Patent Application
20120162011
This invention relates to phase control in radiofrequency transmission and reception using arrayed antenna elements.
It has long been known that arrays of multiple antennas for radar and other radiofrequency transmission and reception offer certain advantages over single-element antennas, such as enhanced spatial selectivity, signal gain, and beam steerability. These and other advantages are greatest when there is precise control over the phases of the antenna elements; i.e., over the relative phase of the wavefront leaving each transmissive element, or of the relative phase, at the detector, of the signal collected by each receptive element.
Conventional methods of phase control include electronic methods based on the transfer function of a reactive circuit, and delay-based methods that use variable-length delay lines to adjust the phase of each radiofrequency (RF) feed to an antenna element. Neither of these approaches is perfectly adapted for all applications. For example, one drawback of electronic methods is that they are limited in bandwidth. One drawback of delay-based methods is that precise, tuneable phase control is difficult to implement.
Accordingly, there remains a need for techniques of phase control that combine high precision with high bandwidth.
We have developed a technique based on optical delay that can provide both high precision and high bandwidth.
In an embodiment adapted for transmission, a light beam is modulated with an RF signal. The light beam is divided into a plurality of beamlets and distributed through an optical network to an array of transmission elements. At each transmission element, at least one beamlet is converted to an RF signal and transmitted.
The optical network includes wavelength-selective elements coupled to optical delay lines. The optical network uses wavelength based routing to deliver each beamlet through a designated amount of delay to a designated transmission element.
In an embodiment adapted for reception, an incoming radiofrequency signal is converted to an electric signal at each of a plurality of reception elements. At each reception element, an optical beamlet is modulated with the electric signal. The respective beamlets are combined into a composite optical signal as a result of propagating them through an optical network of the kind described above. The composite optical signal is detected and further processed, for example by demodulation. While propagating through the optical network, the beamlets are subjected to wavelength based routing to deliver each beamlet through a designated amount of delay before it is combined into the composite optical signal.
An embodiment of the invention comprises an optical network of the kind described above, as adapted for transmission, reception, or both transmission and reception.
A type of optical network useful for the practice of the invention is a network in which passive wavelength-selective optical delay (WSOD) devices are combined with wavelength-shifting devices to provide wavelength-switched optical delay. Such wavelength-switched optical delay networks are known. One example is described in J. D. LeGrange et al., “Demonstration of a time buffer for an all-optical packet router,” J. Opt. Networking, vol. 6, no. 8 (August 2007) 975-982 (LeGrange 2007).
With reference to
WDM device 10 includes an arrayed waveguide grating (AWG) 20 on the input side, and an arrayed waveguide grating 25 on the output side. Each AWG has a number N′ of input ports 30, 35 and a number M′ of output ports 40, 45. (In the view of
Although not essential, it will often be advantageous for gratings 20 and 25 to be symmetrically arranged, such that the number of input ports of AWG 20 is matched to the number of output ports of AWG 25, and likewise that the number of output ports of AWG 20 is matched to the number of input ports of AWG 25. In the discussion below, we will assume the same number N of ports for the input and output sides of both AWG 20 and AWG 25. Accordingly,
As those skilled in the art will understand, an AWG functions as a two dimensional diffraction grating. As such, it can convert spectral routing to spatial routing. A typical AWG is made from two interconnected star couplers. The connection between the star couplers is made by an array of waveguides having linearly increasing lengths.
Due to the diffractive behavior of the arrayed waveguides, a suitable optical input will result in light emerging from each waveguide at a particular wavelength. The wavelengths are determined by the lengths of the respective waveguides, in accordance with the laws of optical interference. The length increments between waveguides are typically set to provide a phase shift of 2πA radians from each waveguide to the next, where A is the diffractive order of the grating.
More particularly, an input signal applied to a given input port will be mapped to different output ports with respective shifts of wavelength. Accordingly, a signal having a given wavelength can enter the AWG on any input port and be routed to a unique output port determined by the given wavelength and by the identity of the input port.
Known designs for the star couplers and waveguide grating enable the AWG to be used as a spectral multiplexer or demultiplexer with minimal crosstalk between channels. The AWG may be used over multiple grating orders, thereby extending the usable wavelength range and making it possible to form multiple beams simultaneously. One source of further information on the AWG is C. R. Doerr, “Planar Lightwave Devices for WDM” in Optical Fiber Telecommunications, volume IVA, edited by Ivan Katninow and Tingyc Li, (Academic Press, New York, 2002), pp 405-476.
Turning back to
Each coupling between an output port 40 and an input port 35 is made through a respective optical delay element 50. Typically, each of the optical delay elements 50 will provide a different amount of delay.
In view of the foregoing, it will be understood that an AWG arrangement such as that shown in
It should be noted that if the mapping between input and output ports of each of the AWGs is different for each operating wavelength, then it may be possible to apply input signals simultaneously to all of the input ports 30 without collision. That is, two signals applied to different input ports 30 will be mapped to the same output port 40 only if they are on different operating wavelengths. If they are on different operating wavelengths, they will not affect each other. Similarly, two input signals can be applied to the same input port 30 without colliding if they are on different operating wavelengths. (Although the AWG is described here with linearly incrementing phase and therefore wavelength shifts from channel to channel, it should be noted that in other embodiments, any router design that results in wavelength selection of the output port could be used.)
Turning now to
The light emerging from each of output ports 200 may be extracted from the optical delay network for further processing and utilization as will be described below, or it may be directed to a next stage of the optical delay network, where it is again split in an optical splitter (not shown), and each output from the splitter is subjected to a further wavelength shifter (not shown) and injected at an input port of a further WSOD device, such as device 210 of the figure.
Each stage of the network of
As shown in inset 350, each sub-network includes an optical splitter 351, a set of wavelength-shifters 352 subject to a control unit (not shown), and a WSOD device 353.
In the design of antenna arrays, it is often advantageous to organize an array having many elements into a plurality of sub-apertures that are organized hierarchically, so that a sub-aperture at a higher level of organization includes a plurality of sub-apertures at a lower level of organization. Advantageously, each of the sub-networks at each stage of the network is associated with a respective sub-aperture of the array. To illustrate this concept,
Turning again to
As noted earlier, two optical signals can enter or exit the same ports of a WSOD device without colliding if they are in different wavelength channels. As a consequence, it may be possible in some implementations to use the optical delay network, or a portion of it, for delay processing of two or more simultaneous signals carrying independent information, if the respective signals are placed on mutually orthogonal sets of operating wavelengths.
For example, those skilled in the art will appreciate that one of the features of an AWG device is the free spectral range (FSR), having the property that if signals of two wavelengths separated by the FSR are applied to the same input port of an AWG demultiplexer, they will be directed to the same output port. Thus, the FSR defines a (weakly wavelength-dependent) periodic band structure for the responsive behavior of an AWG device. Mutually orthogonal sets of operating wavelengths can be selected on the basis of this band structure.
Similarly, it may be possible to use the same WSOD device to simultaneously perform the delay processing of an optical signal for two different sub-apertures, if the sets of operating wavelengths corresponding to the respective sub-apertures are chosen appropriately. This may be advantageous if, for example, the various sub-apertures differ only in their corresponding coarse amounts of delay, but add to the coarse delay the same increments of fine delay. Thus, the total amount of hardware could be reduced by reusing one or more of the WSOD devices that provide fine delay.
It should be noted that if one or more WSOD devices are reused for multiple independent signals or for multiple sub-apertures (at the same level), it will generally be necessary to include one or more wavelength demultiplexers in the network for separating the respective mutually orthogonal sets of operating wavelengths after the last reused device.
As noted above, the spatial selectivity and beam steerability achievable using arrays of multiple antennas are highly advantageous for radar, communications, and other radiofrequency applications. The signal processing that underlies these capabilities of antenna arrays is beamforming, i.e., the coherent combination of the signals going to or from the respective antenna elements.
Beamforming is typically achieved using electronic phase shifters, which are well known. However, the performance of electronic phase shifters is frequency-dependent. For that reason, beamforming is disadvantageously limited in bandwidth when it is performed solely by using electronic phase shifters.
In accordance with the invention, a wavelength-switched optical network such as that described above is used to provide true time delay for at least part of the beamforming. That is, the timing of the phase fronts propagating from individual antenna elements during operation in the transmission mode, or the effective (from the viewpoint of the receiver) timing of the phase fronts propagating toward the individual antenna elements during operation in reception mode, is controlled by optical delay in the signals that the optical delay network directs to or from the antenna elements. Because the optical delays are not affected by the frequencies used for radiofrequency modulation, bandwidths can be achieved that are much greater than those achievable using only electronic phase shifters.
We believe that because of the precise tolerances achievable in the fabrication of optical delay elements, true time delay can be used to provide controllable delay increments over an extremely wide dynamic range, extending from microseconds or more, down to 0.01 ns or even less. In typical switched fabrics of the kind described here, true time delay provided via optical delay elements will be most useful in the range from 0.1 ns to 100 ns. For the finest phase control at the last stage of the network (i.e., at the stage nearest the antenna elements), we believe it will be most advantageous to use electronic phase shifters. (It should be noted in this regard that the performance of electronic phase shifters is limited by the product of bandwidth times interelement separation. Thus, the electronic phase shifters are most advantageous at the finest level of delay processing, where the corresponding antenna elements are typically clustered within a small spatial volume.)
For example,
As seen in the figure, the coarser two stages of delay processing are done in the optical domain by subnetworks 320 and 330. However, the finest stage of delay processing, in which the delay increments are mapped to individual antenna elements, is performed in the electrical domain. Accordingly, each output from stage-2 delay subnetwork 330 is directed to an optical-to-electronic (O/E) converter 500. Devices for performing O/E conversion using high-speed photodiodes, for example, are well known and need not be described here in detail. (Herein, devices for optical-to-electronic conversion as well as devices for electronic-to-optical conversion will be collectively referred to as “optoelectronic devices”.)
The electrical output from O/E converter 500 is directed to electronic phase-shifting device 505. Electronic phase shifters are well known and need not be described here in detail.
The output from phase shifter 505 is directed to radiative antenna element 515, from which it is transmitted as electromagnetic radiation. The signal path from O/E converter 500 to radiative element 515 will typically include one or more electronic amplifiers, which have been omitted to simplify the drawing.
In sub-network 630, after each input signal (i.e., each signal corresponding to one of the individual absorbers 615) has been subjected to optical delay processing, it is shifted onto a common operating wavelength for output from sub-network 630. Accordingly, the output from sub-network 630 is a composite output signal on one operating wavelength. (As noted above, parallel operation is possible in two or more sets of mutually orthogonal operating wavelengths.)
In a like manner, the outputs from a plurality of stage-2 delay networks 630 are collected by stage-1 delay sub-network 620, subjected to still coarser increments of delay, shifted onto a common operating wavelength, and combined into a composite optical signal. The composite optical signal output from stage-1 delay network 620 is directed to receiver 610 for detection and demodulation or other further processing.
By way of example, the WSOD devices in a network having two stages of optical delay might each include 100 waveguides of various lengths to serve as the delay elements. Thus, for example, the coarse WSOD might have waveguides which span 100 ns of delay in 1 ns increments, and the fine WSOD might have waveguides which span 1 ns of delay in increments of 0.01 ns. As noted, electronic phase shifters may be used to provide still liner increments of delay.
With further reference to
The optical signal source, such as source 300, advantageously uses a modulated high-power laser, or alternatively a modulated low-power laser whose output is subjected to optical amplification.
The wavelength-shifting devices may use any of various well-known technologies. One example is provided by a silicon optical amplifier (SOA) wavelength converter. A second example is provided by an electroabsorption modulator (LAM) device.
The EAM device can be used as a wavelength converter by converting the optical data signal to an RF signal via a high speed photodiode. The electrical output of the photodiode is amplified by RF amplifiers and then applied to the EAM. The data modulation is then applied to CW light from a tunable laser transmitted through the LAM, thereby transferring the data modulation to the wavelength of the CW light.
