Flexible Coherent Digital Transceiver for Low Power Space Missions 1, 2

Transcription

1 Flexible Coherent Digital Transceiver or Low Power Space Missions, Christopher B. Haskins, Wesley P. Millard, J. obert Jensen Johns Hopkins University ~ Applied Physics Laboratory (JHU/APL) 00 Johns Hopkins oad Laurel, MD , , chris.haskins@jhuapl.edu, wesley.millard@jhuapl.edu, bob.jensen@jhuapl.edu Abstract A lexible coherent ital transceiver architecture has been developed by the APL in order to enable mission speciic perormance tailoring while maintaining modularity and minimizing program-incurred cost and risk. The new transceiver architecture is based on the heritage X-band transceiver system that is currently integrated into the New Horizons spacecrat. Using this new architecture, a low power, coherent, X-band ital transceiver has been made that meets the requirements or two-way Doppler tracking. The new transceiver contributes less than 0.0 mm/s to the Doppler velocity error measured over a 60-second interval in coherent mode. Secondary power consumption is.8 W in the uplink-only mode o operation including the reerence oscillator. Transceiver designs on the TIMED, CONTOU, and New Horizons spacecrat were noncoherent, which required downlink telemetry in order to support two-way Doppler tracking []. The addition o a coherent capability allows this new architecture to be used on missions where carrier-only Doppler tracking is desired. This paper provides a description o the new transceiver architecture and its demonstrated perormance. TABLE OF CONTENTS. INTODUCTION.... COHEENT ACHITECTUE.... PEFOMANCE EQUIEMENTS PEFOMANCE VEIFICATION CONCLUSIONS AND FUTUE WOK ACKNOWLEDGEMENTS...7 EFEENCES...8 BIOGAPHY...8. INTODUCTION Modern spacecrat radio navigation is typically perormed via two-way Doppler tracking, which is enabled by the use /06/$ IEEE IEEEAC paper #, Version, Updated Dec. 6, 005 o a two-way coherent spacecrat transponder. This navigation technique limits lexibility in the choice o spacecrat communications hardware. Noncoherent navigation techniques [] were developed to enable the use o a wider variety o communications hardware, helping to enable smaller, lower cost satellite missions. Two-way Doppler tracking via telemetry-based noncoherent navigation was irst demonstrated on the TIMED mission [] and implemented on the CONTOU mission []. The New Horizons mission to Pluto and beyond is lying a noncoherent navigation communications system [4] due to uplink radio science requirements and power constraints which were mitigated by the development o a low power X-band uplink receiver [5]. The New Horizons X-Band Digital eceiver consumes. W o secondary power, providing signiicant power savings over commercially available alternatives, while yielding similar perormance and added capability (i.e., regenerative pseudonoise ranging and an uplink radio science instrument). The New Horizons uplink receiver is paired with a noncoherent downlink transmitter, and abricated into two separate integrated electronics module (IEM) cards. Both the uplink and downlink cards share a common requency reerence, the USO. The inclusion o a USO is required on New Horizons to provide radio science capability; it is not a requirement o the communications architecture. Since the uplink carrier tracking signal is not used in the generation o the downlink carrier, the New Horizons transceiver system is noncoherent. To provide or two-way Doppler tracking, additional circuitry measures the uplink requency with respect to the USO and inserts this data into the spacecrat telemetry stream to be downlinked. The Earth station tracks the spacecrat as i it contained a coherent transceiver or transponder and the navigation team then uses a correction actor derived rom the telemetered navigation data. The two-way Doppler tracking perormance is equal to that o a coherent system, assuming that downlink telemetry is available.

2 Figure - Basic block diagram o the carrier tracking circuitry o the coherent transceiver conigured or X-Band Doppler tracking o a spacecrat using noncoherent navigation techniques requires suicient downlink signal to noise ratio (S/N) to transer telemetry rames without errors. Some missions require Doppler tracking capability at times when the downlink S/N is very low (e.g., operation through low gain antennas at long range during orbital operations) or when the downlink signal is carrier-only. The use o a ully coherent transceiver allows two-way Doppler navigation to be perormed using only a carrier. Leveraging rom technologies and ital signal processing techniques incorporated into the New Horizons X-Band Digital Uplink eceiver, a low power two-way coherent X- band transceiver has been developed. The new transceiver requires minimal modiication to the heritage New Horizons X-Band transceiver and meets typical spacecrat Doppler navigation requirements while consuming.8 W secondary power in the uplink-only mode o operation including the reerence oscillator. The new transceiver provides noncoherent, coherent, and radio science modes while maintaining a simple and lexible architecture. An architecture based on requency synthesis and direct ital synthesis (DDS) techniques eliminates the typical constraints o coherent communications transponders by allowing lexibility in the choice o turnaround ratios, channel assignment, and band o operation (e.g. S, X, or Ka-band) without the need or tuning or replacing the reerence oscillator. This lexibility allows the transceiver to be easily tailored to the needs o a given mission without hardware redesign and manuacture (i.e. via sotware or VHDL code changes). The intrinsic nature o a transceiver, as opposed to a transponder, allows the uplink receiver, downlink transmitter, and reerence oscillator to be independently developed, tested, and used in a variety o mission-speciic conigurations. A prototype X-band coherent ital transceiver has been assembled or proo-o-concept using prototype hardware rom the New Horizons noncoherent transceiver system. This prototype transceiver operates in the X-band, though S-Band operation may be achieved by simply reprogramming coeicients in a ield programmable gate array (FPGA) and removing several components. Ka-band band downlink capability may be added by the addition o a multiplier module, which is under investigation. An FPGAbased ital subsystem allows easy incorporation o uture perormance enhancements such as tone-based commanding [6], spectrum analysis via ast Fourier transorm (FFT), MS power detection, and more. The reported power consumption includes an on-board requency reerence, which provides the stability required by many o these uture perormance enhancements. A discussion o the theory behind the new coherent transceiver architecture ollows. The results rom a prooo-concept prototype coherent transceiver test are also presented.

3 . COHEENT ACHITECTUE The New Horizons uplink receiver and the noncoherent navigation enabled downlink transmitter eectively provide a method or spacecrat navigation via two-way Doppler tracking. In order to maintain two-way Doppler tracking capability without the need or noncoherent navigation telemetry, the uplink carrier tracking inormation must be used to control the downlink carrier requency, preerably with a downlink/uplink turnaround ratio o 880/749 to be compatible with ground station inrastructure. Since the uplink carrier tracking loop in the New Horizons uplink receiver is DDS-based, the ital carrier tracking inormation may be processed and passed on to an identical DDS used or downlink carrier generation. It is the processing inside the uplink/downlink interace that removes the onboard reerence contribution to the downlink carrier, making the new transceiver two-way coherent capable. A description o the uplink and downlink sections o the new transceiver ollows. Uplink eceiver Figure provides a block diagram o the new coherent transceiver in an X-Band coniguration. By design o the phase-locked loop in the receiver, the second intermediate requency ( IF ) is orced to re divided by. The eedback requency o the DDS, a, is dependent on the received uplink requency ( X ) and the requency o the reerence oscillator ( re ) as ollows: X re a () Where: M ( ) (or New Horizons) N ; N M N, M, N, and M are divider values or the two uplink requency synthesizers. N and M are divider values or the downlink requency synthesizer. Downlink Transmitter The downlink card architecture (see Figure ) is comprised o a single synthesized requency source, an oset mixer, and a multiplier. The shaded items in Figure illustrate the non-heritage downlink components. The synthesizer reerence requency in the heritage noncoherent downlink system is re, which results in a simple transmit requency equation in terms o the reerence oscillator requency: N ( ); () TX 4 re M Coherent Uplink-Downlink Making the downlink requency coherent with the uplink requency requires that TX is related to X by a constant ratio. To do this, a point on the receiver card carrying both re and X inormation needs to be brought over and mixed into the downlink card. We have chosen to use the actor a rom above to link the two cards together because the ital signal is easily processed and transported. Adding the a term into the downlink card at the reerence to the requency synthesizer was chosen due to circuit similarities in the uplink card. The new equation or TX in terms o re and X is derived as ollows: TX X 4 a' + b' re re () Where: a' ; b' - ( ) From this equation it is clear that b must be zeroed to remove TX dependence on the re requency. However, the ratios that make up b cannot be varied widely, and ideally should not be varied at all to maintain the proper uplink and downlink channels and to maintain a 0 MHz reerence oscillator requency. Thereore, the exact requency control word updating the DDS on the uplink card should not be used to update the DDS on the downlink card. A new DDS requency control word may be derived rom the old word using airly simple mathematics to minimize ital processing. + b a Δ (4) - b b ' + Δ 0 (5) ( ) Δ + (6) ( ) The new requency control word or the downlink card DDS is created rom a and called b. This new word has an extra term Δ that can be varied independent o other terms to zero out b. The Δ term represents the addition o a constant to the requency word coming over to the downlink card rom the uplink card. This is a simple operation to perorm, thereby making the coherent architecture simple to

4 implement. Using this method, the turnaround ratio becomes the ollowing equation: TX X This numeric solution is speciic to the prototype coherent transceiver system, which uses an integer-n downlink requency synthesizer. The turnaround ratio will change based on the DSN channel selected and whether or not ractional-n synthesis is used. The advantage in using ractional-n requency synthesis is that the uplink and downlink carrier requencies may be better centered within the allocated DSN channels. Fractional-N requency synthesis is part o the heritage New Horizons uplink receiver and will be ported to the downlink transmitter as part o the new coherent transceiver design. Error Sources The most signiicant error source in this coherent system is the quantization o Δ. Other quantization and rounding eects exist in the uplink carrier tracking loop; however, these error sources are extremely small and are dominated by system noise and the ollowing error terms. The phase accumulator in the prototype system DDS is bits, leading to a bit representation o Δ. The round-o incurred here causes a small portion o re to contaminate the downlink requency. This error term (Err Δ ) could be as signiicant as - (or part in 0 0 ), creating a ixed velocity oset o 5.8 mm/s. The use o a 48 bit DDS would reduce the worst case possible ixed error to 88 nm/s. The error contribution equation is as ollows: TX ( o + drit ) Δ 4 X + 4 Err The irst term in the error source is the nominal requency re, and the second term is the drit o the reerence oscillator. The nominal value o the reerence oscillator ( o ) will cause a constant oset in the downlink requency that is easily accounted or, and the drit contribution o the reerence oscillator ( drit ) will be insigniicantly small. For the prototype coherent transceiver, the constant oset in TX is ~ 6.6 mhz, or 7.5 parts per trillion, which was veriied during several tests.. PEFOMANCE EQUIEMENTS There are two primary perormance requirements driving the new two-way coherent X-Band transceiver architecture: ) two-way Doppler tracking velocity error o 0. mm/s on a 60-second interval at low S/N and ) low power consumption. The new coherent X-band transceiver consumes.8 W secondary power in receive-only mode, (7) (8) 4 which is a considerable savings over commercially available systems. The two-way rms Doppler velocity error requirement must be converted to a requency stability requirement to be meaningul; the ollowing equation can be used to perorm this conversion [7]: σ σ v (9) c c Where: σ Allan deviation, Hz σ Doppler velocity error, m/s v c speed o c light in vacuum, m/s downlink carrier requency, Hz The minimum achievable velocity error or an X-Band uplink/downlink system is limited by solar phase scintillations or a given Sun-Earth-probe angle as reported in [7], among other eects. For an increasing Sun-Earthprobe angle, this error approaches a minimum or a given integration time. Using the above equation, the Allan Deviation is calculated or the approximate minimum Doppler velocity error due to solar phase scintillations or an X-Band system (Table ); or comparison, the speciied Allan Deviation or the NEA X-band transponder [8] is also included. Inspection o Table reveals that solar corona eects are signiicantly higher than the speciied residual error o the transponder itsel. As a result, the calculated Allan Deviation due to solar corona eects shall be used as an absolute requirement and the NEA transponder speciication shall be used as a goal or the new transceiver. Both meet the typical 0. mm/s over 60 seconds precision requirement o deep space navigation. Table - Allan Deviation equirements or given MS Doppler Velocity Errors Integration Time (s) Velocity Error (mm/s) Allan Deviation (parts) eason E- solar scintillation E-4 NEA transponder spec E- solar scintillation E-5 NEA transponder spec E- Typical nav spec The downlink carrier phase noise contributes to phase error in the receiving Earth station s tracking loop. This phase error, i signiicant, leads to bit error rate (BE) degradation and cycle slipping in the downlink. The eects o transmitter phase noise on the downlink when received by

5 an Earth station may be quantiied using the methods in [9] and [0] through which a downlink carrier phase error speciication was calculated assuming residual-carrier phase modulation (PM) (Table ). BE degradation must be minimized or downlink telemetry transmission, though when operating in a carrier-only downlink mode cycle slipping becomes the limiting actor in speciying allowable phase error. Carrier-only perormance is important during low S/N mission scenarios that preclude downlinking telemetry and that require two-way Doppler tracking o a spacecrat. Further details later in this paper will tie this phase error requirement into a downlink carrier phase noise requirement. E867C uplink signal generator and distributed through the HP 5087A ampliier. The same 0 MHz reerence can also be attached to the Agilent 866A generator on the spacecrat-side to phase-lock the transceiver reerence to the measurement system and perorm baseline measurements in noncoherent and coherent downlink modes. Table - Downlink Carrier MS Phase Error equirements in Ground eceiver Tracking Loop Criteria Max MS Phase Error (degrees) Negligible BE degradation db BE degradation 9. year mean time between cycle 5.4 slips Figure : Proo-o-Concept Transceiver Assembly. In addition to the above requirements, the new transceiver architecture must support a noncoherent downlink mode and regenerative pseudonoise ranging and radioscience modes. Both o these requirements are met by the New Horizons transceiver and are not aected by the modiications required or the ully coherent transceiver. 4. PEFOMANCE VEIFICATION A ully coherent X-band transceiver prototype was assembled or proo-o-concept, using prototype hardware o the New Horizons uplink and downlink cards described above (see Figure ). In this coniguration, the uplink receiver secondary power consumption is. W including the critical command decoder and wideband radio science channel. The downlink transmitter secondary power consumption is 4.7 W; improvements will be made on this or uture missions by applying the technologies already incorporated into the uplink receiver or New Horizons. A second DDS (5 mw) was added to the heritage transceiver system in order to obtain coherency. A proo o concept test was conducted in order to demonstrate the Allan Deviation and downlink carrier phase error requirements described above. The test setup consisted o the demonstration transceiver chassis, a downconverter built rom connectorized F components, and a rack o test equipment. A block diagram o the test system is illustrated in Figure ; the components in red are considered spacecrat systems, the components in black are considered Earth station systems. The Earth station components are linked by a common 0 MHz reerence generated in the Agilent 5 Figure : Transceiver Test System Block Diagram To measure the perormance o the prototype transceiver, an uplink carrier ( X ) is generated by the E867C and transmitted to the uplink receiver. The receiver phase-locks to the uplink carrier, resulting in a coherent downlink carrier requency ( TX ) as deined earlier. The transmitted downlink carrier is downconverted using a connectorized mixer, ilter, and ampliier. The local oscillator or the downconversion is an HP 867B signal generator. The resulting intermediate requency is approximately 0 MHz; this requency is conducive to measuring Allan Deviation, phase noise, and other perormance metrics. Allan Deviation The Allan Deviation o the prototype transceiver system is measured using the time interval analyzer illustrated in

6 Figure. esults rom several key measurements are shown in Figure 4. The dashed lines show the speciications and measured results or the transponder used on NEA. These lines represent our perormance goal or the transceiver. The measurement loor o our test equipment is shown in the solid red line. This result was obtained by directly connecting the E867C to the downconverter box using an 8.49 GHz carrier requency. The solid orange line shows the transceiver Allan Deviation or a strong S/N in the uplink path (same conditions as the NEA transponder measurements). Degradation rom the residual Allan Deviation o the measurement system itsel is primarily due to higher phase noise on the downlink carrier compared to the E867C signal source. The solid green line shows the transceiver Allan Deviation measured under the same conditions using the Deep Space Network Compatibility Test Trailer (CTT). The CTT ground receiver was conigured or a one-sided tracking loop bandwidth o 0. Hz. The downlink provided approximately +5 db-hz carrier to noise power density ratio (Pc/No) to the ground receiver. The ground transmitter provided approximately +0 db-hz Pc/No to the uplink; this corresponds to a received uplink carrier power o -70 dbm at the low noise ampliier input. Only a small portion o the complete curve was assembled due to limited testing time with the CTT. However, the CTT test validates the transceiver test data taken with the Figure test system. Together the green and orange lines show that the new transceiver is capable o meeting and exceeding typical mission requency stability requirements (0. mm/s over 60 seconds) under normal operating conditions (without the presence o 60 Hz signal contamination). A measured Allan Deviation o.6e-4 at 60 seconds is illustrated in this test data, which yields a.9 μm/s Doppler velocity error. Overall these measurements demonstrate very good perormance or a prototype coherent transceiver system, making it acceptable or navigation and communication. Normally the slope o a coherent Allan Deviation measurement should be 0 db per decade, similar to the NEA measured curve. The skew on the slope o our test setup is due to the limited ability o the time interval analyzer to accurately detect the zero crossings o the downlink signal (i.e. a narrowband tracking ground receiver was not available or this test). The lump seen at about 00s is an eect o connecting the entire test system to the same 0 MHz reerence clock. The lump seen at.0s is 60 Hz noise. Figure 4: Measured Allan Deviation or Proo-o- Concept Transceiver Test Phase Noise In order to predict the downlink carrier rms phase error a model o the downlink carrier phase noise was generated in Matlab. The phase noise model includes uplink thermal noise, a model o the uplink carrier tracking loop bandwidth, and uplink and downlink quantization noise. Using the coherent downlink carrier phase noise model a simulation o the downlink phase noise and phase error was completed or varying received uplink signal levels in coherent mode in addition to a simulation o noncoherent mode. The assumptions in the model are as ollows: reerence oscillator phase noise comparable to that used in the proo-o-concept test (Agilent 866A). Ground receiver one-sided loop bandwidth 0.5 Hz. This is used as a worst case with respect to rms phase error estimation. Phase error integration rom 0 - to 0 7 Hz oset. Phase noise was measured on the prototype transceiver or each scenario simulated with the phase noise model. The simulated and measured results are illustrated in Figure 5 and are in good agreement. Figure 5 illustrates that the phase noise model is suicient or use in calculating worst case phase error. The phase error was calculated based on this model or each scenario and is listed in Table. 6

7 dbc/hz X-Band Ground eceiver esidual SSB Phase Noise or APL Coherent Transceiver noncoherent (model) -70 dbm uplink (model) -0 dbm uplink (model) -40 dbm uplink (model) -50 dbm uplink (model) noncoherent (meas) -70 dbm uplink (meas) -0 dbm uplink (meas) -40 dbm uplink (meas) -50 dbm uplink (meas) Frequency (Hz) Figure 5 X-Band Downlink Phase Noise Estimates Comparison o these phase error estimates to the requirements in Table reveals that there is no impact on BE perormance or received uplink signals as low as -0 dbm. In order to limit the downlink BE degradation to 0. db or this transceiver, the uplink carrier power must be - 4 dbm, which is.5 db higher than the typical 7.85 bps command threshold assuming a modulation index o 0.8 radians; improvements may be made on this threshold and are currently under investigation. Below this threshold, a carrier-only uplink-downlink coniguration may be established or two-way Doppler tracking. Since most spacecrat F links are downlink limited, it is assumed that dropping below the uplink command threshold precludes sending downlink telemetry and thus BE degradation is no longer an issue; at this point, cycle slipping is the pertinent perormance metric. The current design will not contribute to cycle slipping or received uplink signals as low as -50 dbm, though improvements may be made on this threshold and are currently under investigation. Table - Downlink MS Phase Error in Ground eceiver Tracking Loop or Coherent Transceiver Transceiver Coniguration MS Phase Error ( ) Noncoherent downlink dbm uplink dbm uplink dbm uplink dbm uplink dbm uplink Application to Spacecrat Navigation The noise in the downlink requency measurement can be compared to that routinely used to support precise orbit determination in deep space. It is common to base orbit determination on downlink requency measurements made over 60-second intervals. The Doppler residuals, rom all sources are typically less than.0 mm/s. At a requency o 8.4 GHz, 0.0 mm/s corresponds to a round-trip requency error o 0.56 mhz, or degrees o phase on a 60-second interval. The rms phase errors reported above are below this level or signal powers above -4 dbm. Even the highest estimated phase errors o 5 degrees correspond to a Doppler velocity error o only 0.0 mm/s. These errors are small enough to support precise orbit determination. 5. CONCLUSIONS AND FUTUE WOK The prototype APL coherent X-Band transceiver has demonstrated acceptable two-way Doppler tracking perormance with a Doppler velocity error o.9 um/s at 60 s and 0.9 um/s at 000 s; this meets typical mission requirements. Phase noise measurements o the coherent downlink carrier have demonstrated that the rms phase error is suiciently low or negligible bit error rate (BE) degradation and cycle slipping as the downlink signal is tracked and demodulated in the receiving Earth station phase-locked loop. Secondary power consumption is.8 W in the uplink-only mode o operation including the reerence oscillator. The new transceiver architecture provides a high perormance and lexible platorm. This lexibility allows the needs o a variety o missions to be met without incurring hardware changes that drive schedules outward and cost and risk higher. The use o ital signal processing as a oundation or this architecture provides an easy way to incorporate uture perormance enhancements without hardware modiications. This allows overall perormance and power consumption to be tailored to the unique requirements o any mission while minimizing incurred cost and risk. 6. ACKNOWLEDGEMENTS The authors would like to thank the NASA Science Mission Directorate or sponsorship o the New Horizons program which supported development o the basic technologies required or the new coherent transceiver architecture. The technologies and ideas developed or the X-Band Digital eceiver or the New Horizons spacecrat orm the basis o the new architecture. We would also like to thank the DSN or their support and or use o the DTF- and CTT- test acilities during the perormance validation discussed in this paper. 7

8 EFEENCES [] J.. Jensen and. S. Bokulic, Highly Accurate, Noncoherent Technique or Spacecrat Doppler Tracking, IEEE Trans. Aerospace and Electronic Systems 5, pp , 999. [] C. C. DeBoy, J.. Jensen, and Mark S. Asher, Noncoherent Doppler tracking: irst light results, Acta Astronautica 56, pp. 5-55, 005. [] J.. Jensen et al., In-Flight CONTOU adiometric Perormance, 00 IEEE Aerospace Conerence, Big Sky, MT, March 00. [4] C. C. DeBoy et al., The F Telecommunications System or the New Horizons Mission to Pluto, 004 IEEE Aerospace Conerence, Big Sky, MT, March 004. [5] Christopher B. Haskins and Wesley P. Millard, X-Band Digital eceiver or the New Horizons Spacecrat, 004 IEEE Aerospace Conerence, March 004. [6] obert S. Bokulic and J. obert Jensen, Tone-Based Commanding o Deep Space Probes Using Small Aperture Ground Antennas, Proc. o the 5 th International Symposium on educing the Cost o Spacecrat Ground Systems and Operations, Pasadena, CA, July 8-, 00. [7] , Module 0, ev. A, DSMS Telecommunications Link Design Handbook, /5/0. [8]. S. Bokulic, J.. Jensen, and T.. McKnight, The NEA Spacecrat Telecommunications System, Proceedings o the 9th Annual AIAA/USV Small Satellite Conerence, September 8-, 995. [9] A. J. Viterbi, Phase-Locked Loop Dynamics in the Presence o Noise by Fokker-Planck Techniques, Proc. o the IEEE, vol. 5, pp , Dec. 96. [0] J. H. Yuen, Deep Space Telecommunications Systems Engineering, Plenum Press, New York, 98. [] F. M. Gardner, Phaselock Techniques, New York: John Wiley, 979. BIOGAPHY Christopher B. Haskins is the lead engineer or the New Horizons Uplink eceiver. He received a B.S. and M.S. rom Virginia Tech in 997 and 000, both in electrical engineering. He joined the Johns Hopkins University Applied Physics Laboratory (APL) Space Department in 000, where he has designed F/Microwave, analog, and mixed-signal circuitry and subsystems in support o the CONTOU, STEEO, and MESSENGE spacecrat. He also served as the lead engineer or the development o F ground support equipment or the CONTOU spacecrat. Prior to working at APL, Mr. Haskins designed low cost commercial transceivers at Microwave Data Systems. Mr. Haskins is a member o the IEEE. Wesley P. Millard is the lead engineer or the ital subsystems in the New Horizons Uplink eceiver. He received a B.S. in electrical engineering and in computer engineering in 999, and a M.S.E. in electrical engineering in 000, both at Johns Hopkins University. Since joining the APL Space Department in 000, Wes has worked on the STEEO, MESSENGE, and New Horizons programs where he has designed mixed signal circuitry and high eiciency DSP algorithms or FPGA implementation Bob Jensen (M'9) received a B.A. rom Cornell College in Mt. Vernon, Iowa, in 97, and the Ph.D. in physical chemistry rom the University o Wisconsin, Madison, in 978. He joined The Johns Hopkins University Applied Physics Laboratory in 978 and worked on a variety o nonacoustic detection problems, principally involving radar perormance analysis, signal processing algorithms, and rough surace scattering. In 989, he joined the APL Space Department and has participated in the TOPEX altimeter pre-light testing, the development and testing o algorithms or the beacon receiver on the MSX satellite, the NEA telecommunications system, and was responsible or noncoherent Doppler aspects o the CONTOU mission. He is a member o the APL Principal Proessional Sta and the IEEE. 8