Identifying and Isolating Signals Using Radio Frequency Photonics

Figure 1: Example of an RF photonic circulator with a signal identifier and separator.

A single antenna can be used for both transmission and reception. To accomplish this, the transmission must be isolated from the reception. In Figure 1, a radio frequency (RF) circulator is connected right after the antenna. The three-port device separates the transmit path from the receive path. After the circulator, a system can be used to identify the frequency of different signals. Once the frequency has been found, a filter with the right pass-band frequency can be used to isolate signals from each other.

RF photonics can be used for RF circulator, frequency identification, and filters. The photonic filters are tunable and narrow. A photonic-based circulator isolates the transmit path from the receive paths. Multiple photonic methods can be used to identify the frequency of the signal. This article discusses these methods for identifying and isolating signals using RF photonics.

The Need for Signal Identification and Isolation

Figure 2: Block diagram of a spectrum analyzer.

New signals continue to fill the available spectrum. Amateur radio and television signals fill the high-frequency (HF), very-high-frequency (VHF), and ultra-high-frequency (UHF) bands. Air Traffic Control (118-136 MHz) and emergency radio communications (138-144 MHz) also use VHF. In the UHF band, 400-MHz frequency is used for time and frequency standard transmission to satellites, while wireless phones use the 900-MHZ frequency band. Above 1 GHz, phones and Wi-Fi also use the 2.5- and 5-GHz bands. Commercial GPS uses the 1.2- and 1.5 to 1.6-GHz frequencies. The 2.7 to 2.9-GHz band is used for Airport Surveillance Radars. As these signals increase usage, isolation is required. Filters can separate out these various signals. Before separation, the frequency of the signal has to be determined in order to set the filter appropriately.

Multiple methods exist for determining the frequency of signals. One commonly used method is the electrical spectrum analyzer. A block diagram of an electronic spectrum analyzer is shown in Figure 2. The signal is low pass filtered (LPF) and then mixed with a local oscillator (LO) from a voltage-controlled oscillator (VCO). The resulting intermediate frequency (IF) is amplified (IF Amp), passed through a band-pass filter (BPF), and detected (DET). Finally, the signal is displayed. The ramp generator sweeps the LO and syncs the output to the display.

RF Photonics for Signal Identification

Figure 3: RF photonic version of a spectrum analyzer.

RF photonics can play a role in frequency identification. Multiple methods exist for accomplishing this task.

A photonics-based spectrum analyzer replaces the electronic components with photonic components. An optical modulator takes the place of the mixer, and a Fabry-Perot (FP) filter now scans, rather than the local oscillator. The resulting output is then sent to a photodetector where a display of power as a function of RF frequency can be obtained. The block diagram is shown in Figure 3.

Another type of photonic spectrum analyzer has been developed based on rare earth doped crystals. The absorption of the crystal can be modified by a laser. Figure 4 shows two laser beams at different angles to the surface of the crystal. They create an absorption grating on the crystal. A third laser is input to the crystal from the opposite side of the other lasers. The absorption grating set up in the crystal then deflects the beam with the RF information onto a photodiode array. The deflection of the beam with the RF signals will be precisely mapped to the photodiode array.

Figure 4: RF photonic spectrum analyzer using an optical crystal.

One method for finding the signal frequency involves optical filters. The fixed optical filter has a sinusoidal response. Combined with two lasers of different wavelengths, the signal frequency can be determined. The first laser’s wavelength is set at the null of the response while the second laser is set at the peak. Generated sidebands appear on complementary slopes of the response. The optical power of each wavelength is demuxed and detected. The ratio of powers from the two photodiodes is called the amplitude comparison function (ACF). It can be used to determine the frequency of the signal.

Another approach uses two. Similar to the two-laser case, the setup provides measurement of the signal frequency; however, it does not require two lasers, which can reduce the power consumption.

Another method uses RF fading due to dispersion as a filter. The dispersion-based filter provides an ACF similar to the one in the previous section. A two-laser method makes the ACF – the ratio of the two different frequency responses generates the ACF.

Yet another version of the dispersion-based system was demonstrated. A single laser is followed by a dispersive element and a photodetector. Due to the dispersion, a frequency-to-time mapping of the dual optical sidebands occurs. The time delay through dispersion must have a linear response as a function of wavelength. Under this condition, the optical sidebands arrive at different times to the photodetector. The difference in time between the first sideband and its twin is proportional to twice the RF frequency.

A combination of the above methods has also been demonstrated. A tunable laser can be used as one of the two lasers in a two-laser dispersion system. A measurement of the power maps the signal frequency using an electronic processor. This is a combination of the swept spectrum analyzer and the dispersion method.

RF Photonics for Signal Separation

Figure 5: Example of an FIR filter using different fiber lengths.

RF photonics can also provide signal separation. Once the signal frequency is determined, a filter can be centered on the signal. The filter can be realized in different ways. One is simply a bandpass filter. Others can use finite impulse response (FIR) to create different filter shapes.

A photonic filter is often used to filter RF signals. The most common metric for filter is the quality factor. Optical filters have been realized in many different ways. The Fiber Bragg grating (FBG) filter is one used frequently. The filter is designed to act as either a bandpass (in reflection) or a notch filter (in transmission).

Another method for generating an RF filter is the use of either a FIR or an infinite impulse response (IIR) filter. The FIR filter is simply the discrete convolution sum of the sampled impulse response of a given filter shape with multiple time-delayed versions of the signal. The IIR filter is the same as the FIR filter, but instead of a finite set of delays, the delays are modeled to continue forever.

Different ways exist to realize an FIR filter. Figure 5 shows an optical source with different optical carriers connected to a modulator. The RF signal appears on each wavelength. A demux separates the different wavelengths into parallel paths. Each path is attenuated and passed through a multiple of one time period delay. The signals are combined with a mux connected to a photodiode. The photodiode sums up the delayed RF signals.

In another method, a multiple wavelength source and modulator are used. The output is connected to a fiber with multiple Bragg gratings. The gratings are spaced by a delay of T/2, providing an integer multiple of delays for each wavelength. The reflected wavelengths appear on a photodiode. An IIR filter can be realized simply by using a feedback loop of a fixed delay. In this case, the signal will ideally be a summation of an infinite number of delay round trips. While this is hard to realize in the electronic domain, the low loss of fiber can provide multiple round trips without a large amount of loss.

Another method to measure the signal frequency uses finite and infinite impulse response filters. A combination of FIR and IIR filters can be used to identify the center frequency of an RF signal. The FIR filter is generated by splitting the light with a delay in one arm. The IIR filter is implemented by the electronic feedback from the photodiode back to the modulator. The detected power increases versus frequency, and the response is similar to methods shown above.

Conclusion

Various advanced techniques have been demonstrated to improve the performance of the photonic links. Nonlinearities can be overcome by using different modulation formats. Optical fiber limits can also be overcome by using different fiber types and isolators. The noise of the erbium-doped fiber amplifier (EDFA) can be characterized and controlled by proper design to reduce the added noise. Finally, the Mach Zehnder modulator (MZM) can be used at different biases to improve the RF performance.

This article was written by Preetpaul S. Devgan, RF/EO Subsystems Branch, Aerospace Components & Subsystems Division, Air Force Research Laboratory, Wright-Patterson Air Force Base, OH. For more information, visit here .