Experimental and Theoretical Results with Plasma Antennas Igor Alexeff, Fellow, IEEE, Ted Anderson, Sriram Parameswaran, Eric P

166 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 34, NO. 2, APRIL 2006 Experimental and Theoretical Results With Plasma Antennas Igor Alexeff, Fellow, IEEE, Ted Anderson, Sriram Parameswaran, Eric P. Pradeep, Student Member, IEEE, Jyothi Hulloli, and Prashant Hulloli

Invited Paper

Abstract—This report is a summary of an extensive research pro- passes through the same reflector. The idea is that a gram on plasma antennas. We have found that plasma antennas are plasma antenna can be so configured that a high-fre- just as effective as metal antennas. In addition, they can transmit, quency, electronic-warfare signal can pass through the receive, and reflect lower frequency signals while being transparent to higher frequency signals. When de-energized, they electrically antenna without appreciable interaction, while the an- disappear. Plasma noise does not appear to be a problem. tenna is transmitting and receiving signals at a lower frequency. Index Terms—Active antennas, antennas, plasma antennas, plasma devices. 6. Mechanical Robustness: We have developed two kinds of robust plasma antennas. In one design, the glass tubes comprising the plasma antenna are encapsulated in a di- I. INTRODUCTORY SUMMARY electric block. In a second design, the plasma antennas are E have had the following experimental demonstrations composed of flexible plastic tubes. We have found that the Wof plasma antennas. Most of these demonstrations are plasma does not damage the plastic tubes over periods of documented on videotape, and are available on request. several hours if the plastic tubes are kept cool. Heat, not 1. Transmission and Reception: We have demonstrated plasma, causes damage to plastic. Mechanical Reconfigurability transmission and reception of operating plasma antennas 7. : We have been able me- over a wide frequency range (500 MHz–20 GHz). The chanically to manipulate the operating plasma antenna surprising results were that the efficiencies are compa- composed of flexible plastic tubes. In particular, we have rable to a copper wire antenna of the same configuration, designed a plasma antenna that may be compressed and and the noise level seemed comparable with a wire an- stowed when not being used. Plasma Waveguides tenna. The noise measurements will be repeated with a 8. : We have demonstrated a coaxial precision noise meter. plasma waveguide. The advantage of such a waveguide is 2. Stealth: When de-energized, the plasma antenna reverts that it reverts to dielectric tubes when de-energized, and to a dielectric tube which has a small radar scattering cross does not have large RADAR cross section. Noise Reduction section. 9. : We have found that plasma-generated 3. Reconfigurability: At 3 GHz, we have demonstrated a noise is in general not a problem. However, to further parabolic plasma reflector. When energized, it reflects the improve the system, we have discovered several new radio signal. When de-energized, the radio signal passes methods of noise reduction. freely through it. 4. Shielding: The plasma reflector, when placed over a re- II. REVIEW OF PREVIOUS RESULTS ceiving horn and energized, prevents an unwanted 3-GHz The first phase of the plasma antenna project started with the signal from entering. When the antenna is de-energized, idea of a coaxial plasma closing switch, shown in Fig. 1. the signal passes through freely In this switch, the outer conductor was a metal shell, and the 5. Protection from electronic warfare: We have demon- inner conductor was a plasma discharge tube. When the tube strated that with a plasma reflector operating and re- was not energized, the outer shell comprised a metal wave- flecting a signal at 3 GHz, a signal at 20 GHz freely guide beyond cutoff, and no radiation was transmitted. When the plasma discharge tube was energized, the apparatus became Manuscript received August 31, 2005; revised January 24, 2006. a coaxial waveguide, and transmission of radio signals was ex- I. Alexeff, E. P. Pradeep, and J. Hulloli are with the University of Tennessee, Knoxville, TN 37996 USA. cellent. The work was done by W. L. Kang, as a thesis project, T. Anderson is with Haleakala Research and Development, Inc., Brookfield, and was presented at a scientific meeting. MA 01506 USA. The second phase of the research started when researchers S. Parameswaran is with Williams-Sonoma Inc., Memphis, TN 38118 USA. P. Hulloli is with Dell, Inc., West Chester, OH 45069 USA. at the Patriot Scientific Corporation, Carlsbad, CA, read of our Digital Object Identifier 10.1109/TPS.2006.872180 work, and called me in as a consultant. They had an ongoing

Fig. 1. Coaxial plasma on switch.

Fig. 3. Early plasma reﬂector.

Fig. 4. Plasma antenna. Fig. 2. Early plasma antenna. well as the parabolic reflector shown in Fig. 3. With this appa- plasma antenna project, in which they wanted to use the dis- ratus, we demonstrated stealth, reconfigurability, and protection appearing feature of the plasma antenna to prevent ringing on from electronic warfare. signal turnoff. Their problem was poor plasma antenna trans- The fourth phase of research was done at the Malibu Re- mission and reception. A version of the first plasma antenna search Corporation, an antenna design facility in Camarillo, CA. is shown in Fig. 2. My investigation showed that under their We felt that precision measurements were required in a proper conditions of operation, the plasma antenna’s resistance was a facility. In Fig. 4, we show a plasma antenna installed in an megohm, and so did not match the 300 resistance of space. electrical anechoic chamber. Also shown is a metal antenna de- The solution was to pulse the plasma antenna to higher currents, signed to be an identical twin to the plasma antenna. The mi- as the plasma discharge has a resistance that decreases with in- crowaves are generated by a line antenna, focused in one di- creasing current. Under the proper conditions, we found that mension by the metal pillbox, and focused in the second dimen- the plasma antenna transmitted and absorbed radiation virtually sion by either the plasma antenna or a metal twin. The results identically to a metal antenna. In addition, the plasma-generated were remarkably successful, as shown in Fig. 5. First, when the noise appeared to be rather low. plasma antenna was on, the transmission efficiency was virtu- The third phase of research started at the University of Ten- ally identical to the metal antenna. Second, the radiation pattern nessee, Knoxville. The work was transferred to the ASI Tech- was also quite similar to the metal antenna. Third, the noise was nology Corporation, Henderson, NV. At the University of Ten- not particularly worse for the plasma antenna over the metal an- nessee, we constructed the plasma antenna shown in Fig. 2, as tenna. However, when the plasma antenna was de-energized, the 168 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 34, NO. 2, APRIL 2006

Fig. 7. Plasma waveguide.

the mast of a ship, yet become transparent to radiation when de-energized. Fig. 5. Radiation pattern. Radar signals would pass through a de-energized waveguide rather than be reflected. In fact, these waveguides could pass in front of operating antennas and be virtually invisible when off. A third plasma antenna application is reconfigurability. The effects of a reconfigurable plasma filter are shown in Fig. 8. In one oscilloscope trace, we observe several spectral lines emitted from an oscillator driven to a nonlinear limit. In the second oscilloscope trace, several of the higher-frequency lines have been removed by the energizing of a plasma interference filter placed between the transmitter and receiver.

IV. RECENT RESULTS We have made remarkable discovery in the operation of plasma antennas (patents pending). In the past, our plasma tubes were ionized by direct current (dc). However, if the tubes are ionized by extremely short bursts of dc, we find the Fig. 6. Embedded plasma antenna. following remarkable improvements. The plasma is produced in an extremely short time—2 s. However, the plasma persists reflected signal dropped by over 20 dB! In other words, the re- for a much longer time—1/100 second. This is the reason why flected signal dropped by over a factor of 100. fluorescent lamps can operate on 60 or 50 Hz electric power. For stealth projects, the first metal reflector could be incased Consequently, if the pulsing rate is increased to 1 kHz, the inside the body of a structure. However, this project is really a tubes are operating at essentially constant density. There are proof-of-principle, rather than a deployable system. three benefits to this new mode of operation. First, the exciting current is on for only 2 s, while it is off for III. OTHER PLASMA ANTENNA PROJECTS 1 ms. Consequently, the discharge current is only on for 0.2% of One of the criticisms directed at the plasma antenna is that it is the time, so current-driven instabilities are not present for most fragile. As a researcher from another company told us, he built of the time. However, the current-driven instabilities in general a glass plasma antenna, but it was no good, because it broke have proven to be not serious. when he installed it underneath an airplane. To make a robust In general, operating the plasma tubes in the noncurrent-car- plasma antenna, we imbedded one in an epoxy block, as shown rying, afterglow state should produce considerably less noise in Fig. 6. This imbedded antenna transmits and receives quite than in operating in the current-carrying state. The decrease in well, and has survived several years of hard treatment. plasma noise is obvious, but detailed measurements have been A second, antenna related, plasma application is a plasma deferred till later in the program. waveguide, as shown in Fig. 7. Here we have an inner con- Second (this was unexpected), the plasma density produced ductor comprising one plasma tube surrounded by an outer shell by the pulsed-power technique is considerably higher than the of eight plasma tubes. When on, the structure transmits radia- plasma density produced by the same power supplied in the tion almost as well as a coaxial cable, but when off, the trans- steady-state. This observation produces two beneficial results. mitted signal decreases by over 100 dB—a factor of . Such We can operate at much higher plasma densities than before plasma waveguides could convey radiation to the antennas on in the steady-state without destroying the discharge tube elec- ALEXEFF et al.: EXPERIMENTAL AND THEORETICAL RESULTS WITH PLASMA ANTENNAS 169

Fig. 10. Experimental apparatus.

Fig. 8. Plasma ﬁlter.

Fig. 11. Pulsing apparatus.

The two signals are received by a horn antenna outside the ring of fluorescent lamps. Fig. 11 is a schematic of our pulsing apparatus. A 0–30 KV supply is connected to a resistance-capacitance (RC) supply Fig. 9. Signals from the two transmitters. comprising 1.5-M resistor feeding a nanofarad capacitor. The resultant high voltage arc over a spark gap to provide pulsed current to the fluorescent lamps—up to 12 wired in series. We trodes. Formerly, using commercial fluorescent tubes, we were find that when operating arc high pulsing frequencies—1 kHz limited to steady-state operation below 800 MHz. Now we can up—the spark gap tends to go over to a steady-state arc. To operate at several gigahertz. The upper frequency limit has not prevent the steady state arc, a small blower is placed on the been explored. spark gap to flush out the ionized air. In practice, this solution We can operate at much higher plasma densities using lower works very well. The Fig. 12 photograph shows the pulsing average consumption. This results both in much lower power plasma tubes and the receiving horn. consumption and reduced heating of the antenna structure. Our In conclusion, our recent inclusion of a pulsed power supply new results are shown in the following figures. for our plasma tubes provides reduced noise, higher steady-state In Fig. 9, the lower trace shows signals from two transmitters dc plasma density, and reduced power consumption. There are at 1.7 GHz and 8 GHz passing through a de-energized plasma possibly minor problems because of a slight plasma density fluc- barrier. The upper trace shows the 1.7-GHz signal being blocked tuation during the pulsing cycle. These problems will be ad- by the energized plasma barrier while the 8-GHz signal is able dressed in future work. to pass through. We note that the received signal with the plasma window on shows considerably more noise than that with the plasma V. P LASMA FREQUENCY SELECTIVE SURFACES window off. Surprisingly, we found that this noise signal ap- Anderson developed a theoretical model of a plasma fre- parently is not present on the microwave signal, but primarily quency selective surfaces and Alexeff provided the experiment. is due to receiver pickup via the power line. Disconnecting the This research is focused on using plasma as a substitute for receiving antenna from the panoramic receiver does not change metal in a frequency selective surface (FSS). FSS have been the observed noise level. used for filtering electromagnetic waves. Each FSS layer has to In Fig. 10, we show the present experimental apparatus. Two be modeled using numerical methods and the layers are stacked transmitting antennas are used with a ring of fluorescent lamps. in such a way to create the desired filtering. Genetic algorithms 170 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 34, NO. 2, APRIL 2006

Flouquets Theorem to connect the elements. We determined the transmission and reflection characteristics of the plasma FSS as a function of plasma density. We utilized frequencies from around 900 MHz to 12 GHz with a plasma frequency around 2 GHz. We pulsed plasma tubes to continuously vary the plasma density and observed the tunability of the reflection and transmission of electromagnetic waves. As the plasma density decays, the amount of transmitted electromagnetic energy increased as expected. However, at electromagnetic signals at frequencies well above the plasma frequency, the plasma FSS was transparent. We also rotated the polarization of the transmitting antenna by 90 and produced a similar but reduced effect. We modeled an array of plasma frequency selective surfaces. Similarly, we made the plasma FSS in the laboratory. Our Fig. 12. Pulsing plasma tubes and receiving horn 20. theory and experiment were in close agreement. A comparison of the theoretical and experimental results is given in Fig. 13. One reason that the theoretical and experimental results are not closer is that the theoretical results were for an infinite surface and the experimental results were for a finite structure. The plasma FSS is unique and new to the field of electromagnetic filtering. Others have developed FSS filters using metal and dielectrics, but we are the first to use plasma and the reconfigurability that it offers. The potential payoff for this technology is high and the risk is moderate. It is moderate since we have developed plasma antennas with transmitters, but the plasma FSS is in some ways easier to develop since they do not require transmitters. The plasma FSS can shield antennas, military electronics, and radar systems in a tunable way. If no shielding is needed, turning the plasmas off causes the shield to be invisible. Plasma FSS allow users to filter out any undesirable radiation, but at the Fig. 13. Comparison between theoretical and experimental results for a same, time enabling operations outside that band. The poten- frequency selective surfaces. tial for technology transfer is significant since the plasma FSS can be tuned to filter out unwanted radiation from commercial are used to determine the stacking needed for the desired products or tuned to filter electromagnetic emissions to meet filtering. This is a complicated and numerically expensive FCC EMC requirements. process. We developed a method to replace metal in a FSS with plasma elements. Our plasma FSS can be tuned to a desired VI. ORGANIZATIONAL DETAILS filtering by varying the density in the plasma elements. This could save much of the routine analysis involved in the standard We have incorporated our company, Haleaka Research, analysis of conventional FSS structures. The user simply tunes Brookfield, MA, in which Anderson is President, and Alexeff is the plasma to get the filtering desired. Plasma elements offer the Chief Scientist. The present work is being supported by SBIR possibility of improved shielding along with reconfigurability and STTR grants from the U.S. Army, the U.S. Navy, and the and stealth. Plasma FSS can be made transparent by turning the U.S. Air Force. The experimental research is being carried plasma off. This extends our previous scientific achievements on in the Electrical Engineering Laboratory, the University in the development of the plasma antenna. of Tennessee. In addition, we have consulting work with the As the density of the plasma is increased, the plasma skin Malibu Research Corporation, Camarillo, CA. depth becomes smaller and smaller until the elements behave The results of our work can be summarized as follows. as metallic elements and we create filtering similar to FSS with 1) Concerning plasma antenna efficiency on transmission metallic elements. Up until the metallic mode for the plasma, and reception, the plasma antennas appear to work just our theory and experiments showed that the plasma FSS had a as well as metal antennas as long as two criteria are continuous change in filtering. We developed a basic mathemat- met. a) The plasma density is sufficiently high—the ical model for a plasma FSS by modeling the plasma elements plasma frequency should be considerably above the radio as half wavelength and full wavelength dipole elements in a frequency being used. b) The coupling to the plasma is periodic array on a dielectric substrate. The theoretical model optimized. The measurements confirming these state- with numerical predictions predicted results in good agreement ments were made at three separate locations by three with our experiments on the plasma FSS. Theoretically we used different, independent, groups. The first group was at ALEXEFF et al.: EXPERIMENTAL AND THEORETICAL RESULTS WITH PLASMA ANTENNAS 171

the Patriot Scientific Research Corporation, San Diego, [11] I. Alexeff et al., “Advances in plasma antenna design,” in IEEE Int. Conf. CA; the second group was at the Electric Boat Corpora- Plasma Sci., Monterey, CA, Jun. 20–23, 2005, p. 88. [12] Kostrov, Pakhotin, and Smirnov et al., Sov. J. Plasma Phys, vol. 21, no. tion, Groton, CT (U.S. submarines), and the third group 5, p. 435, 1995. was The Malibu Research Corporation, Calabasas, CA [13] K. Markov and Smirnov et al., Sov. JTP Lett., vol. 15, no. 5, p. 34, 1989. (antenna development). In the Groton experiments, the plasma antenna communicated between two computers transmitting at 900 MHz. All experiments showed that Igor Alexeff (M’72–SM’76–F’81) received the plasma antenna elements performed within a few B.S. degree in physics from Harvard University, dB of a metal antenna. Recent results using plasma Cambridge, MA, in physics in 1952, and the Ph.D. antennas in the pulsed mode have transmission and re- degree in nuclear physics from the University of Wisconsin, Madison, in 1959. ception taking place in noncurrent carrying plasma. He is a Professor Emeritus with the University 2) Concerning noise in plasma antennas, in the experiments of Tennessee, Knoxville. He has been working in listed above, noise was never a problem. At most, noise plasma and microwave engineering for over 50 years. He has a patent on the Orbitron Microwave was a few decibels above the metal antenna, which was Maser that has operated up to one Terahertz (1/3 always used for reference. In the Groton experiments, the mm). He is an author and co-editor of the book High plasma antenna was compared to a metal antenna in an Power Microwave Sources (Norwood, MA: Artech House). He has over 100 refereed publications and over 10 patents. He has spent considerable time electrical anechoic chamber. In the Malibu experiments, recently on plasma stealth antennas, and is listed on several patents issued to the plasma reflector also was compared to a metal reflector the ASI Technology Corporation. He has worked at the Westinghouse Research in an anechoic chamber. In both cases, the noise level was Laboratory on nuclear submarines, at the Oak Ridge National Laboratory in controlled thermonuclear fusion, and at the University of Tennessee in a few decibels above the metal antenna. industrial plasma engineering. He has worked overseas for extended periods in 3) Concerning the plasma antenna igniting in a pulsed RF Switzerland, Japan, India, South Africa, and Brazil. field, the answer is most interesting. The plasma may or Dr. Alexeff was a co-founder of the IEEE Nuclear and Plasma Sciences So- ciety. He was President of that society in 1999–2000. He is a Fellow of The may not ignite in an intense RF field, depending on the American Physical Society. He also passed the Tennessee State License Exam, design of the plasma antenna! and is a registered professional engineer. 4) Concerning the nonlinear effects of plasma antennas at high-power, no nonlinear effects have been observed so far, but the investigation is still ongoing. Ted Anderson received the Ph.D. degree in physics from New York University in 1986. VII. CONCLUSION He is currently a Research Professor in the Department of Electrical and Com- puter Engineering, University of Tennessee, and CEO of Haleakala Research and Development, Inc., Brookfield, MA. He has done research and published The plasma antenna appears to work well and is becoming in the areas of the foundations of quantum mechanics, atomic physics, fluid dy- the object of study in several laboratories. Plasma antennas are namics, plasma physics, and antenna physics. While working for the U.S. Navy, useful in applications for stealth, reconfigurability, and protec- he invented ten patents on the plasma antenna. Since 1999, he has worked with Prof. Alexeff on plasma antennas. Together, they have invented several patents tion against electronic warfare. on the plasma antenna and plasma waveguides, and they presented several pa- pers at various conferences, which include the IEEE and AIAA organizations.

REFERENCES

[1] G. G. Borg, J. H. Harris, D. G. Miljak, and N. M. Martin, “The applica- Sriram Parameswaran received the B.S. degree in tion of plasma columns to radio frequency antennas,” Appl. Phys. Lett., electrical engineering from University of Madras, vol. 74, pp. 3272–3274, 1999. Chennai, India, the M.S. in electrical engineering [2] G. G. Borg, J. H. Harris, D. G. Miljak, D. Andruyezik, and N. M. Martin, and the M.B.A. degree in logistics from the Univer- “Plasmas as radiating elements,” IEEE Trans. Plasma Sci., submitted for sity of Tennessee, Knoxville, under the guidance of publication. Dr. I. Alexeff. [3] J. Rayner, A. Whichello, and A. Cheetham, “Physical characteristics of a He worked under various projects, which include plasma antenna,” presented at the 11th Int. Conf. Plasma Phys., Sydney, plasma sterilization, ball lightning, and plasma Australia, 2002. antennas. He is currently a Project Engineer with [4] M. Hargreave, J. P. Rayner, A. D. Cheetham, G. N. French, and A. P. Williams-Sonoma Inc., Memphis, TN. Whichello, “Coupling power and information to a plasma antenna,” pre- Mr. Parameswaran received the IEEE Nuclear and sented at the 11th Int. Conf. Plasma Phys., Sydney, Australia, 2002. Plasma Sciences Society graduate scholarship award for the year 2004. [5] A. Cheetham, J. Rayner, B. Gilbert, and G. French, “Software wave ex- citation for plasma antenna applications,” in Proc. 23rd AINSE Conf. Plasma Sci. Technol., Adelaide, Australia, 2000, pp. 17–19. [6] M. Moisan, A. Shivarova, and A. W. Trivelpiece, Phys. Plasmas, vol. 24, p. 1331, 1982. Eric P. Pradeep (S’05) was born in Coimbatore, [7] C. Balanis, Antenna Theory. New York: Wiley, 1997, pp. 165–173. India. He received the B.S. degree in electrical [8] G. G. Borg, J. H. Harris, N. M. Martin, D. Thorncraft, R. Milliken, D. G. and electronics engineering from the University of Miljak, B. Kwan, T. Ng, and J. Kircher, “Plasmas as antennas: theory, ex- Madras, Madras, India. He is currently working periments and applications,” Phys. Plasma, no. 5, pp. 2198–2202, May toward the M.S. degree at University of Tennessee, 2000. Knoxville. [9] I. Alexeff, “Plasma antennas,” presented at the SMi 8th Annu. Conf. He has been working for Dr. Alexeff as a Grad- Stealth, London, U.K., Mar. 15–16, 2004. uate Research Assistant in the Microwave and Plasma [10] , “Plasma antennas,” presented at the SMi 9th Annu. Conf. Stealth, Laboratory for about two years. Apr. 11–12, 2005. 172 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 34, NO. 2, APRIL 2006

Jyothi Hulloli received the B.S. degree in electrical Prashant Hulloli received the B.S. degree in me- engineering from Karnataka University, Karnataka, chanical engineering from Karnataka University, India. She is currently working toward the MBA de- Karnataka, India. After working for a year with gree at the University of Tennessee, Knoxville. leading automobile giant in India, he received the After receiving the B.S. degree, she worked as a M.S. degree in industrial engineering from Univer- Research Associate at one of the top engineering col- sity of Tennessee, Knoxville and the MBA degree in leges of India. She is currently working as a Gradu- supply chain management and marketing from the ated Assistant with Dr. Alexeff to assist in developing same university in 2005. business plans for plasma devices. While working toward the MBA degree, he worked as Graduate Assistant with Dr. Alexeff to assist in developing business plans for plasma devices. Currently, he is with Dell Inc., as Supply Chain Manager/Consultant at West Chester, OH facility.