IES80913B2 - Binary digital communication system using a chaotic frequency-modulated carrier - Google Patents

Description

This invention relates generally to means and methods for transmitting and receiving binary digital information, and more particularly to improved means and methods for transmitting and receiving binary digital information employing a chaotic frequency-modulated signal as the transmission carrier.

With the increasing use of digital computers and systems, the transmission and reception of digital information has taken on new importance. Of course, conventional communication systems are available for transmitting digital information, but the use of such conventional systems is somewhat inefficient and wasteful, since the digital information to be transmitted is of relatively simple form and should not require the complexity of conventional communication systems. Besides, it is often necessary to transmit large amounts of digital information requiring many communication systems and/or channels, so that the use of conventional communication systems involves not only considerable circuitry, but also considerable expense.

The problem is further complicated where secrecy, a high immunity to jamming and low sensitivity to multipath effects (see mobile and indoor radio systems) are required of the system, as in certain commercial and military applications. Spread spectrum systems (see, for example, R.C. Dixon: “Spread Spectrum Communication Systems with Commercial Applications,” Wiley, New York, 3rd edition, 1994) offer a solution to these problems. However, they require expensive and complicated circuitry. Moreover, the synchronization of spreading and despreading codes used at the transmitter and receiver sides, respectively, cannot be achieved and maintained under poor propagation conditions, where the channel is time-varying or if it suffers from doppler effects.

Unlicensed radio is a new and emerging field of radio communications, where the power spectral density of the radiated signal is limited in order to avoid interferences with other radio systems. Spread spectrum systems can be and are used in unlicensed radio, but many applications such as wireless local area networks demand a much simpler and cheaper solution. In other cases, it is difficult to achieve and to maintain synchronization of many users in a multi-user spread spectrum system.

Many features of spread spectrum systems such as secrecy and low sensitivity to multipath effects result from the fact that the signal to be transmitted is spread over a large bandwidth. Many of its disadvantages are due to the synchronization requirements for correct operation.

The U.S. Patent 4,363,130 entitled “Binary Digital Communication System” has proposed a solution to this problem. Instead of a sinusoidal carrier, a wideband non-periodic carrier signal derived from a natural source of random noise is used and the digital information to be transmitted is mapped to the correlation between two pieces of non-periodic carrier radiated successively. To recover the binary information in the receiver, the correlation between the two pieces of carrier, which is proportional to the energy per bit in the noise-free case, is determined.

Due to the finite bit duration and the non-periodic nature of the carrier, the radiated energy per bit is a stochastic variable that has a non-zero variance. This variance reduces the performance of the system in the presence of additive noise in the channel. The variance of the radiated energy per bit can be reduced by increasing the bit duration; however, increased bit duration limits the data rate of the system.

The second disadvantage of the binary digital communication system described in I S 8 091 3 U.S. Patent 4,363,130 is that the generation of a non-periodic carrier at a power level which is suitable for radio transmission requires complex and expensive circuitry.

Instead of using a random noise source, a non-periodic carrier can be derived from a deterministic dynamical system; this is called a chaotic signal. A chaotic signal is a g nonperiodic wide-band signal that can be generated by a very simple circuit at any power level and in any frequency band (see, for example, R.N. Madan, editor, “Chua’s Circuit: A Paradigm for Chaos”, World Scientific, Singapore, 1993).

In recent years, efforts have been directed towards devising simplified digital communication systems which take advantage of the noise-like broadband nature of chaotic signals. Several communication systems using chaotic carriers have been proposed. U.S. Patents 5,291,555 “Communication using synchronized chaotic systems” and 5,432,697 “Technique for controlling the symbolic dynamics of chaotic systems to generate digital communications waveforms” describe typical systems.

However, the systems which have so far been devised do not offer any great simpli5 fication over conventional communication systems and have not appreciably reduced the complexity and expense involved in digital transmitting and receiving systems. Also, these systems do not offer any real secrecy, improved data rate, superior performance under poor propagation conditions, or high immunity to jamming. A robust chaotic synchronization scheme that can operate in a typical radio environment where noise, interferences and distortion are present is not known.

Accordingly, it is the broad object of this invention to provide improved means and methods for transmitting and receiving digital information.

A more specific object of this invention is to provide simplified means and methods for transmitting and receiving binary digital information by the use of a chaotic signal as /-5 the transmission carrier signal.

An additional object of this invention is to provide improved means and methods for transmitting and receiving binary digital information in a manner in which more efficient use is made of transmission time. This can be accomplished by using a carrier of constant amplitude which has been frequency-modulated by a chaotic signal. In this way the :. o transmitted energy per bit is kept constant and the data rate is not limited.

A further object of this invention is to provide improved means and methods for transmitting and receiving binary digital information which permit the use of a chaotic signal as the transmission carrier signal and where the demodulation can be performed without chaotic synchronization. ss Another object of this invention is to provide a binary digital transmitting and receiving system which radiates the signal to be transmitted with a low power spectral density.

Still another object of this invention is to provide a binary digital transmitting and receiving system which has a low sensitivity to multipath propagation effects. ί,ο Yet another object of this invention is to provide a binary digital transmitting and receiving system which can operate via a highly distorted and time-varying channel.

An additional object of this invention is to provide improved means and methods for transmitting and receiving binary digital information so that a high immunity to jamming and a significant measure of secrecy is achieved. a. 5 Another object of this invention is to provide improved means and methods for transmitting and receiving binary digital information which is unaffected by doppler.

Still another object of this invention is to provide a binary digital transmitting and receiving system which is capable of operating even in the presence of large amounts of X interfering noise.

An additional object of this invention is to provide the improved means and methods of the aforementioned objects in relatively simple and compact form.

In a typical embodiment of the invention, the output of a relatively wide band chaotic 4 signal is applied to the input of a frequency modulator. The output of the frequency modulator is divided into two correlated signals, one signal being delayed by a predetermined amount with respect to the other. Electronic switching means are then provided for alternately radiating the undelayed and delayed signals for equal periods of time, the sum of these periods being equal to the repetition rate of the symbols of the binary message. The /1), where T is the same as the time delay provided by the delay network 24; each positive pulse corresponds to a “1” of the binary message, and each negative pulse id corresponds to a “0” of the binary message. The conversion of a binary message into such positive and negative pulses can readily be accomplished in a variety of well-known ways, and the block 12 in Fig. 1 is intended to represent the necessary structure for this purpose.

The switching position of the electronic switch 16 is controlled by the signals B, C and ^5' D in such a way that the electronic switch 16 is in position one when signal C is positive and signals B and D are zero, in position two when signal D is positive and signals B and C are zero and in position three when signal B is positive and signals C and D are zero.

In response to each message pulse, whether positive or negative, the binary modulation driver 14 generates a positive pulse with duration equal to T seconds at its output 14a. In response to each positive message pulse, the binary modulation driver 14 produces a positive pulse with duration equal to T seconds that appears at the output 14b of the binary message driver 14 after a delay equal to T seconds. In response to each negative message pulse, the binary message driver 14 generates a positive pulse with duration equal to T seconds that appears at the output 14c of the binary modulation driver 14 after a id delay equal to T seconds. Binary modulation drivers generating these outputs can be easily provided by those skilled in digital electronics.

The signal Si fed to the electronic switch 16 will hereafter be referred to as the reference signal R and the delayed and amplified signal S2 fed to the electronic switch 16 will be h referred as the message signal M, the inverted message signal being referred to as — M. It will now be evident that these signals R and ±M appearing at the inputs of the electronic switch 16 are alternately fed to the RF transmitter 30 for equal time periods of T seconds as a result of the operation of the electronic switch 16 as will now be described. Φ The electronic switch 16 in cooperation with the binary modulation driver 14 and binary message pulses 12 acts to cause the signals applied to the RF transmitter 30 and radiated from antenna 32 to alternate between the reference signal R and the message signals ±M, the time duration of each signal being equal to T seconds. In response to a positive message pulse, first the reference signal R then the message signal M is radiated. • o In response to a negative message pulse, first the reference signal R then the inverted message signal — M is radiated.

Many different versions of the RF transmitter can be used. The RF transmitter 30 may contain power amplifiers, frequency conversion and filter circuits. These circuits can be provided by those skilled in the field. ' < Now referring to Fig. 2 which illustrates graphs corresponding to similarly designated points in the block diagram of Fig. 1, it will be understood that graph A illustrates the type of binary message pulses 12 of repetition rate 1/(271) which may be employed, the positive pulse “1” in the graph A corresponding to a “1” of the binary message and the negative pulse “0” corresponding to a “0” of the binary message. Only four pulses are xo shown in the graph A, but it will be evident that the description contained herein applies in a like manner to any number of pulses which may be employed.

The outputs 14a, 14b and 14c of the binary modulation driver 14 are designated by signals B, C and D respectively.

Graph B of Fig. 2 illustrates the first output obtained from the binary modulation xs driver 14. It can be seen from graph B that in response to each binary message pulse of graph A, whether positive or negative, the binary modulation driver 14 produces a positive pulse of duration T. The designations 7i, T2, T3, 74, T5, T3, Tj and Tg in the graphs indicate successive periods of time of equal duration T.

Graph C of Fig. 2 illustrates the second output obtained from the binary modulation re driver 14. It can be seen from graph C that in response to each positive binary message pulse of graph A, the binary modulation driver 14 produces a positive pulse of duration T; this pulse is delayed by an amount T relative to the binary message pulse.

Graph D of Fig. 2 illustrates the third output obtained from the binary modulation driver 14. It can be seen from graph D that in response to each negative binary message -Τ' pulse of graph A, the binary modulation driver 14 produces a positive pulse of duration T; this pulse is delayed by an amount T relative to the binary message pulse. The designations Ί\, T2, T3, T4, T5, TG, T7 and T8, which are valid for both graphs C and D of Fig. 2, indicate successive periods of time of equal duration T.

Graph E of Fig. 2 shows the non-periodic output of the chaotic signal source 18 which xo is applied after filtering at the input of the frequency modulator 22.

Now referring to Fig. 3 which illustrates the output signal of the frequency modulator 22 designated by signal F in the block diagram of Fig. 1, graph F-T of Fig. 3 shows the output of the frequency modulator 22 in the time domain; signal F acts as a band-limited chaotic carrier with constant amplitude. The power spectrum of signal F shown in graph x F-F of Fig. 3 covers a wide bandwidth with uniform spectral density.

In Fig. 4, it will be understood that graph G illustrates the signal appearing at the output of the electronic switch which is applied to the RF transmitter, radiated from the transmitter antenna 32, received by the receiver antenna 51 and selected by the RF receiver 52 as a result of the previously described operation of the transmitter system 10 of Fig. 1.

Assuming for explanatory purposes that the bit sequence “1010” forms a complete < binary message which is represented by positive-negative-positive-negative pulses as shown by graph A of Fig. 2, then the electronic switch 16 is controlled by the signals shown in graphs B, C and D of Fig. 2. Thus during the intervals Τι, T3, T5 and T7, the electronic switch 16 will be in position three, i.e. only the terminals 16a and 16d are connected.. During the intervals T2 and T6, the electronic switch 16 will be position one, i.e. only the terminals 16a and 16b are connected. Finally during the intervals TA and Tg, the electronic switch 16 will be position two, i.e. only the terminals 16a and 16c are connected.

It now follows, as shown in graph G of Fig. 4, that during the interval Τΐ a reference signal Ri is radiated, and during the interval T2 a message signal Mi is radiated which is in phase with the reference signal Ri and delayed therefrom by an amount T. Similarly, during the interval T3 a reference signal R2 is radiated, and during the interval T4 a message signal -M2 is radiated which is an inverted and scaled version of the reference signal R2. Then during the interval T5 a reference signal R3 is radiated, and during the interval T6 a message signal M3 is radiated which is in phase with reference signal R3 and delayed therefrom by an amount T. Finally, during the interval T2 a reference signal RA is xo radiated, and during Tg a message signal —MA is radiated which is an inverted and scaled version of the reference signal R4.

Because of the non-periodic nature of the signal obtained from the chaotic signal source 18, successive segments of the reference signal such as Ri, R2, R3 and R4, or successive segments of the message signal Μι, —M2, M3 and —M4 are uncorrelated. It xs will be understood, therefore, that the signal radiated from the transmitter 10 appears as nothing more than a continuous noiselike signal, and an unwanted listener who does not know the principle of operation and/or the delay T will be unable to interpret it.

At the receiver 50, the antenna 51 picks up the signal radiated from the transmitter 10 and feeds it to an RF receiver 52 having a bandwidth which is preferably only large enough to receive the band of the radiated signal. The RF receiver can be either a direct or superheterodyne receiver; such receivers can be readily provided by those skilled in the field.

The output of the RF receiver 52 (as shown in graph G of Fig. 4) is divided into two portions, one portion being fed directly to one input 56a of a multiplier 56 and the other portion being fed through a delay network 54 to the other input 56b of the multiplier 56. The delay network 54 is chosen to provide the same delay T as the delay network 24 in the transmitter 10.

Referring to graphs G and H of Fig. 4, it will now be understood that the signal applied to the input 56a of the multiplier 56 will be that shown in graph G, while the signal applied to the input 56b of the multiplier 56 will be that shown in graph H, which is the signal of graph G delayed by T as a result of passing through the delay network 54. The output of the multiplier 56 is fed to the integrator-with-dump 58 for integration thereof.

The action of the multiplier 56 is such that when in-phase correlated signals are simula f taneously applied to its inputs 56a and 56b, a signal with a positive average value will be produced at its output 56c, and when correlated but inverted signals are simultaneously applied, a signal with a negative average value is produced at the multiplier output 56c. On the other hand, uncorrelated signals applied to the multiplier 56 produce an output therefrom which has a zero average value, i.e. it tends to integrate out to zero. Typical types of multipliers which may be used as the multiplier 56 in Fig. 1 are described in “Phase-Locked Techniques” by F.M. Gardner, Wiley 1979, pp 107-116.

Referring to the graphs G and H of Fig. 4, it will now be understood that during < the time interval T2, the in-phase correlated message and reference signals Mi and Ri are simultaneously applied to the multiplier 56 so as to produce an output from the multiplier 56 with a positive average value as illustrated by “1” in graph J of Fig. 4; and during the interval T4 the correlated but inverted message and reference signals —M2 and R2 are simultaneously applied to the multiplier 56 so as to produce an output from the multiplier <° 56 with a negative average value as illustrated by “0” in graph J. Similarly, during the time intervals T6 and T8, the outputs from the multiplier 56 have positive and negative average values, respectively, as illustrated in graph J of Fig. 4.

During the intervals T3, T5 and TV, correlated signals do not appear at the inputs to the multiplier 56 so that only a signal having zero average value is obtained at the ’ 5' output 56c of the multiplier 56 as illustrated by N in graph J. It should be remembered that because the message and reference signals are derived from the output of the chaotic signal source 18, successive segments of each will be uncorrelated. Thus, signals such as Mi and R2, —M2 and T?3, Af3 and R4, which are simultaneously applied to the multiplier 56 during intervals T3, T5, and T7, respectively, are uncorrelated and produce only a signal having zero average value, as indicated by N in graph J.

It will now be evident that the binary message is carried by the average value of the output 56c from the multiplier 56. The average value of signal J can be recovered by an integrator implemented using either analog or digital circuitry.

If the integrator is not discharged at the beginning of each symbol then neighboring ‘F symbols interfere with one another, i.e. intersymbol interference occurs which degrades the performance of the system. Intersymbol interference can be avoided by discharging the integrator at the beginning of each symbol. This can be achieved by means of a dump circuit in the analog case or by using a moving average technique in the digital domain. A detailed description of the intersymbol interference problem and circuit solutions -° thereto can be found in “Digital Communication” by A.E. Lee and D.G. Messerschmitt, pp. 289-324, Kluwer Academic Publisher, 2nd edition, 1993 and in “Digital Microwave Transmission, ” by I. Frigyes, 2. Szabo, and P. Vanyai, Elsevier Scientific Publishing Co, 1989, pp. 43-55.

Assume for explanatory purposes that an analog integrator-with-dump 58 is used to recover the average value of signal J. As shown by graph K of Fig. 4 the dump circuit resets the output of integrator 58 to zero during the intervals 7\, T3, T5 and T7. During the intervals T2, T4, T6 and TB, the dump circuit is switched off and consequently the output of the integrator 58 becomes positive and negative corresponding to “Is” and “Os” of the binary message, respectively, as illustrated by graph K of Fig. 4. Due to the 0 averaging, the binary message is easily recognizable even in the presence of large amounts of interfering noise.

The integrator used to recover the average value of signal J can be implemented in many different ways either in the analog or in the digital domain. In certain applications, the discharging of the integrator can be omitted to obtain a simpler system configuration; < of course, the system performance is limited by intersymbol interference problem in this case.

The output from the integrator 58, shown in graph K of Fig. 4, may be fed to any suitable type of threshold detector 60 and sampled at the ends of the even-numbered intervals T%, Ί\, Τ6 and Τ& to produce positive and negative pulses of suitable amplitude and duration for application to digital circuitry. The ends of the even-numbered intervals Γ2, T4, T$ and Tg are called decision time instants and they are assigned by the receiver’s clock signal. If signal K is greater than the threshold at the decision time instant then < a positive pulse appears at the output of the threshold detector 60 while if signal K is less than the threshold at the decision time instant then a negative pulse appears at the output from the threshold detector 60. In graph K of Fig. 4, the threshold is assumed to be zero.

The receiver’s clock signal is used to assign the decision time instants and to control the / 0 integrator 58. The clock signal may be recovered from the incoming signal as discussed in “Digital Microwave Transmission,” by I. Frigyes, Z. Szabo, and P. Vanyai, Elsevier Scientific Publishing Co, 1989, pp. 139-144. In very simple digital communication systems the recovery of the clock signal can be omitted. Because the clock recovery circuit is not the subject of this patent, it is not shown explicitly in Fig. 1. 1M From the above description of the invention, various important features of the system will now become evident. First, as mentioned previously, because detection is obtained by multiplying correlated pulses and then integrating, detection of these pulses is possible even in the presence of large amounts of interfering noise, since uncorrelated signals tend to average out to zero after integration. Secondly, because both the transmitted reference a o and message signals receive the same doppler shift (which may result from relative velocity between the transmitter and receiver), doppler shift will have no effect upon the operation of the system.

Thirdly, because both the transmitted reference and message signals pass through the same channel, the system is insensitive to time-varying effects in the channel which occur on a time-scale which is large relative to the delay T.

In other binary digital communication systems which use a chaotic carrier, the data rate is limited by the time required by the receiver to estimate some property of the received signal. The use of frequency modulation makes the transmitted energy per bit constant, i.e. it is not a stochastic variable. The use of frequency-modulation thereby -3 reduces the required bit duration and provides a higher data rate.

Another feature of the system which can be of great importance in some situations is that the system permits any sort of chaotic signal source to be applied to the frequency modulator in order to generate the carrier.

Also, the use of a chaotic signal, particularly one which is spread over a relatively b F wide band, is very difficult to jam because of the wide range of jamming frequencies which must be provided and the small possibility that they will be sufficiently correlated to affect system operation. Further, since the radiated signal appears as a continuous noiselike signal, an unwanted listener who did not know the value of T could not interpret the radiated signal even if the principle of operation were understood.

° Still another feature of the system is that because both the transmitted reference and message signals pass through the same channel, the system is insensitive to distortion caused by the channel.

A further advantage of the system is that the power spectral density of the radiated signal is low; therefore, the system causes minimal interference with other radio systems.

? The system proposed here is suitable for use in unlicensed radio applications.

An additional important feature of the system is that the demodulation is performed without chaotic synchronization. This solution makes the system very simple and robust against channel noise, interferences and distortions.

The last feature of the system follows from the fact that the cross correlations between segments of chaotic waveform are low; therefore, the system has low sensitivity to multipath effects. This is why it is suitable for use in mobile communications, indoor radio and wireless local area networks.

In a possible implementation of the system shown in Fig. 1 the value of T is chosen equal to 1 millisecond, a chaotic analog phase-locked loop is used as the chaotic signal source 18, and the frequency modulator 22 is chosen to have a a center frequency of 20 kilohertz. The bandwidth of the output of the frequency modulator 22 is 30 kilohertz. ι o The gains G and — G are 1 and -1 respectively. These values are only illustrative and should not be considered as limiting the scope of the invention.

It is to be understood in connection with the system described herein that the electronic circuitry and devices designated in block form in Fig. 1 are all of a type which can be readily provided by those skilled in the art. Since the present invention resides chiefly ' S in the combination of these electronic devices and circuitry and not in the design of any particular one thereof, the details of these devices and circuitry will not be given. However, based upon the description and operation of the various systems provided herein, those skilled in the art will have no difficulty in practicing the invention.

It is also to be understood that the invention is not limited to the embodiment dex ° scribed and illustrated herein, since many modifications and variations in the construction and arrangement thereof may be made without departing from the scope of the invention as defined in the appended claims.

Dated this the 28th day of November, 1997,

Claims (5)

1. A transmitter for transmitting a binary message, comprising a) a chaotic signal source, b) means for applying the chaotic signal to the input of a frequency modulator, c) means for dividing the output of the frequency modulator into a first and second signal, d) means for delaying the first signal by a predetermined amount with respect to the second signal, e) means for multiplying the delayed first signal by a positive or negative constant in response to the binary message, f) electronic switching means for alternately radiating, via an RF transmitter, the second and multiplied delayed first signals, and g) means for restricting the bandwidth of the chaotic signal, optionally this means may comprise a low pass filter.

2. A receiver for receiving a binary message comprising symbols, which has been modulated using a chaotic frequency modulated carrier comprising a) a RF receiver for receiving the chaotic frequency modulated message, b) means for dividing the output of the RF receiver into a first and second signal, c) means for delaying the first signal by a predetermined amount with respect to the second signal, d) means for multiplying the delayed first signal and the second signal, e) an integrator for integrating the resulting multiplied signal, f) a threshold detector for retrieving the binary message from the output of the integrator, and g) means for discharging the integrator at the beginning of each symbol of the binary message.

3. A receiver as claimed in claim 2, wherein the integrator is implemented using digital circuitry and the means for discharging is implemented using a moving average technique.

4. 5 4. A receiver for receiving a binary message, which has been modulated using a chaotic frequency modulated carrier, as herein before described and with reference to and, or as illustrated in the accompanying drawings. 5. A transmitter for a transmitting a binary message, using a chaotic frequency

5. 10 modulated carrier, as herein before described and with reference to and, or as illustrated in the accompanying drawings.