US2966583A - Diversity transmission system for beyond-the-horizon signaling - Google Patents

Description

Dec. 27, 1960 MOD.

(a) v g K. F. ROSS DIVERSITY TRANSMISSION SYSTEM FOR BEYOND THE-HORIZON SIGNALING Filed Dec. 12, 1955 v I II I V 21 a RCV 25 DELAY 28 OUTPUT 52%;? 1 Fig. 4

INVENTOR:

Unitedstates Patent? DIVERSITY TRANSMISSION SYSTEM FOR BEYOND-THE-HORIZON SIGNALING My present invention relates to a diversity transmission system for short-wave signaling beyond the radio horizon,

in particular through the use of the so-called forward scattering technique.

In accordance with this recently developed technique, a beam of high-frequency energy from a directive transmitting antenna illuminates an atmospheric volume above the radio horizon, thereby exciting columns of ionic scatters present in greater or less abundance in such volume. The main lobe of the directive pattern of a receiving antenna is trained upon the same volume whereby the latter antenna sees the excited particles and is energized as a result thereof.

The strength of the received signal is subject to more or less predictable long-term (diurnal and seasonal) variations and also to rapid random fluctuations; fading of the signal for periods of the order of several minutes is not uncommon. This situation may be improved by diversity transmission, a standard system of this type comprising a pair of receiving antennas spaced apart along a line transverse to the general direction of the signal path. Diversity signaling by means of different carrier frequencies has been suggested in principle but appears not to have been explored heretofore to an extent.

The present invention has for its general object the provision of an improved diversity transmission system of the character set forth by means of which, apart from a reduction in signal fluctuation, a better overall ratio of received to transmitted power can be obtained.

According to one aspect of this invention, I provide a transmitting antenna and a plurality of receiving an-' tennas whose main lobes are trained upon the transmitted beam so as to intersect the latter at spaced-apart locations. The spacing of the intersections serves to minimize correlation between the two incoming beams but, at the same time, introduces a phase difference between the outputs of the several receiving antennas unless the latter are so positioned that the total path length between the transmitting antenna and each receiving antenna is the same. In the last-mentioned case it may be desirable to employ difierential amplification at the receiving points to compensate for divergences in scattering angle.

Where a positioning of the receiving antennas to equalize the path lengths is impractical, the invention proposes to introduce delay means between the several receiving antennas and a common output circuit therefor, such delay means being preferably inserted in the audio or signal-frequency path of one or more of these antennas and being designed to compensate for the path-length difference.

Another feature of my invention resides in the provision of a plurality of directive transmitting antennas whose beams intersect the main lobe of the or each receiving antenna at spaced-apart locations, delay means being included, if necessary, between the signal source and one or more of the transmitters. If several transmitting antennas as well as several receiving antennas are provided, it is advantageous to maintain all the outgoing beams and also all the incoming beams substantially parallel to one another so that the intersections of any pair of the former with any pair of the latter define a parallelogram, the problem of compensating for path-length differences being thereby considerably simplified. It will be understood that with such arrangement each receiver obtains power from each transmitter, the power ratio in a system according to the invention being thus capable of considerably exceeding that of conventional systems.

It is also an object of my invention to provide an arrangement for frequency diversification which is particularly useful in combination with a multi-beam transmission system as outlined above. In accordance with this aspect of my invention, I provide a source of two carrier frequencies F and F which are so related to each other that half their difference, or (F F )/2, is of the order of the highest signal frequency to be transmitted, such signal frequency being of course substantially lower than either carrier frequency. The linear superposition of these two carriers will then give riseto a carrier (F -I-F )/2 amplitude-modulated with the aforementioned halved difference frequency, this resultant carrier thus reaching a peak or near-peak at least once during each signal-frequency cycle regardless of any spurious phase shifts between the two original carriers. The signal in this case may, for example, comprise a train of pulses each of a duration substantially equal to a cycle ofthe difference frequency (F -F The above and other objects, features and advantages of'the invention will become more fully apparent from the following detailed description, reference being had to the accompanying drawing in which:

Fig. 1 illustrates schematically a beyond-the-horizon transmission system according to the invention, utilizing a plurality of transmitting and receiving beams shown in elevation and intersecting within a terrestrial great-circle plane;

Fig. 2 is a diagram of one of the beam'intersections shown in Fig. 1; i

Fig. 3 shows, schematically, an optional spatial arrangement of the beams of Fig. l as view from the line III--III thereof;

Fig. 4 is a schematic illustration similar to Fig. l but showing a single transmitted beam associated with a pair of received beams; and

Fig. 5 is a set of graphs illustrating the use of frequency diversity in the system of Fig. 1 or Fig. 4.

In Fig. 1 there are shown a pair of highly directive, horizontal transmitting antennas 11 and 12 geographically spaced from each other, the main lobes of their respective beams being indicated at A and A These lobes are directed, within a great-circle plane of the earth E or at a small inclination relative to such plane, toward a region well above the radio horizon as seen from antennas 11 and 12, e.g. in the upper troposphere or the lower ionosphere. The spacing of the antenna 11, 12 is large enough to maintain these lobes well separated from each other throughout the region of interest. The lobes A A will be referred to hereinafter as the outgoing or transmitted beams.

The main lobes B B of the directive patterns of a pair of horizontal receiving antennas 21 and 22, similarly spaced from each other, are trained from the opposite direction upon the aforementioned region above the radio horizon; lobes B and B will be referred to hereinafter as the incoming or received beams.

The beams A A intersect the beams B B at an obtuse angle, their intersections defining four separate Scattering volumes whose crosssectionssin the vertical or great-circle plane have been indicated by vertical hatching; these volumes are designated V (beams A B1), V112 (beams A1, B2), vz (beams A2, B1) and V (beams A B Subject to Whatever corrections may be required to allow for earth curvature and atmospheric refraction, pursuant to conventional practice, the following geometric relationships obtain: Let S be the distance between antenna 11 and volume V S+s' the distance between antenna 12 and volume V R the distance between antenna 21 and volume V and R-l-r the distance between antenna 22 and volume V also, let the center spacing between volumes V V (as well as between volumes V V be s" and let the center spacing between volumes V V (as well as between volumes V V be r". The length of the signal path from either transmitting antenna to either receiving antenna will then be:

Transmitting Receiving Path Length 11 21 S-l-R 11 22 S+r+ R+r' 12 21 s+S+s"+R 12 22 s+S+r+s+R+r Transmitting antennas 11, 12 are part of a sending station comprising, in addition, a signal source 13 conneoted directly to a modulator 14 and through a delay network 15 to a modulator 16. The delay of network 15 is equal to (s'+s")/c where c is the free-space propagation velocity of the radio waves, this network thus compensating, so far as the output of source 13 is concerned, for the phase shift introduced by the relative displacement of the two transmitting antennas. Modulator 16 is also connected to an oscillator 17 producing a carrier frequency F modulator 14 being similarly connected to an oscillator 18 producing a carrier frequency F It will be understood that the two carrier frequencies may be identical, if desired, and that in such event the oscillators 17 and 18 may be combined into a single generator. The output of modulators 16 and 14 is delivered to antennas 11 and 12 by way of transmitters 19 and 20 (including the necessary amplifier stages), respectively.

Antennas 21, 22 similarly form part of a receiving station which further comprises a demodulator 23, connected to antenna 21 via a receiver 24, and a demodulator 25, connected to antenna 22 via a receiver 26. Demodulator 25 works directly into an output stage or load circuit 27, demodulator 23 being connected to that stage through a delay network 28 whose delay is equal to (r'+r)/c, thereby compensating for the phase shift introduced in the signal band by the relative displacement of the receiving antennas. The signals received by the load 27 over the four transmission paths referred to above will therefore be in phase.

As a further means for compensating for the differences in path length, transmitter 20 and receiver 26 may include amplifier means more powerful than those of transmitter 19 and receiver 24, this having been indicated in the drawing by amplifier symbols of different sizes.

The transfer of energy from a transmitter to a receiver by way of a scattering volume has been the subject of much mathematical analysis, not all of it cons stent with experimental observation. Below I shall attempt, at the risk of quantitative inexactitude, to demonstrate with the aid of an oversimplified model of a scattering v'olume'that the system of Fig. 1 will indeed reduce the fading rate to a substantial extent.

Reference is made to Fig. 2 in which I have shown a portion of two beams A, B (which may be considered representative of the beams A and B in Fig. 1) in the immediate neighborhood of their intersection V; the angle between these beams has been designated a, beam A has a width a and beam B a width -b. -A col-ordinate sys em has been plotted so that its x-axis coincides with the lower boundary of beam B, the origin being the junction of this axis with the left-hand (upper) boundary of beam A.

We shall make the following (physically approximately realizable) assumptions:

(1) The width of each beam remains constant throughout the scattering volume;

(2') the base length a/sin a and the height b of the vertical cross section of the scattering volume is small compared to its distance R from the receiving antenna;

and also the following (physically inattainable) postulations:

(3) The scattering volume V is a prism with a parallelogrammatic base and a height z (normal to the plane of the paper in Figs. 1 and 2);

(4) the power of beams A and B is uniform throughout their supposedly rectangular cross sections (a-z and b -z, respectively) and zero beyond the boundaries thereof;

(5 the atmosphere within the scattering volume is uniformly conductive and free from convection currents and other disturbances.

Since the beam A is assumed to originate at a hori- Zontal radiator, we can regard any infinitesimal column z-dXdy as an oscillating dipole traversed by a current :0 cos a+y sin a) from this 'we obtain by double integration, between the limits x. '"=-y cos (x/sin a and x '=(a+y cos a)/ sin a for x and between the limits y =0 and 3 for y, the following expression for the voltage vector E at the receiving It should be observed at this point that, in order for Assumption 2, supra, to be approximately valid, the value of 0: must be considerably greater than zero.

We may now examine the four parameters 2, K, K" and K with regard to their contributions to the phenomenon of fading.

The magnitude of z is dependent upon such constant factors as the transmitter distance S and the carrier frequency F (which, disregarding the existence of sideb'ands, we shall consider invariable) as well as upon such variables as the electron density or the degree of ionization in the scattering volume. Since we have based our calculations upon a supposedly uniformly conductive medium and since changes in the average conductivity of an actual 'medium occur only at a relatively slow rate, we may assume that the parameter 2 is responsible for the diurnal and seasonal variations in signal strength. The coefficient 2.20? sin a wR(1 008M12 isto betermed-hereinafterthe amplitude factorh note that the magnitude of sin cz/ (1cos at) decreases rather rapidly with increasing scattering angles a, in keeping with experimental results.

The parameters K and K are identical in form except in that the former is related to the width a of the incoming beam and the latter to the width b of the outgoing one. Theoretically, on the basis of our oversimplified model, we should expect both of these parameters to remain constant it we disregard the slight variability of the pulsatance to. From a practical viewpoint, however, we must remember that the beam widths a and b are not, in fact, sharply defined and that minor and transitory atmospheric changes in the path of either beam A, B may modify the effective values thereof. Taking K as an example, we note from Equation 4 that its magnitude will go to zero whenever reaches a whole multiple of 11', thus whenever a tan a/Z equals an integral number of half-wavelengths M2. If Equation 3, supra, reflects at least approximately the qualitative aspects of the scattering phenomena, and if 'we are justified in assuming that a and b are not simple geometrical dimensions, as in our model, but are in fact complex mathematical operators based upon unstable arguments, we can expect random fluctuations of K and K" between zero and unity to occur and to produce "proportional variations in the strength of the received .pair of beams, as in Fig. 2, to that of three or more beams intersecting as in Fig. 1. Taking, for example, the two transmitted beams A A and the received beam B we are faced with two scattering volumes V and V each adapted to give rise to a respective voltage vector E E at the receiving antenna 21. Similarly, the combination of beams A A with received beam B provides two scattering volumes V V each adapted to produce a respective voltage vector E E at antenna 22. Complete fading will occur only if all four of these vectors vanish simultaneously.

We shall assume that the distances s and s" are small with respect to S and R and/or are compensated at the sending station by a higher degree of amplification at transmitter 20 compared to transmitter 19; similarly, that the distances r and r" are small with respect to S and R and/or are compensated at the receiving station by a higher degree of amplification at receiver 26 compared to receiver 24. We can then consider the amplitude factors of all the aforementioned voltage vectors to be of the same order of magnitude.

If we apply the theory of our rigid, oversimplified model to each of the four scattering volumes in Fig. 1, We find the outgoing fading factor to have a value for the two scattering volumes V V and to have a value K,'=sin 5 tan a/2 for the two other scattering volumes V V where a and a; are-the Widths of beams A;-and A;, respectively;

similarly, we find the incoming fading factor to have a value K "'=sin tan a/2 for the two scattering volumes V V and to have a value K "=sin tan a/2 for the remaining two scattering volumes V V where b and b are the widths of beams B and B respectively. These factors are paired in the combinations K 'K for vector E Kl Kg for vector E K K for vector E and K 'K for vector E It will thus be seen that complete fading requires that at least three of these four factors go to zero simultaneously.

In a practical system we can expect the outgoing fading factors of, say, volumes V V or the incoming fading factors of, say, volumes V V not to be identical but to exhibit more or less strong correlation, possibly diminishing with increasing spacing of these volumes from each other. The likelihood of total fading is there by reduced still further.

There is, on the other hand, the further possibility of fading through an adverse combination of phase factors where incoming waves from two or more scattering areas are received simultaneously. Thusit is conceivable, for example, that phase factor differs from phase factor at intersections V and V respectively, by a multiple or 11' at the very time when fading factor K is zero or when a similar relationship exists between the phase factors K and K at intersections V and V in which case there would be no signal at either receiving antenna 21, 22. The probability of signal cancellation on one incoming beam, occurring at a time of equal amplitudes and opposite phase angles of the wave energy picked up from two outgoing beams, is roughly as low as that of the simultaneous fading of signals from two scattering volumes; the probability of such cancellation occurring on two incoming beams simultaneously is, of course, very slight.

In Fig. 3 I have shown how the beams A A B and B which we have hitherto considered as co-planar, may be slightly offset with respect to one another so that the axis of each transmitted beam passes between the axes of the two received beams and vice versa. The greatcircle plane has been indicated at G; the intersections of all four beams lie substantially in that plane. As will be apparent from this figure, any change in axial distance between a transmitted beam and a received beam will vary the dimensions a and b of the scattering volume V (Fig. 2) in the region of their intersection. To the extent that the fading factors K, K" are in fact determined by these dimensions, the amplitude of the received wave will be affected by slight shifts in the path of the transmitted and/or the received beam due, for example, to movements of the boundary between atmospheric layers of different electron density.

With the arrangement of Fig. 3, any downward deflection of beam A; or upward deflection of beam A will increase the scattering volume at its intersection with beam B but decrease the scattering volume at its intersection with beam E the converse will be true upon a displacement in the opposite direction. Inasmuch as there is no direct relationship between the size of the scattering volume and the amplitude of the received wave, it will assass n beumde'rstood that the purpose of this arrangement is not simple compensation but diminution of the correlation between conjugate intersections, e.g. as represented by volumes V V or volumes V V An analogous effect will, of course, come into play in the event of upward ordownward deflection of either receiving beam B B as viewed in Fig. 3.

In Fig. 4 I have shown a system similar to that of Fig. 1 but having only a single outgoing beam A, originating at a transmitting antenna 111 and intersecting at V and V respectively, with incoming beams B B terminating at receiving antennas 121, 122. The beams B B are, furthermore, not parallel yet cross each other in such manner that the total path length from transmitting antenna 111 to either receiving antenna 121, 122 is the same. Thus, with S representing the distance between antenna 111 and intersection V and with s being the spacing between intersections V and V it will be seen that antenna 121 is located on a circle centered on V whose radius is R+s whereas antenna 122 is located on a circle centered on V whose radius is R. The need for a delay network at the receiving station is thereby eliminated. The beams 13,, B may be axially offset with respect to beam A in the manner illustrated in Fig. 3.

For reasons to be discussed in connection with Fig. 5, I have provided at the transmitting station a pair of oscillators 117, 118 generating respective carrier frequencies F F and working into modulators 116, 114- which also receive the output of a signal source 113. Modulators 114, 116 feed a transmitter 119 connected to antenna 111. The receiving station includes a receiver 124, connected to antenna 121 and working into a demodulator 123, as well as a receiver 126, connected to antenna 122 and working into ademodulator 125. The outputs of the two demodulators is combined in a load circuit 128. Owing totheless favorable scattering angle of beam B it may be desirable to make the amplifying equipment of receiver 126 more powerful than that of receiver 124, as indicated symbolically in the drawing.

It will be assumed that the carrier frequencies F and F while being sufficiently close together to fall within the range of operating frequencies of antennas 111 and 121, 122, have been selected so that half their difference (F -F )/2 is of the same order as or greater than the highest significant frequency f in the output of signal source-113.

Suppose that the output of source 113 is a train of pulses P of width T, as shown in graph (a) of Fig. 5, and that only the presence or absence of a pulse within an interval 2T is important but not the shape of the pulse nor its exact time position within such interval. The wavedesignated f represents the fundamental frequencyof .the pulse train when the pulses follow one another .at their most rapid rate, i.e. at the rate 1/2T, and constitutesin effect the highest significant signal frequency to be transmitted. Carrier F Fig. 5 (b), and carrier F Fig. 5(c),.differs by one cycle within period T, hence F F =-1/T and, since f =1/2T, (F -F )/2=f Each of these carriers is modulated by pulse P so as to be suppressed except during the period T between I, and t Fig. 5(d-) shows the result of a linear superposition of carriers F and E Given equal amplitudes of the component carriers, the resultant wave has, as is known per se, the form of a carrier of frequency (F +F )/2 amplitude-modulated with a frequency (F F 2 but with periodic'phase reversas and with amplitude peaks occurring at intervals l/ (F F =T. Period T should, of course, be large in comparison with the maximum phase delay. to be encountered on the transmission path between antennas 111 and 121, 122. Even if a substantial phase shift should occur on each of the carriers- F F we may assume that-the relative phase displacement between these carriers. will not exceed a small fraction ofT so that approximate coincidence will be maintained between. them. By virtue of theirdifference in' Y frequency, however, the

8 probability of both carriers fading simultaneously is con. siderably less than that of either carrier fading separately; whenever such fading by one carrier takes place, the other carrier will continue to deliver the signal.

If'both carriers are received simultaneously with approximately equal amplitude, the incoming wave will have the form shown in Fig. 5 (d) except that, owing to relative phase shift, the peaks of that wave may coincide with some different portion of the interval t -t Whether this phase shift be a full cycle of either carrier frequency or a fraction thereof, or even more than one cycle, apeak or near-peak of the composite wave will always occur within that interval so long as the minimum frequency spacing .specified above is observed. Greater fidelity is, of course, obtained if the difference between the carrier frequencies is-increased; thus, .if the input signal were not the pulse P but the Wave i itself, it would be desirable to make (F -F equal to three to four times f rather than to 24 since the amplitude of a sinusoidal wave rises to half its peak and higher during. two-thirds and to more than 70% thereof during one-half of a cycle.

The outputs of signal source 13, oscillator 17 and oscillator 18 in Fig. 1 may likewise be of the character illustrated in Fig. 5, graphs (a), (b) and (c). In such case the two carriers will be transmitted separately via beams A A and will be combined with each other at the intersections V and V Here, as in the system of Fig. 4, the frequency diversity will supplement the geometric diversity to provide improved reception; thus, the risk of "signal cancellation from several scattering volumes '(e.g. V V on the same received beam, due to an unfavorable combination of phase factors as discussed above, will be completely eliminated.

It will be understood 'that, if desired, more than two transmitting and/or receiving antennas may be provided, by simple extension of theprinciples herein disclosed. The invention is, furthermore, susceptible of numerous modifications and adaptations without departing from the spirit and scope 'of the appended claims.

I claim:

1. A radiowave signaling system comprising a source of high-frequency wave energy, a plurality of spacedapart transmitting antennas connected to said source, -a plurality'of spaced-apart receiving antennas positioned beyond the radio horizon of said transmitting antennas, all of said antennas having reflector axes lying substan tially in the same vertical plane, and a utilization circuit for said wave energy connected to said receiving antennas; the reflector axis of each of said transmitting antennas being trained upona region above said radio horizon, the reflector axis of each of said receiving antennas being trained upon said region and intersecting the reflector axis of each transmitting antenna, thereby defining a number of spaced-apart scattering volumes equal to the number of transmitting antennas times the number of receiving antennas; said source of energy inluding first equalizing means for substantially compensating for differences in the respective distances between said scattering volumes and said transmitting antennas; said utilization circuit including second equalizing means for substantially compensating for differences in the respective distances between. said scattering volumes and said receiving antennas; the axes of said transmitting antennas and the axes of said receiving antennas being so close to mutual parallelism that the intersections of any two of the first-mentioned axes with any two of the last-mentioned axes substantially define a parallelogram.

2. A system according to claim 1, wherein said first and second equalizing means each include amplifier means of different power connected to different ones of said transmitting antennas and to different ones of said receiving antennas, respectively.

- 3.- A radiowave signalingfsystem comprisinga source-of high-frequency wave energy, a .pair of geographically separated transmitting antennas connected to said source and provided with a pair of first reflector axes trained in the same general direction upon an elevated region, a source of low-frequency signals connected to modulate the high-frequency wave energy of said transmitting antennas, a pair of geographically separated receiving antennas provided with a pair of second reflector axes trained in the same general direction upon said elevated region, all of said antennas being so positioned substantially in the same vertical plane that each of said first reflector axes is skew to and passes between said second reflector axes while each of said second reflector axes is skew to and passes between said first reflector axes, a common utilization circuit connected to said receiving antennas for recovering said low-frequency signals, and equalizing means connected between said signal source and said utilization circuit, in series with the several signaling paths extending between said transmitting antennas and said receiving antennas, for substantially compensating for difierences in the effective lengths of said signaling paths.

4. A radiowave signaling system comprising a source of high-frequency wave energy, a pair of geographically separated transmitting antennas connected to said source and provided with a pair of first reflector axes trained in the same general direction upon an elevated region, a source of low-frequency signals connected to modulate the high-frequency wave energy of said transmitting antennas, a pair of geographically separated receiving antennas provided with a pair of second reflector axes trained in the same general direction upon said elevated References Cited in the file of this patent UNITED STATES PATENTS 1,954,898 Stone Apr. 17, 1934 2,549,423 Carlson Apr. 17, 1951 2,568,408 Peterson Sept. 18, 1951 2,610,292 Bond et al Sept. 9, 1952 2,629,816 Rabuteau Feb. 24, 1953 OTHER REFERENCES Proceedings of the IRE, vol. 38, April 1950, A Theory of Radio Scattering in the Troposphere, by Booker and Gordon, pages 401-412.

Proceedings of the IRE, vol. 28, October 1940, Experiments on the Propagation of Ultra-Short Radio Waves, by A. H. Waynick, pages 468-475.

Proceedings of the IRE, vol. 43, October 1955, Diversity Reception in UHF Long-Range Communication," by C. L. Mack, pages 1281-1289.

Images