GB2215930A - Sub-carrier reception - Google Patents

Abstract

A method of low-noise reception of a signal comprising a carrier and at least one sideband produced by at least one sub-carrier. The carrier is extracted using a phaselock loop 6 and multiplied with the received signal in a circuit 8 with both in the same phase. The products are filtered by band-pass filters 10-12 of bandwidth equal to the desired noise bandwidth and each centred on a sub-carrier frequency. The filtered components may be processed in different ways depending on the requirement, eg the original received signal can be reconstituted, but with reduced noise content, by multiplying each filtered component with the extracted carrier in circuits 14-16 and added to the extracted carrier in adder 17 for subsequent conventional detection. Alternatively, each filtered component can be detected or processed separately in circuits 18-20 and these outputs added if desired. Phase-shifters 22-24 and attenuator/amplifiers 25-27 can be used to compensate for differential dispersion effects. Amplitude- and phase-modulated signals can be handled, and both symmetrical and assymetrical sideband structures, by suitable modifications. <IMAGE>

Description

Improvements in or relating to sub-carrier reception This invention relates to sub-carrier reception and provides in particular low-noise sub-carrier reception.

Many signals contain sub-carrier modulations which are largely unchanging in themselves but have imposed upon them some form of information which the recipient requires to detect. A typical example would be an amplitude-modulated (AM) carrier in which the modulation, although of relatively high frequency, contains more slowly varying information in the form of sidebands around the sub-carrier producing the modulation.

Such a signal may be produced, for example by causing a sub-carrier of 30 Hz, whose varying phase constitutes the information, to amplitude-modulate a carrier of 100 MHz. In the limit, the slowly varying information can be infinitely slow, so that the information bandwidth about the sub-carrier will tend to zero.

Since the information bandwidth determines the overall signal-to-noise ratio, it is desirable to restrict the receiver bandwidth, at some point, to the value required by the information itself. The usual current practice is to firstly detect the sub-carrier with a pre-detection bandwidth restriction which is adequate to pass the sub-carrier frequency (30 Hz in the above example), then to detect the sub-carrier modulation within a defined post-detection bandwidth (30 Hz in the example), and subsequently to detect the information signal before passing it through a filter having a bandwidth comparable with the defined information bandwidth. This solution involves a sequence of detections and will only -produce the theoretical final bandwidth provided certain conditions are fulfilled, as will be explained.The present invention provides a reception method and apparatus in which these conditions are no longer so severely restricting.

According to the present invention a method of sub-carrier reception comprises: extracting the carrier from a received signal comprising said carrier and at least one sideband produced by at least one sub-carrier; multiplying the received signal by the extracted carrier with both in substantially the same carrier phase; and filtering the products of said multiplication to accept only the component or components of substantially said sub-carrier frequency or frequencies over a desired noise bandwidth or respective bandwidths.

Said accepted component or components may each be multiplied by the extracted carrier and the product or products added to the extracted carrier to reconstitute the received signal but with reduced noise content. Alternatively said accepted component or components may each be detected or otherwise processed and the resulting outputs may be summed.

The relative phases and/or amplitudes of said components may be adjusted to compensate for dispersion effects.

The invention also provides a bandpass filtering method suitable for use in a method as aforesaid comprising: generating a reference signal of approximately the same frequency as an input signal to be filtered; multiplying the input signal directly by a quadrature phase-shifted version of the reference signal and multiplying a quadrature phase-shifted version of the input signal directly by the reference signal; low-pass filtering the respective products of said two multiplications to accept only frequencies not greater than a desired bandwidth frequency; multiplying the filtered product of the first-mentioned multiplication by the reference signal and the filtered product of the second-mentioned multiplication by the quadrature phase-shifted version of the reference signal; and summing the products of said third-mentioned and fourth-mentioned multiplications.

The present invention also provides apparatus for sub-carrier reception and for bandpass filtering as aforesaid.

The present invention will now be described, by way of example, with reference to the accompanying drawings wherein: Fig 1 is a block diagram of an existing sub-carrier detection circuit; Figs 1A and 1B are diagrams showing bandwidths in the circuit of Fig 1; Fig 2 is a block circuit diagram of a circuit embodying the present invention for handling a symmetrical sub-carrier sideband signal; Fig 3 is a block circuit diagram of a bandpass filter usable in the circuit of Fig 2; Fig 4 is a block circuit diagram of a single-sideband mixer for modifying Fig 2 to handle asymmetrical sub-carrier sideband signals.

The circuit of Fig 1 illustrates current receiver practice in sub-carrier detection. A signal of carrier frequency Lr, assumed to be amplitude-modulated with a sub-carrier U m is received by the pre-detection RF or IF circuits 1 having a bandwidth Br as indicated in Fig 1A. Fig 1A also shows the bandwidth Bm, within Br, required to accommodate the sub-carrier frequency xm. The sub-carrier is extracted from the carrier by detector circuit 2, the subsequent circuitry 3, eg filters, having the bandwidth B as shown in Fig 1B. Fig 1B m also shows the bandwidth Bi, within Bm, required to accommodate the information-frequency band.The information is here assumed to be a phase modulation (PM) xm but the principles are the same if Bi relates to amplitude-modulation of w xms or a combination of both forms; the same is true if the modulation of # xr by zm is PM/FM (frequency modulation). Of course forms of detector appropriate to the forms of modulation must be used. The information is extracted from the sub-carrier by detector circuit 4, the subsequent circuitry 5, eg filters, having the bandwidth B..

In the above arrangement the final noise bandwidth can be defined by the final information bandwidth Bi provided the following conditions are met: (1) the signal-to-noise ratio in the small element B of Br m (Fig 1A) is greater than unity; (2) given that (1) holds, the signal-to-noise ratio in the small element Bi of Bm (Fig 1B) is greater than unity.

If these two conditions are not met, the effective noise bandwidth is much greater than Bi. Moreover, if wide-deviation FM/PM is used, the ratio requirements in conditions (1) and (2) are raised to well over unity, in accordance with FM threshold theory.

The required signal-to-noise ratios within the above-stated band elements may not always be obtainable, which can produce problems in some circumstances. One solution would be to increase the transmitted power, or reduce the service area covered, to accord with the required signal-to-noise ratio, but this is not always possible. The present invention provides a solution which gives an acceptable performance provided the signal-to-noise ratio in the element Bi within the band Br is always greater than unity; since this band is usually much narrower than the other pre-detection bands shown in Figs 1A and 1B, the improvement in performance is substantial, allowing the use of reduced transmitter power and/or greater service range.

In the present invention, the double-detection arrangement of Fig 1, ie circuits 2 and 4, is replaced by an arrangement of which one form is shown in Fig 2. In Fig 2 it is first assumed that the received input is an AM signal of the conventional form sin #r t (1 + km cos #m t + k2m cos 2#m t + k3m cos 3#m t) ...(1) where wr is the carrier and w , 20 , 3w are the sub-carriers.

r m m m Usually the sub-carriers are harmonically related as shown, but this is not essential. Their respective amplitudes are shown as km, k2m, k3m. The information signal may be contained in the variable phase of xm etc, or in the variable amplitude of m k etc, or in both. The several sub-carriers usually carry m the same information which, however, may not have the same apparent form in each sub-carrier. This is because the original modulation method may have processed the information differently in respect of each sideband. There can, of course, be more than three sub-carriers. It is also assumed, meanwhile, that the carrier sideband structure produced by the sub-carriers is symmetrical.

The above input signal is fed to a phase-lock loop 6 which extracts the carrier, as cos wrt, in a known manner. The extracted carrier is fed via a goe phase-shifter 7 as sin #rt to a multiplier 8 which also receives the input signal. The product is passed via a low-pass filter 9 which passes the 2wm and 3wm components, but rejects components at #r and above, r to three band-pass filters 10, 11 and 12 having centre frequencies of um , 2 and 3w respectively and bandwidths 1Bi, 2Bi and 3Bi at their centre frequencies. The values of 1Bi, 2Bi etc depend ontheform of processing used for each sub-carrier.If all these forms are identical and linear, then 1Bi - 2Bi -3Bi - nBi I Bi. This is usually the case where each sub-carrier sideband has a small modulation index, but it is not necessarily always so. Large modulation indices will often result in effective B values which depend on the higher values of nBi, ie the higher-order sidebands determine the information bandwidth because they demand a greater information bandwidth. The inclusion of the low-pass filter 9 is desirable to prevent large #r etc components breaking through the band-pass filters.

The subsequent processing depends on the particular requirement. The circuits shown at the RH side of the line 13 are used when it is desired to reconstitute the original signal (1) in a noise-reduced form, effectively with a narrow band-pass filter of width Bi around each carrier sideband. In this case, neglecting meanwhile circuits 22-27, the outputs from filters 10-12 are each multiplied by the extracted sin #rt from phase-shifter 7 in multipliers 14-16 respectively, whose outputs are summed, together with sin #rt from phase-shifter 7, in multipliers 14-16 respectively, whose outputs are summed, together with sin #rt, in adder 17. An application of this requirement will be described later.Due to known dispersion effects, the sidebands as received may have relative amplitudes and phases different from those transmitted. These effects can be compensated by connecting variable phase-shifters 22-24 and attenuators (or amplifiers) 25-27 preceding each multiplier as shown, so that the arrangement corresponds in effect to equaliser requirements. Instead of being located where shown, the attenuator/amplifiers 25-27 can immediately precede the adder 17.

When it is desired to extract the information immediately, the circuits connected at the LH side of line 13 can sometimes be used. The km cos wmt t etc outputs from the filters 14-16 are fed to separate processors 18-20 (amplitude or phase detectors when nB. - Bi) whose outputs (assumed to be parts of a single information signal) are summed at 21. If the several nB.,s are different, this reflects that each information signal has been processed differently during the modulating process, for each sub-carrier sideband. This means that other appropriate processors, eg squarers, multipliers, etc, designed to extract the information from each sideband, may have to be used as circuits 18-20.Although it will usually be desired to sum the outputs of circuits 18-20 as shown, in some applications it may be desired to utilise these outputs separately. The attenuator/amplifiers 25-27 can again immediately precede the adder 21.

In general, such processing makes the arrangement to the RH side of line 13 preferable, since the final wanted output can then be produced by a single conventional detector (not shown) connected to adder 17, which is specific to the form of imposed modulation. Also, even with detectors as circuits 18-20, it will still sometimes be preferable to use the RH side arrangement since, with a multiplicity of parallel detectors, the power applied to each will be correspondingly reduced.

However, if only one or a few detections are required, the LH side arrangement may be more economical to implement.

The foregoing description is limited to the processing of signals where the carrier is amplitude-modulated symmetrically by the sub-carriers. The general expression for a carrier modulated to produce pairs of symmetrical sidebands is sin #r t (1 + km cos k2m cos ) + cos #rt (kn cos #nt + k2n cos 2#nt + ) (2) Usually m - n, but not necessarily so; nor need the sub-carriers be harmonically related as shown, though usually so. It will be seen that the first, ie the sin #rt term, of (2) is the same as expression (1) and is the AM term whose sidebands are in mean phase with the carrier sin #rt. The second term is a PM term whose sidebands are in mean quadrature with the carrier.

In order to process a signal where the carrier is also, or only, phase-modulated symmetrically by the sub-carriers, ie includes the second term in (2) above, some modification to Fig 2 as so far described is required, as shown by the interrupted lines. These result from the above quadrature relationship. A multiplier 81 now receives the input signal together with the extracted cos #rt signal from PLL 6, and the products are fed through low-pass filter 9 to separate band-pass filters101 and 111 and thence, via phase-shifters and attenuator/amplifiers (not shown) corresponding to circuits 22-24 and 25-27 and multipliers 141 and 151, to adder 171 in a similar manner to the circuits on the RH side of line 13 for AM signals.Likewise also, the outputs from filters 101 and 111 can be taken to circuits (not shown) corresponding to 18-20 and 21. For clarity of illustration, circuits for only two sub-carriers #n and 2#n are shown for the FM case.

n n It will be clear that Fig 2 is a general circuit for processing symmetrical sub-carrier signals of the form given at (2), and that in particular cases not all of the circuit shown will be needed. A few examples will now be given.

Example 1 For simple AM by a single sub-carrier (no harmonics thereof and no FM of the carrier by this sub-carrier, though the sub-carrier itself may be FM/PM) all the indices in (2) disappear except (1 + km). In this case only the circuits 6, 7, 8, 9, 10, plus either 14 and 17, or 18, are needed.

Example 2 For simple low-deviation PM/FM by a single sub-carrier (no harmonics thereof and no AM of the carrier by this sub-carrier), all the indices in (2) disappear except kn. In this case only the circuits 6, 81, 91, 101, plus either 141 and 171, ora detector 181 (not shown), are needed.

Example 3 For high-deviation PM/FM by a single sub-carrier m which produces several harmonically related symmetrical sidebands, n - m and (2) becomes sin #r t(1 + k2m cos 2#mt + other even-order terms) + cos #rt (km cos #m t+ k3 cos 3#mt + odd-order terms) In this case appropriate circuits in Fig 2 for both AM and FM processing are needed, and the two respective outputs added in a single adder replacing 17 and 17 Example 4 For high-deviation PM/FM (as in Example 3 with n - m) but with #m treplaced by (wm + AS cos ('S t)t, ie where w is in m q q effect a sub-sub-carrier which phase-modulates w and "S is a m variable modulation index constituting the information signal.

This gives many sidebands of L m which in turn modulate wr.

This is an application, referred to earlier and relevant to navigational beacons such as described in UK Appln No 8713212, where it is preferable to recreate the original signal, using the summed outputs of 17 and 17 , and then employ conventional detection of the sub-carrier and sub-sub-carrier on the lines of Fig 1. Since the Fig 2 circuit has effectively reduced the carrier sideband bandwidth to Bi, the noise level is now low even with conventional FM detection as used in present beacon systems.

Although a conventional form of PLL 6 can be used in the circuit of Fig 2, it is preferred to use the form shown in Fig 4 of copending Appln No 2,187,907 A, which gives automatic amplitude control and more exact phase control. Unlike a conventional PLL, in which the output is in antiphase with the input, the Fig 4 output is in phase within the input. Thus its use requires the phase-shift circuit 7 of the present Fig 2 to be omitted when processing AM signals, and to be inserted between the PLL6 and circuits 81 and 171 when processing FM signals.

At low frequencies there may be difficulty in realising the band-pass filters in terms of conventional passive components. For example, with an AM input signal of the form sin w t (1 + k cos U t) and U - 1 kHz, B. - 1 Hz r m m m i this difficulty would arise. In such cases a filter of the form shown in Fig 3 can be used.

In Fig 3 the cos wmt toutput of filter 9 is led via a 900 m phase-shifter 28 to a multiplier 29 and directly to a multiplier 30. These multipliers also receive a locally generated reference signal cos (w + w )t, where wy represents y y the uncertainty in the value of wm and is not only much smaller than #m but ideally is also smaller than the desired bandwidth Bi; it is assumed that um is known. This reference signal is m fed directly to multiplier 29, and via goe phase-shifter 31 to multiplier 30. The outputs of these two multipliers are fed via respective low-pass filters 32, 33 which pass only frequencies below Bi and are easily realisable, eg in resistor-capacitor form, to respective multipliers 34 and 35.

The latter also receive respectively the phase-shifted reference signal from phase-shifter 31 and the reference signal direct. The two products are fed to an adder 36 where the terms involving zy cancel out as shown. The adder output is y fed via a bandpass filter 37 of centre frequency Xm and m bandwidth normally greater than Bi, to eliminate any spurious terms produced in the multiplication process and deliver the required cos w t output of bandwidth Bi. If Uy cannot be made m y smaller than Bi, the LP filter 33 must be designed to pass Uy y and the resulting noise performance will be limited by this value.

It will be understood that Fig 2 is a generalised form of the present invention, for use with input signals having several pairs of symmetrical sidebands, actually, as shown, three AM pairs and/or two FM/PM pairs. In the lower limit there may be only a single pair of either, in which case the circuit of the present invention may sometimes terminate after the bandpass filters 10 and/or 10 , since the outputs thereof, viz a low-noise information-carrying Xmt tsignal of bandwidth m only Bi, have utility even without subsequent detection or combination with the extracted carrier.An example of such utility would be in known VOR beacons such as Fig 1 of copending UK Application 8713212, where the 30 Hz reference signal received by the aircraft from the centre aerial 1 can be regarded as a sub-carrier of the 114 MHz VHF frequency and whose constant phase constitutes information used for comparison with the phase of the signal received from the other aerials 5. This is an example where the rate of variation of the information is infinitely slow, as mentioned in the present introduction, and a Bi of say 2 Hz would be sufficient. The effect of the present invention would be to place this 2 Hz bandpass filter around the sidebands of the 114 MHz carrier instead of the 30 Hz bandwidth which the receiver would otherwise require and which would itself be difficult to achieve.The resulting noise-reduction would give the beacon greater range, or alternatively allow lower transmitting power for the same range.

The description so far has been concerned with input signals having symmetrical sub-carriers arranged both sides of the carrier, which are not affected by any differential dispersion as regards any given pair of sidebands. On occasion there may be sub-carriers which are asymmetrical, or which are altered as regards amplitude and/or phase by dispersion which affects each of a pair of sidebands in different ways depending on whether it is above or below the carrier frequency. The asymmetrical case can be one of complete asymmetry, in which one of a given pair of sidebands is absent. It can also be one in which each of the pair is present but they are transmitted at different amplitudes and/or phases.For present purposes, the second case can conveniently be grouped with the differentially dispersive case and termed "partially asymmetrical" to distinguish it from the first case, which will be termed "completely asymmetrical".

The circuit of Fig 2 can be modified to handle both forms of asymmetry, by replacing the multipliers 8 and 81 by conventional single-sideband mixers, preferably of the kind shown in Fig 4. In Fig 4 the carrier input from phase-shifter 7 (and/or PLL6) is fed via a 900 phase-shifter 38 to a multiplier 39, and directly to a multiplier 40. The signal input is fed directly to multiplier 39, and via a 90 phase-shifter 41 to multiplier 40. The outputs of the two multipliers are added in adder 42 and subtracted in subtractor 43 to produce two outputs, one corresponding to the upper sidebands and the other to the lower sidebands in a known manner. The desired output is fed to the filters 9 or 91 For a completely asymmetrical input signal sideband structure there will be an output from only one of the two circuits 42, 43. However, it will usually be necessary to retain both circuits because the total input signal will usually include both partially and completely asymmetrical sideband structures. In the ultimate, for an input signal having a partially symmetrical sideband structure, the number of filters and subsequent circuits will be doubled since there will be upper and lower sidebands to handle.

Claims (13)

Claims

1. A method of sub-carrier reception comprising: extracting the carrier from a received signal comprising said carrier and at least one sideband produced by at least one sub-carrier; multiplying the received signal by the extracted carrier with both in substantially the same carrier phase; and filtering the products of said multiplication to accept only the component or components of substantially said sub-carrier frequency or frequencies over a desired noise bandwidth or respective bandwidths.

2. A method as claimed in claim 1 wherein said accepted component or components are each multiplied by the extracted carrier and the product or products added to the extracted carrier to reconstitute the received signal but with reduced noise content.

3. A method as claimed in claim 2 wherein, prior to said multiplication, the relative phases of said components are adjusted to compensate for dispersion effects.

4. A method as claimed in claim 2 or claim 3 wherein, prior to said summing, the relative amplitudes of said components are adjusted to compensate for dispersion effects.

5. A method is claimed in claim 1 wherein said accepted component or components are each detected or otherwise processed.

6. A method as claimed in claim 5 wherein, prior to said detection or other processing the relative phases of said components are adjusted to compensate for dispersion effects.

7. A method as claimed in claim 5 or claim 6 wherein the detected or otherwise processed outputs are summed.

8. A method as claimed in claim 7 wherein, prior to said summing the relative amplitudes of said components are adjusted to compensate for dispersion effects.

9. A bandpass filtering method suitable for use in a method as claimed in any preceding claim comprising: generating a reference signal of approximately the same frequency as an input signal to be filtered; multiplying the input signal directly by a quadrature phase-shifted version of the reference signal and multiplying a quadrature phase-shifted version of the input signal directly by the reference signal; low-pass filtering the respective products of said two multiplications to accept only frequencies not greater than a desired bandwidth frequency; multiplying the filtered product of the first-mentioned multiplication by the reference signal and the filtered product of the second-mentioned multiplication by the quadrature phase-shifted version of the reference signal; and summing the products of said third-mentioned and fourth-mentioned multiplications.

10. Apparatus for use in a method as claimed in any of claims 1 to 8.

11. Apparatus for use in a filtering method as claimed in claim 9.

12. A method or apparatus for sub-carrier reception substantially as described with reference to Figs 2, 3 and 4 of the accompanying drawings.

13. A method or apparatus for bandpass filtering substantially as described with reference to Fig 3 of the accompanying drawings.