WO1997018642A1

WO1997018642A1 - Bi-directional amplifiers and signal channels for frequency division duplex communications

Info

Publication number: WO1997018642A1
Application number: PCT/CA1996/000744
Authority: WO
Inventors: Andrew S. Beasley; Stephen Vandenbrink
Original assignee: Allen Telecom Inc.
Priority date: 1995-11-14
Filing date: 1996-11-13
Publication date: 1997-05-22
Also published as: AU7487996A

Abstract

A bi-directional amplifier arrangement which may be used in an RF repeater in an FDD cordless telephone system repeater has signal paths connecting separate inputs (A, B) through a common amplifier (21). A frequency translator (19, 20) in one of the signal paths changes the frequency on one of the signal paths to counteract oscillation.

Description

Bl-DIRECTIONΛL AMPLIFIERS AND SIGNAL CHANNELS FOR FREQUENCY DIVISION DUPLEX COMMUNICATIONS

1 echnical Field

1 he present invention relates to bi-directional amplifiers and signal channels, and to methods of bi-directional amplification and signal transmission, and is useful in particular, but not exclusively, for bi-directional repeater amplifiers and signal channels for use in frequency division duplex (FDD) telephone systems employing mobile cordless handsets.

Background Art

Many communications systems (e.g. cellular systems) make use of bi-directional RF repeaters that boost a signal being transported between two or more points. In frequency division duplex (FDD) communications links, two different signalling frequencies are employed for signal flows in opposite directions.

In cellular radio systems, for example, it is common to find an RF repeater that receives radio signals from a basestation and then boosts that signal for re-broadcast to a local mobile handset. Similarly, the same RF repeater boosts the radio signal received from the mobile handset for re-broadcast to the basestation. In North America, the basestation- to-mobile handset signals are at a frequency of approximately 890 MHZ while the mobile-to-basestation signals are at a frequency separation of 45 MHZ (nominally 890- 45=845 MHZ).

The prior art typically, but not necessarily, uses two physically distinct amplifiers for RF repeaters for frequency division duplex links. One of the amplifiers handles mobile handset-to-basestation amplification and the other amplifier handles base-to-mobile handset amplification. It is possible to use a single amplifier for a bi-directional FDD, RF repeater as shown in Figure 1. which indicates an amplifier arrangement similar to that shown in Fink et al.. Electronics Engineer Handbook, Second Edition, McGraw-Hill 1 82, page 22-40. While this approach has some obvious advantages (e.g. reduced component count and, therefore, reduced cost and improved reliability), it is generally difficult to implement when the FDD separation frequency is a small percentage of the nominal operating frequency.

For example: For North American cellular telephone systems, the percentage separation would be roughly 45/(890)= 5%. Under these circumstances, the passive filter elements of Figure 1 have finite stopband performance at a frequency mid-way between the two operating bands.

The ability of a 845 MHZ bandpass filter to attenuate signals at 845+(45/2)=867.5 MHZ, and the ability of a 890 MHZ filter to attenuate signals at 890-(45/2)=867.5 MHZ. is quite limited.

As a result, the amplifier gain of Figure 1 is constrained so as to avoid oscillation.

Continuing the cellular example, if the two filters each had - 15 dB of the attenuation at

867.5 MHZ, then random noise at 867.5 MHZ at the amplifier outpul could traverse the two filters and travel to the input of the amplifier reduced by - 15- 15=-30 dB. If the amplifier gain of the amplifier exceeds 30 dB, then the conditions for oscillation exist, since the original random noise can be amplified indefinitely.

A relevant prior art disclosure is contained in U.S. Patent 5,321 ,736 issued to Andrew Beasley, the disclosure of which is incorporated herein by reference. Although this prior patent relates to time division duplex systems, it is of interest because it shows the use of RF repeaters within networks which require both wireless and co-axial/wired/fiber transmission. Disclosure of the Invention

According to the present invention. a bi-directional amplifier arrangement has signal paths connecting an amplifier to first and second inputs in a manner such that a first signal passing from the first input to the second input, and a second signal passing from the second input to the first input, are both amplified by the amplifier, a frequency translator being provided in one of the signal paths for changing the frequency of one of the first and second signals.

By imposing a frequency shift, by means of the frequency translator, the risk of oscillation can be avoided.

In a bi-directional signal channel, e.g. between a basestation and a mobile cordless handset in a mobile telephone system, in which the bi-directional amplifier is provided as an RF repeater at one end of a signal conduit, e.g. a co-axial cable or an optical fiber cable forming part of a cable television (CATV) plant, the frequency translated signal can be translated back to its original frequency after transmission through the signal conduit.

Brief Description of the Drawings

The invention will be more readily understood from the following description thereof and taken in conjunction with the accompanying drawings, in which:

Figure 1 shows a block diagram of a bi-directional amplifier without frequency translation;

Figure 2 shows a block diagram similar to that of Figure 1 , but incorporating a frequency translator, in accordance with the present invention;

Figure 3 shows a pair of amplifiers, similar to that of Figure 2, in an RF repeater arrangement in which the amplifiers are connected through a signal conduit;

Figure 4 shows a modification of the arrangement of Figure 3;

Figure 5 shows a further modification of the arrangement of Figure 3;

Figure 6 shows a still further modification o the arrangement of Figure 3; and

Figures 7, 8 and 9 show modifications of he bi-directional amplifier of Figure 2; and

Figure 10 shows a still further modification of the arrangement of Figure

Description of the Best Mode

In Figure 2, there is shown a bi-directional amplifier indicated generally by reference numeral 10, which has a first input 1 1 and a second input 12. The amplifier 10 includes a filter network, which comprises first and second bandpass filters 14 and 1 , which are connected to the input 1 1 , and third and fourth bandpass filters 16 and 17, which are connected to the second input 12.

The bandpass filter 14 is connected through a frequency translator comprising an oscillator 19 and a mixer 20 to the third bandpass filter 16 and, aLso, to the input of an amplifier 21. the output of which is connected to the second and fourth bandpass filters 15 and 16.

The bandpass filters 15 and 16 each have a center frequency H, the bandpass filter 14 has a center frequency L and the bandpass filter 17 has a center frcquenc;/ L'. the frequency translator 20 being arranged to convert a signal from the bandpass filter 14 from the frequency L to the frequency L\

With this arrangement, a first signal passing from the input 1 1 along a first signal path 5 through the bandpass filter 14, the amplifier 21 and the bandpass filter 17 to the second input 12 is frequency converted by the oscillator 19 and the mixer 20 from the frequency L to the frequency L^" and amplified by the amplifier 21. while a second signal passing through a second signal path at frequency H through the bandpass filter 16. the amplifier 21 and the bandpass filter 15 to the first input 1 1 is also amplified by the amplifier 21 but 1 0 is not frequency converted.

By imposing the frequency translation on the first signal path in this way, the possibility of oscillation in the amplifier arrangement is removed, allowing this amplifier arrangement to be used in a large gain RF repeater for certain FDD applications. Any 1 5 random noise at 877.5 MHZ at the amphfier output will now travel to the amplifier input, reduced in level by -3()dB as before, due to attenuation by the filters 14 and 1 5, but now no longer at the original frequency. Hence the original random noise is nol amplified even if the amplifier gain exceeds 30 dB.

20 This frequency translation is reversed after transport of the signal by a second frequency translation, as described below with reference to Figure 3.

Figure 3 shows an RF repeater arrangement, in which the amplifier 10 is connected to an RF repeater antenna 24. so that the antenna 24 serves as the first input of the amplifier i arrangement, corresponding to the first input 1 1 of Figure 2, and the second input 12 is connected to one end of a signal conduit 26.

The signal conduit 26. in the present embodiment of the invention. comprises a coaxial cable, but the present invention may be adapted for use with signal conduits comprising 30 optical fiber cables, cable TV distribution plant, or microwave links The other end of the signal conduit 26 is connected to an input of a second amplifier arrangement indicated generally by reference numeral 10a. For convenience, the components of the second amplifier arrangement 10a are indicated by reference numerals corresponding to those used in Figure 1 but with the addition of the suffix "a".

In the second amplifier arrangement 10a. the bandpass filters 16a and 17a are connected to the input of the amplifier 21a. whereas the output of the amplifier 21a is connected to ihc bandpass filter 14a and. through frequency translator 20a, to bandpass filter 15a. The bandpass filters 14a and 17a have a center frequency H, the bandpass filter 16a has a center frequency IΛ the bandpass filter 15a has a center frequency 1 - and the frequency translator comprising the oscillator 1 and the mixer 20a effects a frequency translation from frequency L^* to frequency L.

Thus, the second amplifier arrangement 1 a serves to boost the signals passing to and from the signal conduit 26, and also frequency translates the signals passing from the signal conduit 26 from frequency L^" to frequency L.

In this case, the input 12a of the second amplifier arrangement 1 a is connected to a basestation 28 which, in turn, is connected in the manner to a public switched telephone network (not shown).

As a practical matter both amplifier arrangements 10 and 10a may share the same power source (not shown in Figure 3 ), using the signal conduit 26 to transport ac or dc power.

The amplifier arrangement 10 may. for example comprise, cellular radio RF repeater where a mobile handset signal is shifted from 855 MHZ to 500 MHZ. The following consequences result:

1 . The operating bandwidths of the amplifiers 21 and 21a must be increased so as to amplify both the 890 MHZ signals and the 500

MHZ signals. 2. The amplifier gains may be greatly increased since the earlier conditions for oscillation are removed.

3. The mobile handset's signals are now available for transmission to the basestation 28 at an increased level, but not at the original frequency of 855 MHZ. A second frequency shifting is therefore affected at the basestation 28 to translate the 500 MHZ signal back to 855 MHZ.

4. Because of the frequency shift, it will not be possible to re-radiate the boosted signal from the mobile handsets, since regulatory authorities have not cleared 500 MHZ broadcast for cellular use. Therefore transport ofthe 500 MHZ signal is best performed over the signal conduit 26. A network type that best supports the present invention is as illustrated in reference U.S. Patent 5.321 ,736.

5. The use of a second frequency shift (as shown in Figure 3), can recreate the conditions for an oscillation path that is much more complex than before.

Referring to Figure 3, the path A-(a)-(b)-D-E-(d)-(c)-B-(d)-(c)-E-D-(a)-(b)-A forms a loop which can support the conditions for oscillation if the net gain is greater than 1.

Oscillation may be precluded by:

1 ) Use of a basestation with physically separate transmit and receive ports 28a and 28b as shown in Figure 4.

2) Preventing the net gain (i.e. gain of amplifier arrangements 10, 1 Oa less signal conduit losses), from exceeding the limit permitted for stability.

For example consider the amplifier gain between (a) and (b) in Figure 3 to be 60 dB, and the amplifier gain between (d) and (c) to be 10 dB, and the filter losses at the nominal oscillation frequency to be - 1 5 dB each. Then the oscillation path has a net gain of:

- 1 5+60- 15-15+ 10- 1 5- 15+10- 1 5-15+60- 15-(twice the signal conduit losses) = 20 - (twice the signal conduit losses).

Thus, provided the signal conduit 26 provides at least - 10 dB of loss, oscillation cannot occur.

Note that the net gain from A to B. for signals centred at frequencies L or 11 would still be 60+ 10- 1 0=60 dB. As a practical matter, co-axial cable losses of -70 dB might be encountered: thus the margin for stability is generous in most practical circumstances.

If this method of ensuring that signal conduit losses are .sufficient to balance loop gain is used, then it is important to monitor the signal conduit losses and to adjust the amplifier gain accordingly This is not trivial for a number of reasons:

the signal conduit losses at frequency L^" may be quite different than at frequency FI

- the signal conduit network itself may be quite complex and possibly modified by the network owner over time

the RF repeater may have adjustable gain and hence the gain at the first moment of installation is important and not just the gain targeted by the installer the signal conduit network may have more than one RF repeater such as the amplifier arrangement 1 connected to the signal conduit.

Consequently, a practical implementation of this invention for dealing with the loop oscillafion may require pilot signals injected onto the co-axial cable as illustrated in Figure 5, in which reference numerals 10c and l Od indicate generally modification of the amplifier arrangements 10 and 10a. respectively.

Referring to Figure 5. a pilot signal is injected at the amplifier arrangement 10c by an intermittent pilot signal generator 30 at a frequency L' (Pilot), which is close to the frequency L\ The level of the pilot signal is then detected, in the amplifier arrangement l d at the opposite end of the signal conduit 26, by a pilot level detector 32 at a central site ( i.e. close to the basestation) as indicated. The use of an intermittent pilot signal allows other intermittent pilot signals, from other amplifier arrangements (not shown) to be similarly supplied through the co-axial cable 26, so that a plurality of RF repeaters such as the amplifier arrangement 10c can be supported.

The output of the level detector 32 is fed back through a microprocessor 33 and a modulator 35; through the co-axial cable 26; through a demodulator 37 and a further microprocessor 39 to an attenuator 41 to effect gain adjustment. The attenuator 41 may be located as shown or may be placed anywhere else within the loop that would otherwise adju.st the net gain from A to the detector 32. Utilized in a system with multiple RF repeaters connected to a single coaxial cable network as disclosed in reference U.S. Patent No. 5.321,736. it is further necessary that the pilot be active intermittently so that356X2 multiple units can use L¹ (pilot) on the same coaxial cable network.

Pilot signals may also be used to set the gain in the path B to A.

As an alternative to the use of pilot signals to prevent oscillation, a high isolation techniques in the second frequency shifter may be employed. An example of such a technique would be the use of a directional coupler 34 as shown in Figure 6, in an amplifier arrangement l Oe which is. thus, a modification of the amπlifier arrangement 1 a to connect the bandpass filters 15a and 1 7a to the input 12a and. thus, to the basestation 28. The directivity of a practical coupler is typically about 30-40 dB, i.e. a signal trying lo transit from (c) to (d) via the directional coupler 34 will be attenuated 30 lo 40 dB. However the loss factors from (c) to (B) and from (B) to (d) arc very much less. Thus, the oscillation loop associated with Figure 3 is brought under control without much impact on the gain of the system.

Although the frequency translation is advantageous when placed as in Figure 2, olher locations for the frequency translation can also be appropriate: Figures 7, 8 and 9 gives some examples.

Figure 2 requires that the amplifier 21 to operate at frequencies L^' and H, and these frequencies may be widely separated. Figure 7 shows a modified amplifier arrangement 1 Of which requires the amplifier 21 operate only at frequencies L and H, which as discussed above are typically close in frequency.

Figure 8 shows a modified amplifier arrangement l Og which permits heterodyned transport of the signal associated wilh frequency H over the signal conduit 26, as well as heterodyning of frequency L.

Figure 9 shows a further modified amplifier arrangement 1 Oh whic n permits transport over the signal conduit 26 at frequencies H' and 1 / while permitting the amplifier 21 to operate only at frequencies I I and L.

In Figures 3, 4, 5, 6 (and by extension Figures 7, 8, 9), the question of accuracy of the heterodyne frequency translation can be posed, i.e. is there any resid al frequency error in translating from L to L^" and then from L' to L.

This can be controlled either by rigorous specification of the frequency heterodyne element, or by the use of a common frequency reference as in Figure 10. The use of a common frequency reference is known, for example, from Canadian Patent Application No. 2.059.370 filed by Nicholas F. Hamilton-Piercy et al.

In Figure 1 0. a reference frequency source 44 is connected to the mixer 20a of the amplifier arrangement 10a through a frequency translator 49a, e.g. a phase locked loop circuit, or directly. The use of such frequency reference source is common in the art to phase and frequency lock heterodyne elements.

By use of a bandpass filter 46, the same frequency reference is also placed onto the co¬ axial cable 26. and is thus freely available to the amplifier arrangement 1 through a PLL frequency translator 49 or directly. By use of a second bandpass filter 48, the frequency reference is available to control the phase and frequency ofthe mixer 20 of the amplifier arrangement 10.

As a consequence of using a common reference frequency at the two amplifier arrangements 10 and 1 0a. any errors in the translation L to L\ may be balanced by corresponding errors of opposite sign in L^" to L, so that the net effect is to ensure error- free transport of frequency L from (A ) to (B).

As will be apparent to those skilled in the art, the invention described above may be varied within the scope of the appended claims.

Claims

A bi-directional amplifier arrangement, comprising signal paths connecting an amplifier (21 ) to the first and second inputs ( A,B) in a manner such that a first signal passing from the first input (A) to the second input (B ), and a second signal passing from the second input (B) to the first input (A), are both amplified by the amplifier (21 ) characterized by a frequency translator (19, 20) in one of the signal paths, the frequency translator ( 19, 20) being effective to change the frequency of one of the first and second signals.

A bi-directional amplifier arrangement as claimed in claim 1. characterized in that Ihe signal paths include a filter network known per se and the amplifier (21 ) is connected across the filter network.

A bi-directional amplifier arrangement as claimed in claim 2, ciaractcrized in that the filter network comprises a first filter ( 14) and a second filter ( 15). which are connected to the first input ( A), and a third filter ( 16) and a fourth filter ( 1 7), which are connected to the second input (B ), the first and third filters ( 14, 16) being connected to one another, the second and fourth fihers ( 15,17) being connected to one another, the amplifier (21 ) being connected between the first and fourth filters ( 14, 17) by the signal paths, and the signal paths comprising a first signal path extending from the first input (A) through the first filter ( 14), the amplifier (21 ) and the fourth filter ( 17) to the second input (B) and a second signal path extending from the second input (B) through the t rird filter ( 16), the amplifier and the second filter ( 1 5) to the first input (A).

A bi-directional amplifier arrangement as claimed in claim 3. characterized in that the frequency translator ( 19,20) is provided in the first signal path between the first filter ( 14) and the fourth filter ( 16) in series with the amplifier (21 ).

A bi-directional amplifier arrangement as claimed in claim 1 , characterized by a further frequency translator in the other of the signal paths, the further frequency translator being effective to change the frequency of the other of the first and second signals.

6. Λ bi-directional amplifier arrangement as claimed in claim 1. characterized by a directional coupler (34) connecting one of the first and second inputs (A.B) to the filter network.

7. A bi-directional amplifier arrangement as claimed in claim 1. characterized by a pilot signal source (30)connected to the one of the signal paths.

8 A bi-directional signal channel, comprising a signal conduit connecting the first and second amplifier arrangements to one another, characterized in that the first and second amplifier arrangements ( 10,10a) each comprise a two-way amplifier arrangement comprising first and second inputs (A.D and E.B), an amplifier (21 or 21 a). signal paths connecting the amplifier (21 or 21 a) to the first and second inputs (A,D and E.B) in a manner such that a first signal passing from the first input (A or E ) to the second input (D or B) and a second signal passing from the second input (D or B) to the first input (A or E) are both amplified by the amplifier (21 or 21 a); and a frequency translator (19, 20, or 19a, 20a) in one of the signal paths, the frequency translator being effective to change the frequency of one of the first and second signals the first and second amplifier arrangements ( 1 0.10a) being so connected by the signal channel that the one of the first and second signals is converted from a first frequency (L ) to a second frequency (Ll ) in the first amplifier arrangement ( 10) and is converted back from the second frequency (L) to the first frequency (Ll ) in the second amplifier arrangement ( 10a).

9. A bi-directional signal channel as claimed in claim 8, characterized in that the first amplifier arrangement ( 10c) is provided with a pilot signal source (30) connected to the one of the signal paths thereof, and the second amplifier arrangement ( l Od) is provided with a pilot signal level detector (32) connected to the one of the signal paths thereof, the channel including means (37, 41 ) for adjusting the signal gain in the one of the signal paths in the first amplifier arrangement ( I OC) in response to a pilot signal level detected by the pilot signal level detector (32).

10. A bi-directional signal channel as claimed in claim 9, characterized in that the pilot signal source (30) is an intermittent pilot signal source

1 1 . A bi-directional signal channel as claimed in claim 8. characterized by a directional coupler (34) connecting the second input of the second amplifier arrangement ( l Oe) to the signal paths thereof.

12. A bi-directional signal channel as claimed in claim 8, characterized by a reference frequency signal source (44) connected to supply a common reference frequency to the first and second amplifier arrangements ( 10, 10a) for effecting phase and frequency locking thereof.

13. A method of bi-directional amplification, which comprises passing a first signal along a first signal path from a first point (A) to a second point (B ) through an amplifier (21 ); passing a second signal along a second signal path from the second point (B ) through the amplifier (21 ) to the first point (A), whereby the amplifier (21 ) amplifies both of the first and second signals, c haracterized by the step of converting Ihe first signal on the first signal path from a first frequency to a second frequency.

14. A method as claimed in claim 13, characterized by transmitting the first and second signals in opposite directions through a signal conduit (26) connected at one end thereof to one of the first and second points (A.B). and. at an opposite end of the signal conduit (26). effecting bi-directional amplification of the first and second signals and frequency translation of the first sigαal from the second frequency to the first frequency.

1 5. A method as claimed in claim 14, characterized by broadcasting the one of the first and second signals as a radio signal, and receiving the other of the first and second signals as a radio signal, at one end of the signal conduit (26) and exchanging the first and second signals with a basestation al the other end of the signal conduit (26).

1 A method as claimed in claim 15, characterized by frequency converting the second signal, passing the frequency converted second signal through the signal conduit ( 26) and reverse frequency converting the second signal after passage thereof through the signal conduit.

I 7. An RF repeater arrangement for a frequency division duplex cordless telephone system, comprising an RF repeater having an antenna for exchanging telephone signals with a mobile handset, the RF repeater comprising a first RF amplifier arrangement for amplifying the telephone signals; a second amplifier arrangement; a basestation connected to the second amplifier arrangement; and a signal conduit connecting the first and second amplifier arrangements to one another; the first and second amplifier arrangements each being a two-way amplifier arrangement comprising first and second inputs, an amplifier, and signal paths connecting the amplifier lo the first and second inputs in a manner such that a first signal passing from the first input to the second input and a second signal passing from the second input to the first input are both amplified by the amplifier; characterized in that said first and second amplifier arrangements ( 10,

1 0a) each include a frequency translator ( 19.20; 19a,20a) in one of the signal paths, the frequency translator ( 19,20; 19a,20a) being effective to change the frequency of one ofthe first and second signals; and the first and second amplifier arrangements (10, 10a) being so connected by the signal channel that the one of the first and second signals is converted from a first frequency to a second frequency in the first amplifier arrangement ( 10) and is converted back from the second frequency to the first frequency in the second amplifier arrangement ( 10a).

1 . Λ bi-directional signal channel as claimed in claim 17, characterized in that the first amplifier arrangement ( 10) is provided with a pilot signal source (30) connected to said one of the signal paths thereof, and wherein the second amplifier arrangement ( 10a) is provided with a pilot signal level detector (32) connected to said one of the signal paths thereof.

19. A bi-directional signal channel as claimed in claim 18, characterized in that the pilot signal source is an intermittent pilot signal source (30).

20. A bi-directional signal channel as claimed in claim 1 7, characterized by a directional coupler (34) connecting the second input of the second amplifier arrangement ( lOe) to the signal paths thereof.

21 . A bi-directional signal channel as claimed in claim 17. characterized by a reference frequency signal source (44) connected to supply a reference frequency to the first and second amplifier arrangements ( 10, 10a) for effecting phase and frequency locking thereof.