition. When the NRZ signal has no transition, i.e. if consecutive bits have the same value, a transition is made in the encoded signal nine encoding clock cycles (M) from the last transition. The variable aperture coded signal (VAC) is illustrated as the second waveform in fig. 2.

The variable aperture coded signal (VAC) is differentiated by a differentiator 20 to produce a series of pulses that are time aligned with transitions of the VAC signal. The differentiator also produces a 90 degree phase shift of the VAC modulated signal. The rising edge transition of the pulse produces a positive going pulse and the falling edge transition produces a negative going pulse, all in a known manner. Differential VAC signalIllustrated as the third signal in fig. 2.The signal is level-detected by a level detector 25 to produce a series of three-level pulses with successive amplitudes. When differentiating the VAC signalIs greater than the positive threshold value,generating a level signal having a high value; when differentiating the VAC signalIs less than the negative threshold value, a level signal is generated having a low value, otherwise the generated level signal has an intermediate value, all generated in a known manner. The level signal is exemplified as the fourth signal (level) in fig. 2.

The level signal modulates a carrier signal from a local oscillator 40 in a mixer 30. The positive pulse generates a carrier signal pulse having a first phase and the negative pulse generates a carrier signal pulse having a second phase. The first and second phases are preferably substantially 180 degrees out of phase. The carrier signal pulses are preferably substantially one code clock cycle in length, and in the exemplary embodiment the carrier signal has substantially the duration of NRZ bit period 1/9. The signal frequency of the local oscillator 40 is selected so that preferably at least 10 periods of the local oscillator signal can be generated in the carrier signal pulse time period. In fig. 2, the carrier signal CARR is illustrated as a bottommost waveform, wherein the carrier signals are represented by vertical hatching within the respective rectangular envelope. In the CARR signal illustrated in fig. 2, the phase of the carrier pulse generated corresponding to a positive-going level pulse is represented by a "+" and the phase of the carrier pulse generated corresponding to a negative-going level pulse is represented by a "-". "+" and "-" merely represent substantially 180 degrees phase difference and they are not used to represent any absolute phase.

The BPF50 filters all the "out-of-band" fourier components in the CARR signal, as well as the carrier component itself and one sideband, leaving only a single sideband signal. Thus, the output signal OUT of the BPF50 is a Single Sideband (SSB) phase or frequency modulated signal representing the NRZ data signal at the input IN. The signal may be transmitted to the receiver by any known transmission technique.

Fig. 3 is a block diagram of a receiver that may receive the modulated signal shown in fig. 1. IN fig. 3, the input terminal IN is coupled to the source of the modulated signal described above with reference to fig. 1 and 2. The input IN is coupled to an input of the BPF 110. The output of the BPF110 is coupled to the input of the integrator 120. The output of the integrator 120 is coupled to the input of the limiting amplifier 130. An output of the limiting amplifier 130 is coupled to an input of a detector 140. An output of the detector 140 is coupled to an input of a decoder 150. An output of the decoder 150 reproduces the NRZ signal represented by the modulated signal at the input IN and is coupled to an output OUT.

In operation, the BPF110 filters out-of-band signals, passing only the modulated SSB signals. Integrator 120 inverts the 90 degree phase shift introduced by differentiator 20 (of fig. 1). The limiting amplifier 130 limits the amplitude of the signal from the integrator 120 to a constant amplitude. The signal from the limiting amplifier 130 corresponds to the carrier pulse signal CARR illustrated in fig. 2. The detector 140 is either an FM discriminator or a Phase Locked Loop (PLL) for demodulating the FM or PM modulated carrier pulse signal, respectively. Detector 140 detects the carrier pulses and generates a bi-level signal having transitions represented by the phase and timing of the pulses. The output of the detector 140 corresponds to the variable bit width signal of the VAC signal in fig. 2. The decoder 150 performs the inverse operation of the encoder 10 (of fig. 1) and generates an NRZ signal corresponding to the NRZ signal of fig. 2 at an output terminal OUT. The above-mentioned us patent application (RCA 88,945) describes a decoder 150 for use in fig. 3. At the output OUT, the NRZ signal is then processed by using a loop (not shown).

Since the carrier pulses (signal CARR in fig. 2) are generated at precisely defined times with respect to each other and the duration of these pulses is limited, it is possible to enable the detector 140 only at the desired times of the pulses. For example, in the exemplary embodiment, each pulse has substantially the duration of time 1/9 between the number of transitions of the NRZ signal, as described in detail above. After 8/9 of the time between the receipt of a carrier pulse into a transition of the NRZ signal, successive pulses are expected to be generated only at 9/9 (no transition) or 10/9 (rising edge) of the time between the transition of the NRZ signal from that pulse, due to the preceding carrier pulse (representing the falling edge). Similarly, at 10/9 the time between the receipt of a carrier pulse into a transition of the NRZ signal, a successive pulse is expected to occur only at 8/9 (falling edge) or 9/9 (no transition) the time between the transition of the NRZ signal from that pulse, due to the preceding carrier pulse (representing a rising edge). The detector 140 only needs to be enabled when a carrier pulse is desired and only in the temporal neighborhood of the desired pulse duration.

The window timer, illustrated as 160 in fig. 3, has an input coupled to the state output of detector 140 and has an output coupled to the enable input of detector 140. The window timer 160 monitors the signal from the detector 140 and enables the detector only when a carrier pulse is expected and only in the time domain of pulse duration described above.

In an exemplary embodiment, the energy of the modulated signal is primarily between 0.44(8/18) and 0.55(10/18) times the bit rate, and thus has a bandwidth of 0.11 times the bit rate. This will result in a nine-fold increase in data rate in bandwidth. Other compression ratios are readily achieved by varying the ratio of encoding clock to bit rate, and the trade-off selection and limitation will be readily appreciated by those skilled in the art.

The system described above can be implemented in both the transmitter and receiver by a less complex loop than M-ary PSK or QAM modulation techniques. More specifically, in the receiver, after extracting the modulated signal, a limiting amplifier (e.g., 130) may be used, which has less expense and saves energy. Also, both encoding and decoding of NRZ signals may nominally be performed by fast Programmable Logic Devices (PLDs). These devices are relatively inexpensive (currently $ 1 to $ 2). In addition, there is no intersymbol interference in this system, so no waveform shaping is required. Furthermore, no tracking loop is required other than the clock recovery loop.

As described above, since carrier transmission occurs only on bit boundaries and is not continuous throughout a bit period, a time window can be used in the receiver to detect received carrier pulses only as many times as desired. Thus, there is no multipath problem in the current system.

One application of the above modulation technique is the transmission of CD quality digital music simultaneously with FM mono and stereo broadcast audio signals. Fig. 4 is a spectral diagram that facilitates an understanding of the application of the modulation technique shown in fig. 1 and 2. Fig. 4a shows the power envelope of a U.S. FM broadcast signal. In fig. 4a, the horizontal lines represent frequencies and represent some VHF band portion between about 88MHz and about 107 MHz. The vertical direction represents the signal strength. The figure shows the allowed spectral envelopes of two adjacent broadcast signals. Each carrier is shown as a vertical arrow. Surrounding each carrier is a sideband carrying FM modulated broadcast signals on the carrier.

In the united states, FM broadcasters may broadcast mono and stereo audio on sidebands at full power in the 100kHz carrier range. In fig. 4a, these sidebands are illustrated without shading. The broadcaster can broadcast other information in a 100kHz to 200kHz sideband, but the transmit power in this sideband must be 30dB below full power. These sidebands are illustrated with shading. Neighboring stations (in the same geographical area) must be separated by at least 400 kHz.

The uppermost sideband of the lower frequency broadcast signal carrier in fig. 4a is illustrated in the lower spectral diagram of fig. 4 b. In fig. 4b, the vertical direction represents the modulation percentage. In fig. 4b, the mono audio signal L + R audio signal is transmitted in the 0 to 15kHz sidebands with a modulation level of 90%. The L-R audio signal is transmitted as a double sideband suppressed carrier signal around the suppressed subcarrier 38kHz frequency at a 45% modulation level. The lower sideband (1sb) is in the range of 23kHz to 38kHz and the upper sideband (usb) is in the range of 38kHz to 53 kHz. A 19kHz pilot tone (half the suppressed carrier frequency) is also contained in the sidebands near the main carrier. Thus, 47kHz is reserved in the upper (fig. 4b) and lower (not shown) sidebands near the main carrier (i.e., 53kHz to 100kHz) for the broadcaster to broadcast other information at full power. As mentioned above, the power transmitted from 100kHz to 200kHz must be 30dB below full power.

Using the modulation techniques illustrated in fig. 1 and 2 described above, a 128 kilobits per second (kbps) signal, which contains an MP3CD quality audio signal, may be transmitted within a bandwidth of less than 20 kHz. As illustrated in fig. 4b, the digital audio signal may be placed in the space between the upper sidebands (e.g.) 53kHz and 100kHz and transmitted as a subcarrier signal along with the normal broadcast stereo audio signal. In fig. 4b, the digital audio signal is the SSB signal described above centered around 70kHz, and the range of the digital audio signal is approximately from 60kHz to 80 kHz. This is in the 100kHz range of the main carrier and can therefore be transmitted at full power. Such signals are referred to as in-band channel (IBOC) signals.

Fig. 5 is a block diagram of an FM broadcast transmitter incorporating an in-band channel digital transmission channel implemented in accordance with the modulation techniques described above with reference to fig. 1 through 3. In fig. 5, these elements, which are identical to the elements shown in fig. 1, are enclosed by a dashed rectangle labeled "fig. 1", are assigned the same reference numerals and are not described in detail below. Encoder 10, differentiator 20, mixer 30, oscillator 40 and BPF50 combine to produce an SSB phase or frequency modulated signal (CARR of fig. 2) representing a digital input signal (NRZ of fig. 2), all as described above with reference to fig. 1. The output of BPF50 is coupled to the input of amplifier 60. An output of the amplifier 60 is coupled to a first input of a second mixer 70. A second oscillator 80 is coupled to a second input of the second mixer 70. The output of the second mixer 70 is coupled to the input of the first filter/amplifier 260. An output of first filter/amplifier 260 is coupled to a first input of signal combiner 250.

An output of the broadcast baseband signal processor 210 is coupled to a first input of a third mixer 220. The third oscillator 230 is coupled to a second input of the third mixer 220. The output of the third mixer 220 is coupled to the input of a second filter/amplifier 240. An output of the second filter/amplifier 240 is coupled to a second input of the signal combiner 250. An output of the signal combiner 250 is coupled to an input of a power amplifier 270, wherein the power amplifier 270 is coupled to a transmit antenna 280.

In operation, the encoder 10 receives a digital signal representing a digital audio signal. In a preferred embodiment, the signal is an MP3 compliant digital audio signal. More specifically, the digital audio data stream is Forward Error Correction (FEC) encoded using a Reed-Solomon (RS) code. The FEC encoded data stream is then packetized. This packet data is then compressed by the loop shown in fig. 1 into SSB signals as described in detail above.

The frequency of the signal generated by oscillator 40 is selected to be 10.7MHz so that the digital information from encoder 10 is modulated to a center frequency of 10.7 MHz. The modulation frequency may be any frequency, but a more specific selection of the frequency may correspond to the frequency of an existing low-cost BPF filter. For example, typical BPF filters have center frequencies of 6MHz, 10.7MHz, 21.4MHz, 70MHz, 140MHz, and so on. In an exemplary embodiment, 10.7MHz is selected as the modulation frequency, and BPF50 is applied as an existing 10.7MHz filter. The BPF50 filtered SSB signal is amplified by amplifier 60 and combined by second mixer 70 and second oscillator 80 for up-conversion. In the exemplary embodiment, second oscillator 80 generates a 77.57MHz signal and upconverts the SSB to 88.27 MHz. The signal is filtered and further amplified by a first filter/amplifier 260.

The broadcast baseband signal processor 210 receives a stereo audio signal (not shown) in a known manner and performs the signal processing necessary to form a baseband composite stereo signal comprising an L + R signal at baseband, a double sideband suppressed carrier L-R signal at a carrier frequency of 38kHz and a pilot tone of 19 kHz. The signal is then modulated onto a carrier signal designated as the FM station frequency. The third oscillator 230 generates a carrier signal designated as a broadcast frequency, which in the preferred embodiment is 88.2 MHz. The third mixer 220 generates a modulated signal modulated with the composite mono and stereo audio signals as illustrated in fig. 4 b. The modulated signal at the carrier frequency 88.2MHz, with standard broadcast audio sidebands as illustrated in fig. 4b, is then filtered and amplified by a second filter/amplifier 240. This signal is combined with the SSB modulated digital signal from the first filter/amplifier 260 to form a composite signal. As shown in fig. 4b, the composite signal comprises a standard broadcast stereo audio sideband modulated at 88.2MHz on a carrier wave, and an SSB modulated signal carrying a digital audio signal at a center frequency of 70kHz on a carrier wave (88.27 MHz). The composite signal is then power amplified by a power amplifier 270 and provided to a transmit antenna 280 for transmission to an FM broadcast receiver.

Fig. 6 is a block diagram of an FM broadcast receiver that may receive signals modulated by the FM broadcast transmitter of fig. 5. In fig. 6, these elements, which are identical to those illustrated in fig. 3, are indicated by dashed rectangles labeled with the letters of fig. 3 and designated by the same reference numerals, and will not be described in detail below. In fig. 6, a receive antenna 302 is coupled to an RF amplifier 304. An output of the RF amplifier 304 is coupled to a first input of a first mixer 306. An output of the first oscillator 308 is coupled to a second input of the first mixer 306. The output of the first mixer 306 is coupled to respective inputs of the BPF310 and the tunable BPF 110. The output of the BPF310 is coupled to the input of an Intermediate Frequency (IF) amplifier 312, which may be a limiting amplifier. An output of Intermediate Frequency (IF) amplifier 312 is coupled to an input of FM detector 314. An output of the FM detector 314 is coupled to an input of an FM stereo decoder 316.

In operation, the RF amplifier 304 receives and amplifies RF signals from the receive antenna 304. The first oscillator 308 generates a signal having a frequency of 98.9 MHz. The combination of the first oscillator 308 and the first mixer 306 operate to down-convert the 88.2MHz main carrier signal to 10.7MHz and the SSB digital audio signal from 88.27MHz to 10.63 MHz. The BPF310 passes only FM stereo sidebands (L + R and L-R) around 10.7MHz in a known manner. The IF amplifier 312 amplifies the signal and provides it to an FM detector 314 that produces a baseband composite stereo signal. The FM stereo decoder 316 decodes the baseband composite stereo signal in a known manner to produce a mono and/or stereo audio signal (not shown) representative of the transmit audio signal.

In an exemplary embodiment, the tunable BPF110 is tuned to a center frequency of 10.63MHz and passes only digital audio signals around that frequency. In an exemplary embodiment, the passband of the BPF110 ranges from 10.53MHz to 10.73 MHz. The BPF110, integrator 120, limiting amplifier 130, detector 140, decoder 150 and window timer 160 operate in combination to extract the modulated digital audio signal and demodulate and decode the signal to reproduce the digital audio signal in the manner described above with reference to fig. 3. The digital audio signal from decoder 150 is processed in a suitable manner by a further loop (not shown) to produce an audio signal corresponding to the transmitted digital audio signal. More specifically, the signals are grouped and any errors introduced during transmission are detected and corrected. The corrected bitstream is then converted into a stereo audio signal, all in a known manner.

The above-described embodiments provide equivalent compression performance for 1024QAM systems. However, in practice, QAM systems are limited to around 256QAM because of the difficulty in resolving noise and multipath inter-symbol interference due to tight constellation space. The above system has no ISI problem due to the narrow and wide distance carrier pulses. In short, higher data rates can be transmitted in narrower bandwidth channels that do not have the problems associated with other techniques, such as QAM.

Referring back to fig. 2, in the CARR signal, it may appear that the gaps between carrier pulses are relatively wide, and no carrier signal is transmitted in the gaps. These gaps may be used in alternative embodiments of the present invention. Fig. 7 is a more detailed waveform diagram useful for understanding the CARR signal of modulator operation according to this alternative embodiment. As described above, in the encoder illustrated in fig. 1, the encoded clock signal has a period of NRZ signal bit period 1/9. The dashed vertical lines in fig. 7 represent the encoding clock signal periods. The allowed time positions of the carrier pulses are indicated by dashed rectangles. The carrier pulse may be generated 8,9 or 10 clock pulses after the previous pulse. Thus, the carrier pulse may be generated in any one of three adjacent clock cycles. Carrier pulse a is assumed to be an 8 clock pulse from the previous pulse, carrier pulse B is assumed to be a 9 clock pulse from the previous pulse, and carrier pulse C is assumed to be a 10 clock pulse from the previous pulse.

As described above, when the carrier pulse is an 8 clock pulse from the previous pulse (a), this indicates a falling edge of the NRZ signal, and may quickly follow a 9 clock pulse interval (D), which indicates no change in the NRZ signal, or a 10 clock pulse interval (E), which indicates a rising edge in the NRZ signal. Similarly, when the carrier pulse is a 10 clock pulse from a previous pulse (C) indicating a falling edge of the NRZ signal, it may be immediately followed by only an 8 clock pulse interval (E) indicating a rising edge of the NRZ signal, or a 9 clock pulse interval (F) indicating no change in the NRZ signal. When the carrier pulse is 9 clock pulses (B) from the previous carrier pulse, this means there is no change in the NRZ signal, and may be followed immediately by an interval of 8 clock pulses (D) representing the trailing edge in the NRZ signal, 9 clock pulses (E) representing no change in the NRZ signal, or 10 clock pulses (F) representing the leading edge in the NRZ signal. These are shown in figure 7. It will be appreciated that in nine code clock cycles of the NRZ bit period, one of the three adjacent pulses can potentially contain a carrier pulse, while the other six cannot.

In the intervals in which no carrier pulse is generated in the CARR signal (from period t4 to t10), other auxiliary data may be modulated on the carrier signal. This is illustrated in fig. 7 as a circular rectangle (auxiliary data) indicated by a vertical dashed line. A guard period t following the last potential carrier pulse (C) around the pause and preceding the next successive potential carrier pulse (D) is maintained to minimise potential interference between carrier pulses (a) - (F) carrying the digital audio signal and carrier modulation (ancillary data) carrying the ancillary data.

Fig. 8 is a block diagram of an embodiment of the invention that may be implemented to modulate the auxiliary data contained in the encoded data stream. In fig. 8, these elements that are the same as those in the example shown in fig. 1 are designated with the same reference numerals and will not be described in detail below. In fig. 8, an auxiliary data (AUX) source (not shown) is coupled to an input of a first-in-first-out (FIFO) buffer 402. An output of the FIFO buffer 402 is coupled to a first data input of a multiplexer 404. The output of multiplexer 404 is coupled to the input of mixer 30. An output of level detector 25 is coupled to a second data input of multiplexer 404. A timing signal output of encoder 10 is coupled to a control input of multiplexer 404.

In the exemplary embodiment, the auxiliary data signal AUX is assumed to be capable of directly modulating the carrier signal. Those skilled in the art will understand how to encode or otherwise prepare a signal for modulating a carrier wave in a manner best suited to the characteristics of the signal. Additionally, in the illustrated embodiment, the auxiliary data signal is assumed to be in digital form. However, this is not essential. The auxiliary data signal may also be an analog signal.

In operation, encoder 10 includes an internal timing loop (not shown) that controls the relative timing of the pulses. While pulses are potentially generated in the CARR signal, the timing loop may be modified in a manner understood by those skilled in the art to generate a signal having a first state in three adjacent encoding clock cycles t1 through t4 and to generate a signal having a second state in the remaining encoding clock cycles t4 through t 10. This signal may be used to control the multiplexer 404 to couple the output of the differentiator 20 to the input of the mixer 30 during periods (t1 to t4) when a pulse is likely to occur, and to control the multiplexer to couple the output of the FIFO buffer 402 to the mixer 30 during periods (t4+ Δ t to t10- Δ t) otherwise. During the period (t1 to t4) when the output of the differentiator 20 is coupled to the mixer 30, the loop of fig. 8 is of the structure illustrated in fig. 1 and operates as described in detail above.

During the period (t4+ At to t10-At) in which FIFO buffer 402 is connected to mixer 30, data from FIFO buffer 402 modulates the carrier signal from oscillator 40. The FIFO buffer 402 is used to receive the digital auxiliary data signal at a constant bit rate and buffer the signal for a period of time (t1 to t4) during which the carrier pulses (a) - (C) can be generated. The FIFO buffer 402 then provides the stored auxiliary data to the mixer 30 At a higher bit rate during the time period (t4+ At to t10-At) in which the auxiliary data is to be transmitted. The net throughput of the auxiliary data burst through the CARR signal must match the constant net throughput of the auxiliary data from the auxiliary data signal source (not shown). Those skilled in the art will understand how to match throughput and how to handle both overrun and underrun, all in a well-known manner.

Fig. 9 is a block diagram of a receiver that may receive signals generated by the system of fig. 8. In fig. 9, the same elements as those in the example shown in fig. 3 are assigned the same reference numerals and are not described in detail below. In fig. 9, the output of detector 140 is coupled to the input of controllable switch 406. An output of the controllable switch 406 is coupled to an input of the decoder 150. A second output of the controllable switch 406 is coupled to an input of a FIFO 408. The output of the FIFO408 generates auxiliary data (AUX). As in fig. 3, the output of the window timer 160 is not coupled to the enable input of the detector 140, but to the control input of the controllable switch 406.

In operation, detector 140 in FIG. 9 is always enabled. The window signal from the window timer 160 corresponds to the timing signal generated by the encoder 10 in fig. 8. The window signal has a first state at the period (t1 to t4) when the carrier pulses (a) - (C) can be potentially generated, and a second state at the period (t4 to t10) otherwise. The switch 406 may be controlled to couple the detector 140 to the decoder 150 during periods (t1 to t4) when the carrier pulses (a) - (C) can potentially generate the timer 160 condition. This structure is the same as the structure illustrated in fig. 3 and operates as described in detail above.

During the remaining bit periods (t4 to t10), the detector 140 is coupled to the FIFO 408. During this period, the modulated auxiliary data is demodulated and provided to the FIFO 408. In a corresponding manner to the FIFO402 (of fig. 8), the FIFO408 receives the auxiliary data burst from the detector 140 and generates the auxiliary data output signal AUX at a continuous bit rate. The auxiliary data signal represents auxiliary data encoded for modulating a carrier. Further processing (not shown) may be necessary to decode the received auxiliary data signal into the desired format.

Claims (15)

1. A digital data modulator, characterized by:

a digital data signal source (IN);

an encoder (10) for encoding digital data using a variable pulse width code;

a pulse signal generator (20, 25) for generating respective pulses representing edges of the encoded digital data signal; and

a carrier signal generator (30, 40) for generating a carrier signal having carrier pulses corresponding to the respective pulses.

2. The modulator of claim 1, wherein the variable pulse width code is a variable aperture code.

3. The modulator of claim 1, wherein:

an encoder (10) for generating an encoded digital data signal having a leading edge and a trailing edge;

a pulse signal generator (20, 25) for generating a positive pulse in response to a first edge in the digital data signal and a negative pulse in response to a different second edge in the digital data signal; and

a carrier signal generator (30, 40) generates carrier pulses having a first phase in response to positive pulses and a second phase in response to negative pulses.

4. The modulator of claim 3, wherein the first phase is substantially 180 degrees out of phase with the second phase;

the first edge is a leading edge; and is

The second edge is a trailing edge.

5. The modulator of claim 1, wherein the pulse signal generator comprises:

a differentiator (20) connected to the encoder; and

a level detector (25) connected to the differentiator.

6. The modulator of claim 1, wherein the carrier signal comprises:

a carrier wave oscillator (40); and

a mixer (30) having a first input connected to the pulse signal generator (20, 25) and a second input connected to the carrier oscillator (40).

7. The modulator of claim 6, further characterized in that a bandpass filter 50 is connected to the output of the mixer (30).

8. A digital data demodulator, characterized by:

a modulation signal source (IN), wherein the modulation signal has carrier pulses spaced apart from each other to represent a variable pulse width encoded digital data signal;

a detector (140) for generating a variable pulse width encoded signal in response to receiving the carrier pulses;

a decoder (150) for decoding the variable pulse width encoded signal to produce a digital data signal.

9. The demodulator of claim 8, wherein the variable pulse width code is a variable aperture code.

10. The demodulator of claim 8 wherein the carrier pulse has one of a first phase and a second phase.

11. The demodulator of claim 10, wherein the first phase and the second phase are substantially 180 degrees out of phase.

12. The demodulator of claim 8, wherein coupled between the modulated signal source and the detector are:

a band-pass filter (110);

an integrator (120); and

a limiting amplifier (130).

13. The demodulator of claim 8, wherein:

a window timer (160) coupled to the detector (140) for generating a window signal in a temporal neighborhood in which the carrier pulse is expected; and, wherein:

the detector (140) is activated by a window signal.

14. A method of modulating digital data, the method comprising the steps of:

providing a source of digital data signals;

encoding digital data using a variable pulse width code;

generating respective pulses representing edges of the encoded digital data signal; and

a carrier signal is generated whose carrier pulses correspond to the respective pulses.

15. A method for demodulating digital data, the method comprising the steps of:

providing a source of a modulated signal having carrier pulses spaced apart from each other to represent a variable pulse width encoded digital data signal;

generating a variable pulse width encoded signal in response to receiving a carrier pulse;

the variable pulse width encoded signal is decoded to produce a digital data signal.