Betty Lise Anderson (614-292-1323, anderson@ee.eng.ohio-state.edu)
Stuart A. Collins, Jr. (614-292-5045, collins@ee.eng.ohio-state.edu)
Charles A. Klein (614-292-6808, klein@ee.eng.ohio-state.edu)
Elizabeth A. Beecher (beechere@er4.eng.ohio-state.edu)
Stephen B. Brown (browns@er4.eng.ohio-state.edu)
Implementing true time delays using photonics has received a great amount of interest lately with limited success [1]. Approaches tend to fall into two categories: those using fibers and those using long free-space paths. In the fiber approaches, drawbacks include use of multiple optical switches (expensive and lossy), broadcasting the light over all possible paths at once (enormous insertion losses), or wavelength multiplexing (needs complex tunable lasers). In free-space systems, beam diffraction requires physically large components or multiple expensive spatial light modulators (SLM's) are used. Finally, none of the optical systems reported have included amplitude control for transmit beam shaping and receive pre-processing for null steering.
We propose here a system based on a completely new approach, using a new optical TTD device that a) reduces TTD system complexity by allowing compact, massively parallel control of multiple independent time delays, b) allows design flexibility for both long and short time delays, c) provides for amplitude control in receive pre-processing for nulling countermeasures, d) has the option of ruggedized solid block construction, e) has few physical components and does not require fiber couplers, and f) supports multiple beam antennas.
The basic device is designed to feed the individual antenna elements. One device provides the required independent time delays and amplitude control for multiple antenna elements, to direct the beam in the transmit mode and in the receive mode to follow targets and null out interference. The device itself is controlled by input from the beamforming computer processor.
The material in this paper has been submitted in an Invention Disclosure, "Device for producing electronically/optically controlled incremental time delays," to the Ohio State University Office of Technology Transfer, disclosure #95ID45U.
The switching mirror operates in conjunction with the polarizing beamsplitter. After the desired number of passes the light is imaged to a certain spot on the SLM. The SLM then changes the polarization of the light at that spot; when this happens the polarizing beamsplitter reflects the light out of the cell. By choosing the correct beam spot on the SLM, the optical path length can be selected for the corresponding optical delay. The spatial light modulator itself can be either optically or electronically addressed, and each path is independently controlled, since no spots overlap.
There are two true time delay implementations of this basic configuration. The additive delay device (ADD) is described first; in this scheme the total delay for a given beam can be any integer multiple multiple of the single pass transit time. The differential delay device (DDD) is appropriate for very small delays and will be described second.
In the additive delay device, the scheme shown in Figure 1, a beam is introduced into the cell by focusing it onto an input turning mirror. This beam expands to just fill mirror M1, which focuses the beam back to a spot on the spatial light modulator. The beam is reflected and expands again to fill mirror M2, whereupon it is again refocused to a new spot on the SLM. If the polarization is not changed by the SLM, the beam returns to M1 and so forth. Each pass through the cell results in an independent spot on the SLM. The maximum number of delay increments is determined by the spot size, mirror orientation, and the SLM resolution; upwards of 64 possible delays are expected.
A key advantage of the design is that there can be many independently controllable paths in the cell (again, more than 64); each of these corresponding to an individual antenna element in the array. Figure 2 shows the spot arrangement on the front of the SLM. Spots from various input beams will be interleaved, bouncing from left to right as well as top to bottom on the SLM surface. The individual delays are controlled by selecting the appropriate pixel on the SLM. The "x's" show the locations of spots for a particular light beam.
Amplitude control for each independent path is achieved by only partially changing the polarization of that beam before turning it out of the cell. Thus some controlled fraction of the power is sent to the detector.
In the second scheme (DDD), very small delays can be obtained using a differential approach. In this case, a second pair of spherical mirrors is added as shown in Figure 3. The second "cell" has a different optical path length- either due to different physical distances to the mirrors or to a medium of different refractive index from the first cell. Instead of removing the light after some number of passes, all beams go all the way through the device, being finally removed by an output turning mirror (the turning mirrors are shown in Figure 2). The SLM is used to control how many of N total passes are through the shorter cell and how many through the longer cell. Using this scheme, very small differential delays between the various beams (tens of picoseconds) are possible. Again, the amplitudes are controlled by varying the polarization of each output beam before it is sent to a polarization analyzer.
In Figure 3 we have also shown a ruggedized solid block construction for robustness; the spherical mirrors are ground onto the outer surfaces of the optical material, such as glass. There are various spatial light modulator technologies available [3], with different switching speeds. Response times of 50 us have been measured for ferroelectric liquid crystals (FLC's) [4, 5], and response times as fast as 13 us are predicted [6]. The FLC's have the low loss associated with liquid crystals and can be made in arrays as large as 256 x 256. Contrast ratios of 70:1 are measured with greater values expected [7].
We have constructed a laboratory proof-of-concept demonstration ADD device, using discrete optics in the arrangement shown in Figure 4. The light source was a 3 mW Argon laser, which was gated using an electro-optic shutter. The light pulses were introduced into the cell via the input turning mirror, and the number of bounces determined by selecting the appropriate pixel on the liquid crystal light valve. This particular light valve is optically addressed, in this case with an incandescent lamp. Figure 5 shows superimposed scope traces of the leading edges of several pulses delayed by the ADD. This plot shows an average delay increment of ~7.1 ns, which agrees with the predicted value within experimental error. The first pulse represents the pulse passing through the ADD with no delay. The subsequent pulses show delays of ~7.2 ns, ~14.0 ns, and ~21.3 ns, respectively. The lower trace is the trigger pulse. The actual time delay size is dependent on the physical component parameters and is easily scaleable.
For complete antenna integration, either of our TTD devices has straightforward interface requirements for electronic/optical conversions. The ADD and DDD photonic devices use separate optical inputs for or from each element for transmit or receive respectively. The light sources can be of any wavelength and either coherent or incoherent; they need only be polarized.
An example of a potential antenna system configuration incorporating our TTD device is shown in Figure 6. Figure 6 (a) shows the transmit case, where N identical RF modulated pulsed light beams are focused into the TTD device. Each input beam is modulated with the RF signal. It is also possible to modulate one laser beam and then use a 1 X N splitter to generate the N beams. The beams are given N individual time delays corresponding to the desired steering angle. Each independently time-delayed beam is detected by a photodetector that drives a transmit module (amplifier and antenna array element). Amplitude control allows beam shaping.
Figure 6 (b) shows the receive case, where N array element signals modulate N independent light beams, which are input into the TTD device. The beams are individually delayed to correspond to the desired receive angle. Amplitude control can be used for null steering during this receive pre-processing operation.
We have described two photonic true time delay devices capable of providing independent delays for at least 64 parallel beams, with greater than six bit resolution. Amplitude control for beam shaping and receive nulling is integrated into the delay device. The Additive Delay Device is most appropriate for longer delays, greater than a nanosecond, while the Differential Delay Device is best suited for short delays of picoseconds up to the subnanosecond range.
Experimental results for the ADD device were presented, showing controllable delays in increments of approximately 7.1 ns. The actual delay increment is determined by mirror choice.
There are several additional advantages offered by these TTD devices. Reduced system complexity has been stated as one of the most important of these requirements for prospective photonic solutions [8] . The ADD uses only one spatial light modulator, one polarizing beamsplitter, one lens, and two mirrors in a compact ruggedized unit. The DDD approach requires only an additional pair of mirrors. Switching times on the order of 13 ms should be achievable. The light source requirements are flexible. Either of these two designs can provide photonic true time delay for the high performance wideband phased array antennas of the future.
