EP2003727A1

EP2003727A1 - A diplexer for a radio communication apparatus

Info

Publication number: EP2003727A1
Application number: EP20070290736
Authority: EP
Inventors: Giorgio Levati; Giuseppe Cereda; Ezio Peroni
Original assignee: Alcatel Lucent SAS
Current assignee: Alcatel Lucent SAS
Priority date: 2007-06-11
Filing date: 2007-06-11
Publication date: 2008-12-17

Abstract

It is disclosed a diplexer for a radio communication apparatus, comprising a first channel filter and a second channel filter, characterized in that said diplexer further comprises a waveguide, a first probe and a second probe, said first probe and said second probe being arranged in said waveguide, wherein said first channel filter is connected to said first probe, and said second channel filter is connected to said second probe.

Description

Technical field

The present invention generally relates to the field of radio communication apparatuses. More particularly, the present invention relates to a diplexer to be employed in a frequency-multiplexing radio communication apparatus.

Background art

As known, in a radio communication system information is transported as electromagnetic waves propagating in air. Carrier frequencies of such electromagnetic waves typically lie in the radio range or in the microwave range, according to the characteristics of the radio communication system (system capacity, span length, etc.). In frequency-multiplexing radio communication systems, different electromagnetic waves at different carrier frequencies, corresponding to different channel frequencies, are multiplexed and transmitted as a single multi-frequency electromagnetic wave.
An apparatus of a radio communication system typically comprises an antenna for transmitting and/or receiving channel signals and a branching network. In the following description, the term "channel signal" will designate an electromagnetic wave transporting information and having a given carrier frequency corresponding to a given channel.
In case of a radio communication apparatus dealing with a single channel, the branching network is adapted to separate a first channel signal to be transmitted by the antenna and a second channel signal which has been received by the antenna. In case of a radio communication apparatus dealing with frequency-multiplexed channels, the branching unit is adapted to multiplex different first channel signals in a first single multi-frequency signal to be transmitted by the antenna, to demultiplex a second multi-frequency signal received by the antenna thus obtaining different second channel signals, and also to separate the first multi-frequency signal from the second multi-frequency signal.
Generally speaking, a branching network should induce low distortion (both amplitude distortion and phase distortion) on each channel signal, i.e. a branching network should have low insertion loss and low group delay distortion. Further, a branching network should provide high isolation between different channel signals.
Different solutions are known for implementing a branching network. One of the most advantageous known solutions is based on the so-called "diplexers".
A diplexer is a component with three ports (a common port and two channel ports), having two channel filters (typically, but not exclusively, two band-pass filters) whose respective central frequencies are two different channel frequencies. Each channel filter is connected to a respective channel port and to the common port.
According to a first use, the diplexer receives a transmitting channel signal through a channel port, and a received channel signal through the common port. In this case, the diplexer outputs the transmitting channel signal through the common port, and it outputs the received channel signal through the other channel port. Therefore, in this case the diplexer separates the transmitting channel signal from the received channel signal.
According to a second use, the diplexer receives a first transmitting channel signal through one of the channel ports, and a second transmitting channel signal through the other channel port. In this case, the diplexer outputs through the common port a two-frequency signal comprising the spatial superimposition of the first and second channel signals. Therefore, in this case the diplexer acts as a multiplexer.
According to a third use, the diplexer receives a two-frequency signal comprising the spatial superimposition of a first and second received channel signals through the common port. In this case, the diplexer outputs the first received channel signal through one of the channel ports, and the second received channel signal through the other channel port. Therefore, in this case the diplexer acts as a demultiplexer.
Therefore, a diplexer is properly working if:

each transmission coefficient between the common port and a respective channel port has a module as close as possible to 1 (0 dB) within a band centered at the respective channel frequency; and
the reflection coefficient of each channel port has a module as close as possible to 1 (0 dB) outside a band centered at the respective channel frequency.

The first condition ensures that insertion loss of the diplexer are as low as possible, whist the second condition ensures that isolation between the two channel ports is as high as possible, so that a channel signal entering the diplexer through one of the channel ports or the common port is not output through the other channel port.
In case a diplexer acts as a multiplexer or a demultiplexer, channel signals to be multiplexed/demultiplexed typically correspond to adjacent channels. In such case, their frequency distance may be particularly reduced, especially in case the diplexer is used in a radio communication apparatus for a high capacity radio communication system, e.g. a radio communication system at 100 Mbit/s or more. In this case, the frequency distance between adjacent channels (i.e. the channel spacing) may be few tens of MHz (25-50 MHz). Further, for exploiting as much as possible the available bandwidth for each channel, in these systems it is desirable to use channel filters wherein each filter has a bandwidth which is substantially equal to the channel spacing (i.e. passing bands of filters corresponding to adjacent channels are very close each other). For instance, in case of a 28 MHz channel spacing, a band-pass filter with a 25,5 MHz passing band would advantageously guarantee exploitation substantially of the whole available bandwidth.
The paper "A Novel Channel Branching Network Approach for Contiguous (28 MHz spacing) and Co-channel operation in SDH Radios", of Rosenberg et al., Proceedings of the 7th European Conference on Fixed Radio Systems and Networks (ECRR 2000), Dresden, Germany, 12-15 September 2000, pages 227-232, describes a diplexer consisting of two channel filters which are directly coupled to a waveguide branching that is suitably formed by a common waveguide flange. The filter design considers a waveguide port for the interconnection with the waveguide branching, and a coaxial port for interconnection with the dedicated receiver or transmitter via semi-rigid cables.

Summary of the invention

However the above known diplexers have some drawbacks, in particular when they are used as multiplexers or demultiplexers of adjacent channels.
In particular, in case of the above mentioned high capacity applications, since the passing bands of the two filters are very close the one to the other, the spectral characteristics of the two filters are mutually dependent, i.e. they undergo linear mutual distortions. Further, nearness of the filter passing bands causes mutual distortions in the group delay of their spectral characteristics. This increases complexity and duration of the tuning of the two filters, since filters have to be tuned alternately until a suitable tuning is found for both filters.
Further, disadvantageously, known diplexers have a fixed mechanical structure, since the operation of these diplexers depends on respective positions and orientations of filters. Therefore, disadvantageously, known diplexers do not allow to modify their structure to comply with constraints of space available for diplexer accommodation within a radio communication apparatus.
Therefore, the invention addresses the problem of providing a diplexer for a radio communication apparatus which overcomes the aforesaid drawbacks.
More particularly, the invention addresses the problem of providing a diplexer for a radio communication apparatus wherein spectral characteristics of the two channel filters are substantially independent the one from the other, this is so even in case of high capacity applications wherein channel spacing is very small (e.g. 28 MHz) and the filter bandwidth is substantially equal to the channel spacing (e.g. 25,5 MHz), while having good performance in terms of insertion loss, group velocity distortion and isolation.
According to a first aspect, the present invention provides a diplexer for a radio communication apparatus, comprising a first channel filter and a second channel filter. The diplexer is characterized in that it comprises a waveguide, a first probe and a second probe, the first probe and the second probe being arranged in the waveguide, wherein the first channel filter is connected to the first probe, and said second channel filter is connected to the second probe.
Preferably, the waveguide has a first end which is closed by a wall being at least partially conductive.
In one embodiment, the diplexer further comprises a first coaxial transition and a second coaxial transition, the first channel filter being connected to the first probe through the first coaxial transition, and the second channel filter being connected to the second probe through the second coaxial transition.
According to preferred embodiments, the first coaxial transition and the second coaxial transition are fixed to the at least partially conductive wall by means of respective sleeves.
Preferably, the waveguide has a second end corresponding to a common port of the diplexer, the common port being a waveguide port.
Preferably, the waveguide supports a single propagation mode, the propagation mode being a transversal-electric mode.
Profitably, the first probe and the second probe are arranged in proximity of the first end of the waveguide at a given distance from the at least partially conductive wall.
Preferably, the first probe and the second probe are arranged in symmetrical positions relative to a symmetry plane comprising a longitudinal axis of the waveguide.
According to preferred embodiments, the first coaxial transition and the second coaxial transition have circular section.
Profitably, the first probe and the second probe are fixed to the first coaxial transition and to the second coaxial transition, respectively, by welding and/or by screw means.
According to a second aspect, the present invention provides a branching network for a radio communication apparatus, characterized in that it comprises a diplexer as set forth above.
According to a third aspect, the present invention provides a radio network apparatus for a radio communication system, characterized in that it comprises a branching network incorporating a diplexer as set forth above.

Brief description of the drawings

The present invention will become more clear by reading the following detailed description, given by way of example and not of limitation, to be read by referring to the accompanying drawings, wherein:

Figure 1 schematically shows a diplexer particularly suitable for being used according to the second and third use, according to an embodiment of the present invention;
Figure 2 shows a perspective view of the waveguide comprised in the diplexer of Figure 1;
Figure 3a, 3b, 3c show experimental measurements of transmission coefficients, reflection coefficients and group delay distortion, respectively, for a diplexer according to an embodiment of the present invention; and
Figure 4 show an exemplary multiplexer employing two diplexers similar to the diplexer of Figures 1 and 2.

Detailed description of preferred embodiments of the invention

Figure 1 schematically shows a diplexer D according to an embodiment of the present invention.
The diplexer D of Figure 1 comprises a first channel filter F1, a second channel filter F2 and a waveguide WG. Further, the diplexer D has a first channel port p1, a second channel port p2 and a common port cp.
Each channel filter F1, F2 has two ports. In particular, the first channel filter F1 has a first port corresponding to the first channel port p1, and a second port pcx1. Similarly, the second channel filter F2 has a first port corresponding to the second channel port p2, and a second port pcx2. Preferably, ports pcx1, pcx2 are coaxial ports, as it will described in detail herein after.
Each channel filter F1, F2 may be a band-pass filter, a low-pas filter, a high pass filter, or a combination thereof. In the following, it is assumed that the channel filters F1, F2 are band-pass filters. The filters F1 and F2 preferably have passing bands centered at f1 and f2 respectively. Preferably, F1 and F2 are filters of a same type, and therefore they have similar spectral characteristic, except for a translation in the frequency domain, equal to the frequency distance between f1 and f2. For instance, F1 and F2 may be multiple-cavity Chebyshev filters. Preferably, F1 and F2 are tunable filters.
The waveguide WG comprises a body having a hollow cross-section, having a conductive inner wall. The waveguide WG has a longitudinal axis parallel to a first direction z, which corresponds to the propagation direction of signals traveling along the waveguide WG. The length of the waveguide WG along the direction z is indicated in Figure 1 as L. The length L may be of few tens of mm.
The hollow section of the waveguide WG may be for instance rectangular, and it lies on a plane xy perpendicular to the direction z. The hollow section size is chosen so that the waveguide WG is a single mode waveguide in the range of frequencies f1 and f2. For instance, the waveguide WG may be a WR137 (R70) waveguide (size along x direction: about 35 mm, size along y direction: about 16 mm), supporting the mode TE₁₀.
The waveguide WG has, at a first end a, a port corresponding to the common port cp. Preferably, the common port cp is a waveguide port. This advantageously allows to directly connect the diplexer D to a circulator or to another microwave component of the branching network in which the diplexer is incorporated.
The waveguide WG further has a second end b, which is shown in perspective in Figure 2. According to embodiments of the present invention, the end b is closed by a wall c, which is preferably a conductive wall. Further, at the end b of the waveguide WG, two probes P1, P2 are arranged within the waveguide WG.
Such probes P1, P2 preferably have a same shape. Figure 2 shows two probes with a parallelepiped shape elongated in the y direction. However, probes may have any other shape (spherical, cubic, prismatic, etc.). Preferably, the external surfaces of the probes P1, P2 are in a conductive material, e.g. silver.
The probes P1, P2 are arranged at a distance d from the wall c, and they are preferably located in symmetrical positions respect to a symmetry plane sp, which is parallel to y and z and which comprises the longitudinal axis of the waveguide WG.
Each probe P1, P2 is connected to a respective coaxial transition Cx1, Cx2. Connection between each probe P1, P2 and the respective coaxial transition Cx1, Cx2 may be implemented through welding or through screw means. Anyway, connection between the probe and the respective coaxial transition should ensure electrical contact between the probe and the ground of the coaxial transition. Preferably, the coaxial transition has a circular shape.
Each coaxial transition Cx1, Cx2 projects outwards the end b of the waveguide WG along the z direction, through respective holes provided on the wall c. Each coaxial transition Cx1, Cx2 is fixed to the wall c by means of a respective sleeve s1, s2, which are shown in Figure 2.
Referring again to Figure 1, the end of the coaxial transition Cx1 projecting outward the end b of the waveguide WG is coupled to the port pcx1 of the filter F1. Similarly, the end of the coaxial transition Cx2 projecting outward the end b of the waveguide WG is coupled to the port pcx2 of the filter F2.
Herein after, by referring to Figure 1, the operation of the diplexer D will be described. It is assumed that the diplexer D is used as a demultiplexer. The use of the diplexer D as a multiplexer will not be described in detail, since it is operationally symmetrical with respect to the use as demultiplexer.
The diplexer D receives through the common port cp a two-frequency signal Sin(f1, f2) comprising a spatial superimposition of two channel signals corresponding to adjacent channels, which are to be demultiplexed before reception. The signal Sin(f1, f2) is coupled to the waveguide WG, where it propagates along the z direction according to the propagation mode TE₁₀. This means that the electric field of the signal is directed along y, while its magnetic field is directed along x.
When the signal Sin(f1, f2) reaches the end b of the waveguide WG, a portion of the signal Sin(f1, f2) whose spectrum lies within the passing band of F1 (which substantially corresponds to Sin(f1)) is coupled to the coaxial transition Cx1 through the probe P1. Similarly, a portion of the signal Sin(f1, f2) whose spectrum lies within the passing band of F2 (which substantially corresponds to Sin(f2)) is coupled to the coaxial transition Cx2 through the probe P2. Therefore, the coaxial transition Cx1 couples Sin(f1) to F1, which outputs it through the first channel port p1. Similarly, the coaxial transition Cx2 couples Sin(f2) to F2, which outputs it through the second channel port p2.
For minimizing insertion loss and group delay distortion of the diplexer D1, the following parameters are preferably adjusted: size of probes P1, P2, distance d between probes and the wall c, and distance between probes and the symmetry plane sp. Preferably, the minimization of insertion loss and group delay distortion is performed at a frequency which corresponds to the central frequency of the channel pattern to be multiplexed/demultiplexed by the branching network wherein the diplexer is incorporated. This advantageously allows to use the diplexer D for multiplexing/demultiplexing any couple of adjacent channels of the channel pattern with an optimized insertion loss and group delay distortion.
More particularly, for minimizing insertion loss and group delay distortion of the diplexer D1, the coaxial transition Cx2 coupled to the probe P2 is connected to a load acting as a short circuit. Then, the three above cited parameters are adjusted for maximizing the transmission coefficient s13 and minimizing the reflection coefficient s11 in the passing band centered at f1. Then, the coaxial transition Cx1 coupled to the probe P1 is connected to a load acting as a short circuit. Then, the three above cited parameters are re-adjusted for maximizing the transmission coefficient s23 and minimizing the reflection coefficient s22 in the passing band centered at f2.
The diplexer shown in Figure 1 and 2 has several advantages.
First of all, the diplexer D is very compact in size, since connection between waveguide and filters is performed through probes and coaxial transitions. Therefore, more voluminous devices such as T junctions or the like are avoided.
Besides, advantageously, the operation of the diplexer D is substantially independent of relative positions and orientations of filters F1, F2. This advantageously allows to adjust relative position and orientation of filters and waveguide to comply with constraints on available space inside the radio network apparatus in which the diplexer is incorporated.
Furthermore, advantageously, tuning the filter F1, F2 is easier and simpler, since each filter can be autonomously tuned, without affecting tuning of the other filter.
Further, advantageously, even in case of high capacity applications, the diplexer according to embodiments of the present invention has good performance in terms of insertion loss, group delay distortion and isolation, as it will be shown in greater detail herein after.
Figures 3a, 3b and 3c show experimental measurements relative to a diplexer whose structure is similar to the one shown in Figures 1 and 2. The diplexer comprises a waveguide WR137 (R70), whose length L is 14 mm. The filter F1 and F2 are six-cavity Chebyshev band-pass filters with a return loss of -26 dB, and with an equi-ripple bandwidth of 25,5 MHz. The filters are tuned at f1=6.200 GHz and f2=6.228 GHz, respectively. The channel spacing therefore is 28 MHz.
In the follow description:

s13 will designate the transmission coefficient between the first channel port p1 and the common port cp;
s23 will designate the transmission coefficient between the second channel port p2 and the common port cp;
s11 will designate the reflection coefficient at the first channel port p1; and
s22 will designate the reflection coefficient at the second channel port p2.

As already mentioned, for a proper operation of the diplexer D, s13 should be as close as possible to 1 (0 dB) within the passing band centered at f1, s23 should be as close as possible to 1 (0 dB) within the passing band centered at f2, while s11 should be as close as possible to 1 (0 dB) outside the passing band centered at f1 and s22 should be as close as possible to 1 (0 dB) outside the passing band centered at f2.
Figure 3a shows a graph of the transmission coefficients s13, s23 versus frequency. The graph is logarithmic, and transmission coefficients s13, s23 are expressed in dB. In particular, Figure 3a shows that the transmission coefficient s13 between the common port and the first channel port of the diplexer has a maximum of -1.1 dB at f1, and its value is almost constant within the whole passing band comprised between f1-Df=6.187 GHz and f1+Df=6.212 GHz (Df being the half bandwidth of the filters, i.e. 25 MHz/2 = 12.5 MHz). Further, the transmission coefficient s23 between the common port and the second channel port of the diplexer has a maximum of -1.05 dB at f2, and its value is almost constant on the whole passing band comprised between f2-Df=6.215 GHz and f2+Df=6.240 GHz. Therefore, the diplexer has acceptable performance in terms of insertion loss.
Figure 3b shows a first graph (on the left side) of the reflection coefficient s11 of the first channel port, and a second graph (on the right side) of the reflection coefficient s22 of the second channel port. The graph is logarithmic, and reflection coefficients s11, s22 are expressed in dB. In particular, Figure 3b (left side) shows that s11 is equal to 0 dB for frequencies higher than f2-Df=6.215 GHz, while it exhibits oscillations at -30 dB about in the passing band between f1-Df=6.187 GHz and f1+Df=6.212 GHz. Similarly, Figure 3b (right side) shows that s22 is equal to 0 dB for frequencies lower than f1+Df=6.212 GHz, while it exhibits oscillations at -30 dB about in the passing band between f2-Df=6.215 GHz and f2+Df=6.240 GHz. Therefore, the diplexer has acceptable performance in terms of isolation between channel ports.
Figure 3c shows a graph of the group delay distortion versus frequency for the transmission coefficients s12 and s23, respectively. In particular, Figure 3c shows that the transmission coefficient s13 has a minimum group delay of 0.51 ns about at the frequency f1, and its maximum value in the passing band between f1-Df and f1+Df is equal to 95 ns about, at the frequency f1+Df. Similarly, the transmission coefficient s23 has a minimum group delay of 0.50 ns about at the frequency f2, and its maximum value in the passing band between f2-Df and f2+Df is equal to 90 ns about, at the frequency f2-Df.
As already mentioned, the diplexer D advantageously allows to build branching networks which are particularly small in size and which exhibit a particularly reduced insertion loss and group delay distortion, since they comprise a reduced number of components. Figure 4 shows an exemplary branching network comprising two diplexers similar to the diplexer D of Figures 1 and 2.
The exemplary branching network of Figure 4 is a multiplexer MUX for multiplexing four channel signals to be transmitted. The multiplexer MUX comprises a circulator C and two diplexers D1, D2. The circulator C has three ports 1, 2 and 3. The diplexer D1 has a common port cp1 and two channel ports p1, p2. The diplexer D2 has a common port cp2 and two channel ports p3, p4. The diplexer D1 comprises two channel filters F1, F2 centered at frequencies f1, f2 respectively, Similarly, the diplexer D2 comprises two channel filters F3, F4 centered at frequencies f3, f4 respectively.
Port 1 of the circulator C is connected to the common port cp2 of the diplexer D2. Port 2 of the circulator C1 is connected to the common port cp1 of the diplexer D1. Port 3 of the circulator C is connected to a transmission port pT of the multiplexer MUX, which can be connected, for instance, to an antenna of a radio communication apparatus, which is not shown in Figure 4.
The multiplexer MUX receives from four transmitters (which are not shown in Figure 4) respective transmitting channel signals Sout(f1), Sout(f2), Sout(f3), Sout(f4) through the channel ports p1, p2, p3, p4 respectively. The diplexer D2 multiplexes the channel signals Sout(f3), Sout(f4), thus outputting one two-frequency signal Sout(f3, f4), which comprises the spatial superimposition of signals Sout(f3) and Sout(f4). The signal Sout(f3, f4) is then received by the circulator C through the port 1, and it is output by the circulator C through the port 2.
The diplexer D1 then receives channel signals Sout(f1) and Sout(f2) from respective transmitters (not shown in Figure 4) through the channel ports p1 and p2, and it also receives the signal Sout(f3, f4) through the common port cp1. Then, the diplexer D1 outputs a four-frequency signal Sout(f1, f2, f3, f4) through the common port cp1, which is sent to circulator C and then to the transmission port pT of the multiplexer MUX.
Therefore, by employing diplexers D1, D2, the multiplexer MUX advantageously comprises only one circulator, in case of four channels. In general, if n is the overall number of channels to be multiplexed/demultiplexed, the number of required circulators is (n/2)-1, under the assumption that n is an even integer. If only filters and circulators are employed, i.e. no diplexers are used, the number of circulators is n-1.
Therefore, using diplexers advantageously allows to reduce the number of components (in particular, of circulators) of a branching network, thus reducing both the branching network size and the insertion loss.

Claims

A diplexer (D) for a radio communication apparatus, comprising a first channel filter (F1) and a second channel filter (F2), characterized in that said diplexer (D) further comprises a waveguide (WG), a first probe (P1) and a second probe (P2), said first probe (P1) and said second probe (P2) being arranged in said waveguide (WG), wherein said first channel filter (F1) is connected to said first probe (P1), and said second channel filter (F2) is connected to said second probe (P2).
The diplexer (D) according to claim 1, characterized in that said waveguide (WG) has a first end (b) which is closed by an at least partially conductive wall (c).
The diplexer (D) according to claim 1 or 2, characterized in that said diplexer (D) further comprises a first coaxial transition (Cx1) and a second coaxial transition (Cx2), said first channel filter (F1) being connected to said first probe (P1) through said first coaxial transition (Cx1), and said second channel filter (F2) being connected to said second probe (P2) through said second coaxial transition (Cx2).
The diplexer (D) according to claim 3, characterized in that it further comprises a first sleeve (s1) and a second sleeve (s2), said first coaxial transition (Cx1) and said second coaxial transition (Cx2) being fixed to said at least partially conductive wall (c) by means of said first and second sleeves (s1, s2), respectively.
The diplexer (D) according to any of preceding claims, characterized in that said waveguide (WG) has a second end (a) corresponding to a common port (cp) of said diplexer (D), said common port (cp) being a waveguide port.
The diplexer (D) according any of preceding claims, characterized in that said waveguide (WG) supports a single propagation mode, said propagation mode being a transversal-electric mode.
The diplexer (D) according to any of claims 2 to 6, characterized in that said first probe (P1) and said second probe (P2) are arranged in proximity of said first end (b) of said waveguide (WG), at a given distance (d) from said at least partially conductive wall (c).
The diplexer (D) according to any of preceding claims, characterized in that said first probe (P1) and said second probe (P2) are arranged in symmetrical positions relative to a symmetry plane (sp) comprising a longitudinal axis of said waveguide (WG).
The diplexer (D) according to any of preceding claims, characterized in that said first coaxial transition (Cx1) and said second coaxial transition (Cx2) have circular section.
The diplexer (D) according to any of claims 3 to 9, characterized in that said first probe (P1) and said second probe (P2) are fixed to said first coaxial transition (Cx1) and to said second coaxial transition (Cx2), respectively, by welding.
The diplexer according to any of claims 3 to 9, characterized in that said first probe (P1) and said second probe (P2) are fixed at said first coaxial transition (Cx1) and to said second coaxial transition (Cx2), respectively, by screw means.
A branching network (MUX) for a radio communication apparatus, characterized in that it comprises a diplexer (D) according to any of claims 1 to 11.
A radio network apparatus for a radio communication system, characterized in that it comprises a branching network (MUX) incorporating a diplexer (D) according to any of claims 1 to 11.