GB2185667A - Apparatus for transmitting digital signals - Google Patents

Abstract

Apparatus for transmitting digital signals comprises means for presenting groups of the digits in parallel to modulators fed with different carrier frequencies such that each digit of a group is used to control a parameter of the carrier at the output of a corresponding modulator. The outputs of the modulators are combined and passed to a common amplifier and then transmitted. The invention provides an auxiliary signal generator which adds auxiliary signals containing tones whose frequencies interlace with said carrier frequencies to the signal at the input to the amplifier so that the resultant signal is of substantially constant amplitude. <IMAGE>

Description

SPECIFICATION Apparatus for Transmitting Digital Signals This invention relates two apparatus for transmitting digital signals and in particular to the type of apparatus comprising means for presenting groups of digits in parallel to modulators fed with different carrier frequencies such that each digit of a group is used to control a parameter of the carrier at the output of a corresponding modulator and in which the outputs of the modulators are combined and passed to a common amplifier and then transmitted.

This type of system has been proposed for transmitting digital information via high frequency radio signals over distances such that reflections off the ionosphere need to be relied upon. An advantage of this system, as compared with a system in which the digits are transmitted serially on a single carrier, is that the duration of each transmitted digit is relatively long. This means that the effects of dispersion and of multipath propagation in the ionosphere are reduced. A disadvantage is that the input to the amplifier has an amplitude which fluctuates widely. This precludes the use of a Class C amplifier and also means that the amplifier which is used must be capable of handling a much higher peak power than the mean transmitted power.

According to this invention there is provided, in an apparatus of the aforementioned type, an auxiliary signal generator constructed and arranged to generate an auxiliary signal containing tones whose frequencies interlace the aforesaid carrier frequencies and whose amplitudes and phases are selected such that, when the auxiliary signal is added to an information carrying signal at the input to the amplifier, the resulting signal is of substantially constant amplitude.

The parameter referred to will normally be amplitude but other possibilities such as frequency would be practicable.

By employing the invention the required mean power level can be transmitted by use of an amplifier having a modest peak power capability. Furthermore the amplitude variations of the signal to be amplified can be reduced, by employing the invention, to an extent which makes it possible to use a Class C amplifier, which is relatively efficient. The Class C amplifier may need to be preceded by a limiter.

The aforementioned digits will normally be binary but this is not essential. If the digits are binary it is possible for each digit of a group to control a corresponding modulator. In some forms of the invention it would be possible for binary digits at the input to be converted into non-binary digits used to control the modulators.

The nature of the auxiliary signal can be determined using formula to be described later.

One way in which the invention may be performed will now be described by way of example with reference to the accompanying drawings in which: Figure 1 illustrates in continuous lines a transmitter constructed in accordance with the invention and a corresponding receiver, with the broken lines showing additional components which can be included in place of the limiter denoted by reference numeral 9; Figure 2A shows amplitude variations of a signal at the output of component 8 on Figure 1; Figure 2B shows a typical spectrum of the signal at the output of component 8; and Figure 2C shows a typical spectrum of a signal at the output of component 19 on Figure 1.

Referring to the drawings, serial binary digits, constituting the signal to be transmitted, are clocked by a clock signal S1 into a serial to parallel convertor 2 consisting of cells 3. The clock signal So, which defines the serial bit rate, is generated by a timing unit 4. When the shift register has been filled with new digits, a synchronizing signal S2 is applied to a phase value computer 5. After a short processing delay, the computer 5 produces set of phase values 1, 02...ON calculated in accordance with the equations:

0, for n=O n-l 2Tx (n-j)Pj, n=1,2...N-1 j=o (if N is odd) Oncl= . n-i n -2n for (n-j)Pj,±n, n=1,2,...N-1 j=O N (if N is even) where Pj =fj/ E fk, forj=1,2,...N, k=1 Fj being 1 or 0 according to the jth binary digit in the shift register. These equations are in accordance with a theory described in a paper by M. R. Schroeder (IEEE Transactions on Information Theory Volume IT-16 pages 85--89 1970). When the calculation of the phase values is complete a synchronizing signal S3 is applied to the phase value computer, causing it to apply the phase values to its outputs and thence to the phase controlling inputs of modulators 6.The same signal S3 causes data to be clocked from the serial to parallel convertor 2 to a parallel register 7. The contents of that register 7 are applied continuously to the amplitude controlling inputs of modulators 6. Each modulator 6 is fed with a different carrier frequency fl4N and produces a constant amplitude sinusoid at that frequency whose amplitude and phase are determined by respective inputs marked "amplitude control" and "amp control" on Figure 1. Since the amplitude inputs are binary the output of each modulator is either a common constant amplitude (indicating binary 1) or zero (indicating binary 0).

The outputs of the modulators 6 are combined by addition in a summing amplifier 8. The output 9 of the adder 8 has a much smaller amplitude variation than it would have had without the phase adjustment of its inputs effected in the modulators 6. It does however still have some amplitude variation. In order to remove this residual amplitude variation the adder output 9 is added at 19 to another signal, generated in a manner to be described later, and of a nature such that the output of the adder 19 is of sufficiently constant amplitude for it to be amplified, without significant loss of data, by a Class C amplifier 10 before transmission in form of electromagnetic radiation from an aerial 11.

The radiation is received, after multipath propagation, by a receiving aerial 12 and the signal from this aerial 12 is passed to individual receivers 13 tuned to the frequencies f1, f2. . .fN. A timer 14 synchronized with the timer 4, produces a signal S4 (similar to S3) which causes each receiver to produce digits on its output the same as the digits which were applied to the corresponding modulator 6. These digits are entered into a parallel to serial convertor 15 by the timing signal S4 and the contents of the parallel to serial convertor 15 are clocked out by a signal S5 (similar to S1 ) at the serial bit rate.

Figure 2A shows the envelope of a typical waveform, which still has amplitude variations, generated at the output of circuit 8. The waveform is shown on Figure 2A for a period NT where N is the number of modulators and T is the period of the timing signal S1. NT thus is the period during which the inputs to the modulators remain constant.

Figure 2B shows very diagrammatically a spectrum from the output of the circuit 8 during the period illustrated in Figure 2A. This diagram illustrates how some tones (fl,f2, f4, f5 and fN in the illustrated example) have full amplitude, indicating binary 1; and how others (f3 and f6 in this example) have zero amplitude, indicating binary 0.

D. C. Chu & . W. Goodman have suggested (in Applied Optics Volume 11 pages 171--1724 August 1972) a theory for the generation of digital holograms in which Fourier components constituting the hologram have constant magnitude. Their method is based onthe principle of interlacing two sequences: one being a data sequence representing the information to be stored and the other being a parity sequence derivable from the data sequence in accordance with formulas given in the abovementioned paper.

The inventor has realised that this approach can be used to generate signals of constant amplitude by the use of parity tones interlaced in frequency with the frequency components of the information bearing signal.

The parameters i.e., the amplitudes and phases of the parity tones, are computed in accordance with the formulas given by Chu and Goodman by a computer 16. The inputs to this computer are the amplitudes A1--AN and phases 01--ON and the outputs, read out by timing signal S3, are used to control the amplitudes and phase of modulators 17 (identical in design to 6) and fed with frequencies gl--gN.

The outputs of modulators 17 are summed at 18 and then added at 19 to the output of 8 as previously described.

The formulas given by Chu and Goodman offer a range of possible sets of amplitude and phase values for the parity tones any of which will fulfil the function of making the transmitted signal of uniform amplitude. In an alternative system (not shown) additional information can be transmitted by using the particular choice made as a code carrying this additional information.

There is a dependence between the information carrying tones and the auxiliary tones and hence, in another alternative system the auxiliary tones can usefully be received by additional receivers and used to increase the reliability of communication.

In another alternative system the phase value computer 5 and the phase control inputs to the modulators 6 could be omitted. In such a system some facility would need to be provided to supply the phase value inputs to the computer 16. In such a system the amplitude fluctuations at the output of the adder 19 would not be sufficiently small for receipt by the Class C amplifier. It may therefore be desirable to insert a limiter between the adder 19 and amplifier 10.

Claims (2)

1. Apparatus for transmitting digital signals comprising means for presenting groups of digits in parallel to modulators fed with different carrier frequencies such that each digit of a group is used to control a parameter of the carrier at the output of a corresponding modulator and in which the outputs of the modulators are combined and then passed to a common amplifier and then transmitted, characterised by an auxiliary signal generator constructed and arranged to generate an auxiliary signal containing tones whose frequencies interlace the aforesaid carrier frequencies and whose amplitudes and phases are selected such that, when the auxiliary signal is added to an information carrying signal at the input to the amplifier, the resulting signal is of substantially constant amplitude.

2. Apparatus according to Claim 1 in which the amplitudes and phase of the auxiliary signal tones are calculated.