High Power Microwave (HPM) weapons and other sources of intense microwave power pose a growing threat to modern RF receivers. To address this problem, all-dielectric photonic-assisted receivers have been proposed and demonstrated. Here, we describe a new configuration of this type with 15 dB better sensitivity over prior designs. The complete lack of metal and electronics in the front-end offers immunity against damage from intense electromagnetic radiation. In this experiment, detection of C band electromagnetic signal at 6.54 GHz with a sensitivity of -112 dBm/Hz is demonstrated.
© 2008 Optical Society of America
The continuing trend towards reduced feature size and voltage in integrated circuits renders modern electronics highly susceptible to damages caused by High Power Microwave (HPM) weapons and other microwave based directed energy weapons. These weapons induce high voltage transient surges of thousands of volts which can punch through the gate insulator in the transistor and can destroy the circuit’s metal interconnects.
To immunize electronic systems against such threats, a new RF front-end technology is needed which would not have the limitations of the conventional techniques aimed at Electro-Static Discharge (ESD) protection. To achieve this, transistors and metal interconnects, the “soft spots” in a conventional receiver front-end, must be eliminated. Such a technology has recently been demonstrated and is based on a dielectric antenna coupled to an electrooptic field sensor [1,2]. In this paper, we report on a new design that optimizes the coupling between the two structures and results in substantial improvement in the receiver sensitivity.
It’s been known that LiNbO3 resonant modulators can be used as RF receivers [3,4] but these devices were based on metallic electrodes and antennas, that is, they did not employ an all-dielectric front-end. In addition, the ability of optics to provide charge isolation was not identified and made use of in [3,4] hence the technology was not targeted towards electromagnetic damage tolerant receivers. Even earlier works were aimed at creating a fast sampler for repetitive electrical transients by sending short laser pulses through a LiNbO3 crystal which is exposed to the electrical signal . The present work is substantially different from electrooptic sampling through the use of the dielectric antenna and the continuous-wave laser modulation. The application is also different - here we make use of charge isolation property of optics to create an electromagnetic pulse-tolerant analog RF receiver, as opposed to an ultrafast sampler of electrical transients.
2. All-dielectric non-electronic radio front-end (ADNERF) technology
The basic concept of this photonic-assisted all-dielectric RF front-end technology is shown in Fig. 1. The dielectric resonator antenna (DRA) in the front-end, functions as a concentrator of incoming electromagnetic field. When the electromagnetic (EM) field excites the resonance of DRA, a mode field pattern is built up inside the structure. The E-O resonator is placed at the location of the peak field magnitude inside the DRA (Fig. 2). The E-O resonator converts the received EM signal to an intensity modulated optical signal which is then carried away from the antenna front-end via an optical fiber. At the remote location, the signal is converted back to an RF signal which is then amplified and processed using conventional techniques.
This front-end design significantly increases the threshold for damage associated with high power microwave signals. The lack of metal interconnects eliminates the one source of failure. In addition, the charge isolation provided by the optical link protects the electronic circuitry. Good sensitivity can be achieved due to signal enhancement provided by the microwave resonance in the DRA and optical resonance in the E-O resonator.
With the laser frequency (f 0) slightly tuned off of the minimum but within a null, RF modulation is achieved as long as the modulation sidebands are located in the adjacent nulls. This implies that modulation frequency (f RF) must be equal to the optical FSR. However, the modulating E-field (E RF) applied to the resonator should not be uniform across the disk otherwise no modulation occurs. To prevent this from happening, the E-O resonator is placed off center from the symmetrical axis of DRA as shown in Fig. 2.
In general, larger modulating electric field inside the EO resonator results in better modulation efficiency. This in turn would lead to better receiver sensitivity.
Due to the resonance effect in the DRA, it can concentrate the electric field of the impinging wave inside its modal volume. These high field regions can be seen in Fig. 2(b) which is the cross sectional electric field pattern of TM011+δ mode.
In the current configuration, the EO resonator is placed inside the DRA (by drilling a slot into the DRA) making use of the high field region inside the DRA. As the electric field penetrates the lithium niobate from DRA material, it gets even larger (by the ratio of εDRA/εLiNbO3 which is about 3) compared to its magnitude inside the DRA resulting in a better modulation efficiency. In particular, simulation also shows that the magnitude of the electric field inside the EO disk in this configuration is much larger than the case when it was placed underneath the DRA.
The DRA material used was a composite ceramic (BaTiO3 based) with relative permittivity εr=100. The dimensions are fine tuned (D = 8mm and h = 7.2mm) to give TM011 + δ resonance at 6.54 GHz to match the free spectral range (FSR) of the optical resonator. A coaxial electric probe is used to excite and monitor the resonant frequency of the structure which is removed after DRA and optical disk are combined. The electro-optic disk is made of a Z-cut LiNbO3 (lithium niobate) with a radius r = 3.2 mm.
In our experiment, light from an external-cavity tunable diode laser operating at 1.55 µm is coupled into and out of the microdisk resonator using a prism. The laser frequency tuned to bias the resonator at the maximum slope of the optical resonance, which occurs when the optical output about 25% of its peak value. The average optical power that reaches the photodetector is 0.510 mW. A horn antenna is used to transmit the RF signal onto the receiver. With horn antenna feed power and the distance to the receiver known, the power impinging on the receiving front-end can be readily calibrated using the standard procedures.
The horn antenna transmits a single tone at 6.54GHz. Figure 3 shows the received signal spectrum with and without the DRA (i.e. with the E-O sensor only). As the DRA is removed from the front-end, signal level vanishes. The observed noise floor of -134 dBm/Hz (-94 dBm over 10 kHz instrument bandwidth) is generated by the RIN of the source laser. The noise floor of -140 dBm/Hz is dictated by the thermal noise of the amplified photoreceiver (New Focus model 1544A).
In order to calculate the receiver sensitivity, we have to sweep the free space power impinging on the dielectric antenna cross-section to get to a point where CNR (carrier-to-noise ratio) at the photodetector becomes unity. This free-space power is obtained by calculating the fraction of the power fed into the horn antenna falling onto the dielectric antenna using a standard model for the radiation pattern of the horn antenna. It turns out that the minimum detectable free-space RF power is -72 dB corresponding to a minimum detectable field = 2.2×10-4 V/mHz1/2. This corresponds to a sensitivity of -112 dBm/Hz, which is 15 dB better than previous works [1,2].
4. Thermal consideration and dielectric strength
The temperature rise due to the high power microwave pulse should be considered so as not to damage the material nor degrade the performance of the system. The dielectric power dissipation in a material is proportional to the dielectric loss factor of the material. A dielectric heating equation relating power dissipation, applied electric field properties and dielectric properties of the material can be derived from Maxwell’s equation ,
Where P is the power dissipation (W m-3), f is the electric field frequency (6.54 GHz), ε o is the permittivity of free space (8.85×10-12 F m-1), ε″ is the dielectric loss factor of the material (ε″=ε′·tanδ) and E is the average electric field strength (V m-1). Equation (1) assumes that the electric field is uniform throughout the material. The sample dimensions, therefore, should be sufficiently small so that the effect of penetration depth, which is inversely proportional to the dielectric loss factor, is negligible. The dissipation of microwave power within the material results in an increase in the temperature of the material. The heating rate is given by the equation,
Where dT/dt is the heating rate (°C s-1), P is the power dissipation (W m-3), ρ is the density of the material (Kg m-3) and Cp is the specific heat capacity of the material (J g-1 °C-1).
For the dielectric resonator antenna of our experiment, power dissipation of Eq. (1) should be calculated based on the peak electric field inside the resonator to give an idea of the maximum temperature rise occurring inside it. Moreover, it should be noted that the ratio of the peak electric field inside the resonator to that of the incident wave is defined as field enhancement factor and was around 20 in our experiment. Therefore Eq. (1) is modified in this fashion:
Where β is the field enhancement factor. On the other hand, incident wave electric field magnitude is related to its irradiance through I=1/2·cεo|Eo|2 giving rise to the new equation for power dissipation:
Now, for a short pulse (Δt) of high power microwave, maximum temperature rise inside the resonator (ΔT) can be written as:
But I.Δt is called single pulse fluence (with the dimension of J/m2) and is a measure of the total energy (per unit area) that is incident on the antenna.
Given that the single pulse fluence in air is limited by the air’s dielectric breakdown threshold, it is possible to estimate the highest power density and energy fluence that can be beamed on a target a given distance away by an HPM source of given size, frequency and pulse width.
Taking the dielectric breakdown threshold into account, the theoretical maximum of single pulse fluence at 300m from the source is around 103 mJ/cm2 for a 1 µs pulse .
The fluence of existing laboratory sources are still several orders of magnitude below this breakdown threshold. Assuming a source of one order of magnitude below this threshold and using Table 1 for the material parameters, Eq. (6) leads us to a peak temperature rise of about 2°C inside the DRA and about 13°C inside the LiNbO3. Both of which are tolerable by the materials. Moreover, this temperature rise inside the dielectric material results in a slight shift in its resonant frequency which may bring it out of the operating frequency range of the HPM source.
The electric field inside the DRA (and the LiNbO3) will be a factor of 103–104 lower than that occurring in the front-end transistor. The dielectric strengths of DRA and LiNbO3 materials are about 105 V/cm [8,9], similar to those in common semiconductors and dielectrics used in integrated circuit manufacturing. However, in a typical integrated circuit, voltage surge appears across the gate insulator of the front-end transistor, which is on the order of 100 nm or less in a high speed transistor. In the dielectric antenna, this dimension is on the order of the free space RF wavelength divided by the square root of the dielectric constant, which leads to a value of about 6 mm (10 GHz signal and a dielectric constant of 36).
An All-Dielectric Non-Electronic Radio Front-end (ADNERF) receiver can be realized by using a dielectric resonator antenna in conjunction with a resonant electro-optic field sensor. The technology addresses the vulnerability of conventional wireless receivers to high power microwave weapons and other sources of intense microwave energy. ADNERF removes the two “soft spots” in a conventional wireless receiver, namely metal interconnects and transistors amplifiers. Although amplifiers eventually do appear in the signal path, they do so after the optical link in which case they are protected by charge isolation provided by the optics.
This work was supported through grants from DARPA and the U.S. Army. We are grateful to Dr. Lute Maleki of the Jet Propulsion Laboratory for providing the LiNbO3 microdisk resonator and to Prof. Tatsuo Itoh and Dr. Daniel Solli of UCLA for helpful discussions.
References and links
1. R. C. J. Hsu, A. Ayazi, B. Houshmand, and B. Jalali, “All-dielectric wireless receiver,” in 2007 IEEE MTTS Int. Microwave Symp. Dig.2007, pp. 221–224.
2. R. C. J. Hsu, A. Ayazi, B. Houshmand, and B. Jalali, “All-dielectric photonic-assisted radio front-end technology,” Nat. Photonics 1, 535–538 (2007). [CrossRef]
3. D. A. Cohen, M. Houssein-Zadeh, and A. F. J. Levi, “Microphotonic modulator for a microwave receiver,” Electron. Lett. 37, 300–301 (2001). [CrossRef]
4. V. S. Ilchenko, A. A. Savchenkov, A. B. Matsko, and L. Maleki, “Sub-MicroWatt Photonic Microwave Receiver,” IEEE Photon. Technol. Lett. 14, 1602–1604 (2002). [CrossRef]
5. J.A. Valdmanis and G. Mourou, “Subpicosecond electrooptic sampling: principles and applications,” IEEE J. Quantum Electron. 22, 69–78 (1986). [CrossRef]
6. R. Meredith, Engineers’ Handbook of Industrial Microwave Heating (The Institution of Electrical Engineers, London, United Kingdom, 1998). [CrossRef]
7. H. K. Florig, “The future battlefield: A blast of gigawatts?,” IEEE Spectrum. 25, 50–54 (1988). [CrossRef]
8. S. Panteny, R. Stevens, and C. R. Bowen, “Characterisation and Modelling of Barium Titanate-Silver Composites,” Integr. Ferroelectrics 63, 131–135 (2004). [CrossRef]
9. T. Tamir, Guided-Wave Optoelectronics (Berlin Heidelberg: Spring-Verlag, 1988). [CrossRef]