Radio Science
  • Open Access

Intercepted signals for ionospheric science


Corresponding author: F. D. Lind, MIT Haystack Observatory, Route 40, Westford, MA 01886, USA. (


[1] The ISIS array (Intercepted Signals for Ionospheric Science) is a distributed, coherent software radio array designed for the study of geospace phenomena by observing the scatter of ambient radio frequency (RF) signals. ISIS data acquisition and analysis is performed using the MIDAS-M platform (Millstone Data Acquisition System - Mobile). Observations of RF signals can be performed between HF and L-band using the Array nodes and appropriate antennas. The deployment of the Array focuses on observations of the plasmasphere boundary layer. We discuss the concept of the coherent software radio array, describe the ISIS hardware, and give examples of data from the system for selected applications. In particular, we include the first observations of E region irregularities using the Array. We also present single-site passive radar observations of both meteor trails and E region irregularities using adaptive filtering techniques.

1 Introduction

[2] Recent advances in software radio, distributed synchronization, and data networks enable the development of coherent software radio arrays. Such systems consist of software defined radios, which are operated with a common distributed time and frequency standard. Networks of coherent software radio systems are a highly flexible class of distributed sensor that can provide wide spatial coverage and high spatial resolution for observation of the Geospace environment. Other examples of passive and active radio arrays for monitoring of the space environment have proven of great utility. For example, Global Positioning System (GPS) monitoring and mapping of total electron content [Mannucci et al., 1998] and distributed HF radar monitoring of ionospheric convection [Greenwald et al., 1995] (i.e., SuperDARN).

[3] Software radio is an approach to building highly flexible radio systems which can receive and process signals in a manner defined primarily by generally programmable digital elements [Mitola, 1995]. Applications such as spectral monitoring, ionospheric radar, and satellite beacon reception can share a common platform where the primary differences are in the software which implements the necessary signal processing. The computing systems implementing such signal processing can also be shared between multiple instruments.

[4] Scientific applications of software radio are highly tolerant to latency when compared with communications applications. This difference can be exploited to good advantage. In particular, radio frequency (RF) bandwidths can be manipulated in several key ways. RF data can be buffered locally while lower data rates are processed in realtime to identify interesting periods. Stored voltage level data can also be reprocessed at a later time using improved algorithms and techniques. This processing can take place long after the original voltage level data is captured (i.e., even years, decades, or potentially longer). It is also possible to process data multiple times using iterative or adaptive optimizations. By transporting coherent voltage level representations of radio signals over large distances using digital communication networks, it is possible to synthesize a very flexible instrument. The resulting system is useful for a range of scientific observations using well established techniques as well as enabling new approaches.

[5] We have implemented the Intercepted Signals for Ionospheric Science (ISIS) array to investigate the utility of a coherent software radio array to study the plasmasphere boundary layer. A key motivation for construction of the ISIS array is to allow the real world evaluation of the effectiveness of a software radio sensor network for observation of the Geospace environment. The Array complements the Millstone Hill UHF Radar System, a 2.5 MW incoherent scatter radar, and can be operated to provide complementary multistatic radar measurements at both VHF and UHF frequencies.

1.1 Coherent Software Radio

[6] The Coherent Software Radio is the fundamental component of a distributed software radio array. The coherent software radio system consists of an analog interface to the electromagnetic environment and a coherent interface layer, which linearly translates signals to the digital domain with precise global time alignment, high stability in frequency, and minimal addition of noise. This is a key boundary which helps to isolate details of the data acquisition hardware from the signal processing and software. Local computation and buffering memory is available to process signals in real time or store them for later use. A network interface connects the receiver to the wider network and provides for remote control and reconfiguration. These radio systems are combined into a networked array that can operate collectively to make observations.

[7] The general pattern of a sensor array enabled through distributed coherence [Alexander, 1977] reoccurs in many modern radio and radar instruments. It is by no means unique to Geospace radio science applications and has been widely exploited by other communities (e.g., Radio Astronomy and DoD). Applications of such approaches to astronomical observations are being extensively investigated, and the current generation of distributed radio telescope arrays (e.g., MWA, LOFAR, LWA) exploit this essential architecture in a range of configurations [Bowman et al., 2007; Butcher, 2004]. Active adjuncts to such arrays are also under consideration (e.g., LOIS) [Thidé, 2007].

1.2 Array Coherence

[8] It is possible to combine the software radio approach with the precision timing possible using signals from the Global Positioning System (GPS) [Lewandowski et al., 1999]. The resulting coherent software radio system can align its data and mode changes very precisely in time and maintain a high stability in frequency even between systems separated by large spatial scales (∼1000s km). This is a similar approach to the use of atomic clocks for very long baseline interferometry (VLBI) derived coherence in radio astronomy observations and has been an important basis for astronomical radio synthesis imaging techniques for many years [Counselman et al., 1977; Clark et al., 1979]. Synchronization of distributed software radio networks must be of sufficient precision to make meaningful comparison of signals between sites in the array (e.g., [Rogers and Moran, 1981]).

1.2.1 GPS Performance and Limitations

[9] The Global Positioning System (GPS) provides a precise atomic clock broadcast from multiple satellites with known orbits. In the current generation of commercial receivers, the synchronization is limited to pulse per second (PPS) alignments in the range of 20 to 50 ns for high quality units. Lower quality receivers typically achieve on the order of 100 ns. Filtering of the GPS derived PPS signals can improve this to some extent, and stabilization of oscillators on long time scales with Allan variances in the one part in 10E11 or a few parts in 10E12 range can typically be achieved. This translates into networked coherence appropriate for a wide range of ionospheric measurements under most conditions.

[10] In some cases the performance of GPS as a synchronization source can be degraded. For example, Space Weather events can disrupt the ability of GPS receivers to maintain lock on the satellite signals [Skone and De Jong, 2000; Lambour et al., 2003]. It is also possible for the GPS synchronization method to introduce artifacts which impact the measurements of an array. An example of this is that the filtering involved in oscillator stabilization can introduce unintentional variations in the stabilized oscillators. In some cases, this can be observed in cross-correlated signals as a wandering of frequency alignment between sites over seconds to tens of seconds. The occurrence of such variations are specific to the details of the hardware involved. The effects can be characterized in the laboratory using a modified or dynamic Allan variance [Nunzi et al., 2007; Allan, 1987]. They can introduce artifacts into the measurements made by a coherent radio sensor array.

1.2.2 Self Synchronization Techniques

[11] In some applications and conditions, it is desirable for an instrument array to achieve coherence using direct reception of signals in the RF environment or to utilize characteristics of observed RF signals to achieve synchronization. Holdover for the case where GPS lock is lost is one example, but some systems might wish to dispense with the cost of GPS synchronization entirely. This is particularly true for very low power networked sensor elements, and a number of possible strategies have been investigated by other authors [Elson and Rōmer, 2003].

[12] One example of this is the derivation of high precision time synchronization from common pulsar observations [Taylor, 1991; Arzoumanian et al., 1994], but this takes a substantial and flexible aperture. Another approach is to use the narrow and near zero Doppler bandwidth of ground clutter or multi-path observations to correct frequency differences between networked radar systems. Man-made signals will also often contain explicit or hidden synchronization information that can be received and used for self synchronization.

[13] An alternative approach for a computationally powerful software radio system might be to directly receive GPS signals and then use software based signal processing to create the required system coherence by solving for the synchronization as part of the overall data analysis.

1.3 Wide Area Voltage Level Transport

[14] A key aspect of a coherent software radio array is the translation of the RF signal coherently into an associated representation as a complex RF digital voltage. Essentially RF bandwidth is translated to data network bandwidth which enables these signals to be transported, analyzed, and compared at a fundamental level. This translation is most general when it is fully coherent in time and frequency and where a detailed characterization of the analog and digital channel between the electromagnetic environment and the conversion point in the system is available. It is also useful to associate metadata information with the voltage level data. For example, such metadata can include information about receiver location, analog path transfer functions, coherence metrics, and specifics of the data acquisition hardware configurations.

[15] Manipulation and processing of data at the voltage level is fundamental to the multi-role applications of software radio arrays and is also a key to the application of general purpose computing technology for array signal processing. Many of the challenges of this approach (e.g., limited network bandwidth and the required computing power) are being alleviated through technological advances.

1.3.1 Network Transport Considerations

[16] We first stumbled upon the transport of coherently aligned voltage samples over wide area networks in the early development of the Manastash Ridge Radar system [Sahr and Lind, 1997]. At that time, the fundamental nature of this approach was not clear to us. The available data network was fairly limited in bandwidth (500 kb/s), and only two sites were connected and processed for the resulting passive radar applications. This level of available bandwidth is useful, but can be fairly limiting for wide RF bandwidths sampled at high dynamic ranges. It is currently possible to do better than this in many locations. However, in practice, it can be quite difficult or expensive to achieve high performance networking in remote locations (e.g., dedicated fiber or expensive satellite uplinks). Wireless approaches do offer some deployment advantages for a coherent software radio array, and the potential for self networking arrays is very interesting.

[17] As networks have become ubiquitous and bandwidths have increased, it has become clear that voltage level RF transport is a key element in enabling highly flexible and distributed software radio sensor networks. In retrospect, a clue to the general importance of wide area voltage transport could have been recognized in VLBI techniques, which evolved from transport of data using tape, then disk, and most recently networked transport of large coherent RF bandwidths with low dynamic range (eVLBI) [Szomoru et al., 2004].

[18] In a networked coherent software radio array the useable bandwidth for transporting RF signals is in general limited and often asymmetric between sites. This limitation forces selective and intelligent use of the available network bandwidth. Typical bandwidths in the ISIS Array currently range from about 1 to 10 Mb/s with relatively low latencies and good connectivity. This was achieved by preferentially selecting well connected sites.

[19] In coping with bandwidth limits and asymmetries an obvious approach is to limit array performance in some manner. A simple means to do this is to limit the array duty cycle or operate in a burst mode of operation. Another method is to transport limited bandwidths in order to identify time intervals which would be valuable for intensive processing. Multiple signal distributed compressive sensing [Baron et al.2009] also provides some indication of a useful means of data bandwidth reduction for a distributed software radio array. Applying such techniques in an array wide sense is a future direction for investigation.

1.4 Software Radio Signal Processing

[20] Software radio offers great flexibility in the signal processing and analysis of radio signals from an instrument array. We have discussed patterns for Software Radar signal processing in previous work [Grydeland et al., 2005], and many of the same approaches and patterns work equally well for networks of software radio systems.

[21] Signal processing implemented in software using general purpose computers helps to avoid many of the costs associated with the re-implementation of systems due to changes in the underlying hardware technology. It also allows us to take advantage of improvements in computing and networking capacity to increase the capabilities and performance of radio science systems and to share computing capacity between systems.

[22] This flexibility comes at a price as the initial implementation of the needed software can be complex and limited in performance. The later problem can be mitigated to some extent using specialized processors to accelerate selected operations (e.g., digital filtering) where the implementation cost and difficulty can be kept low. However, for our applications, we have explicitly decided to limit the RF bandwidths that we capture and process to the relatively narrow limits (i.e., kHz to MHz), which can be handled with off the shelf computing systems and networks.

2 ISIS Array Overview

[23] The ISIS Array is a prototype coherent software radio array. It is designed to operate as a flexible multi-role distributed radio science instrument. Operational modes involving active and passive multistatic radar imaging, satellite beacon observation of total electron content (TEC) and scintillation, and radio intercept and time difference of arrival (TDOA) applications are supported over a wide range of operating frequencies (0.5 to 1500 MHz). Appropriate antennas must be used, and some analog reconfiguration can be required for specific experiments or applications.

[24] The array is constructed using a series of MIDAS-Mobile data systems which can coherently capture wide bandwidths of RF signals from a variety of antennas. The array is capable of applying high performance Grid based computing to the real time and batch processing requirements of a variety of experiments. The underlying software radar technology allows a consistent platform for data collection, processing, and control.

[25] Seven primary MIDAS-M nodes were constructed and installed between 2004 and 2007. An additional equivalent capability is also in use at the Millstone Hill Incoherent Scatter Radar as part of the IS radar data acquisition system. Two of these systems were used to replace legacy receivers that were used in the Manastash Ridge Radar System. These systems form the core of the ISIS array as it currently exists, although one of the sites (NRAO Green Bank) has recently been withdrawn from the field.

2.1 Scientific Motivation

[26] The plasmasphere boundary layer (PBL) [Carpenter and Lemaire, 2004] is a critical region for inner and outer (auroral) magnetospheric plasmas. The important processes of this region are the associated electric fields, the evolution and redistribution of thermal plasmas, the formation of density irregularities via instability mechanisms, and their related Space Weather effects. The key scientific goal of the initial ISIS Array deployment is to experimentally investigate this boundary layer to characterize the scale sizes of electron density perturbations and associated electric fields from micro to mesoscales (hundreds of meters to the order of 1000 km).

[27] The role of instabilities in the evolution of the plasmasphere, its erosion, and recovery is not yet well investigated or understood [Carpenter and Lemaire, 1997]. Storm enhanced density plumes and associated electric fields and density gradients can provide the conditions necessary in the E and F regions of the ionosphere for the onset of different types of instabilities and the generation of a wide range of perturbations in the underlying plasma. These plasmasphere perturbations can exist as structures on a wide range of scale sizes from less than 40 km [Jacobson and Erickson, 1993] to 1000 km or more in size [Moldwin et al., 1995].

[28] This structuring can become significantly enhanced during active geomagnetic conditions and storms. In the E region, the footprint of plasmaspheric electric fields and sharp density gradients drive two-stream and gradient drift instabilities. These are coupled into the closure of field-aligned currents in this layer. The interaction of these electric fields with a low ionospheric conductivity can lead to the formation of regions of intense electric fields and enhanced field aligned currents [Streltsov and Foster, 2004].

[29] The ISIS Array will be used to experimentally investigate the electric field structures in the plasmasphere boundary layer. Measurements will be made to characterize the scale sizes of the electric field structures that occur. It will also enable investigation of the linearity relationships and background parameters of Eregion coherent scatter parameters when probed simultaneously at UHF and VHF frequencies. This comparison has not been conducted with modern instrumentation at mid-latitudes in the PBL region.

2.2 Array Deployment

[30] The ISIS Array deployment is listed in Table 1, which notes selected site characteristics. The ISIS Array nodes were deployed at sites where we felt that a sufficient level of infrastructure would exist to support the systems, where their network connectivity would be good, and where suitable scientific and technical collaborations could be encouraged. This is probably typical of the approach for smaller arrays but less systematic than would be warranted for a larger scale deployment.

Table 1. Overview of ISIS Array Sites and Basic Details
SiteNodeLatitude (°)Longitude (°)Altitude (m)Notes
Millstone Radar00042.6232288.5112146.002.5 MW active UHF radar
Green Bank00138.4378280.1638792.6446 m steerable antenna
University of Washington00247.6548237.693334.70discone ring array
Manastash Ridge00346.9511239.27531195.5116 λ interferometer
Siena College00442.7195286.247983.17log periodic ring array
Dartmouth College00543.7056287.7138164.51FM and dual pole UHF antennas
Zeman Lab00642.4954283.5687334.00 
Haystack Observatory00742.6232288.5112129.27 

[31] In the case of the first ISIS node (001), we deployed the system to exploit the availability of the NRAO Green Bank 46 m antenna. This system was withdrawn in 2012 after completing a series of bistatic active radar experiments. This deployment limited the spatial coverage of the Northeast ISIS Array by diverting a node from deployment in that region. However, it did allow for a range of special experiments with the large aperture located in a National radio quiet zone. The Northwest ISIS Array deployment was used to replace and modernize the systems at the University of Washington and at Manastash Ridge Observatory. This provided educational opportunities for students involved as well as expanding the capabilities of an already successful field site [Lind et al., 1999; Meyer and Sahr, 2004].

2.3 Array Coverage

[32] Array coverage is determined by the antennas and configurations which are used at a given field site, the frequency of signals which are involved, and the propagation conditions which may include refraction and scattering effects. Thus, in most cases, array coverage is a strong function of a particular experiment configuration. Additionally, the physical constraints of different sites (i.e., different roof top layouts) resulted in variation of the antenna deployments between sites.

2.3.1 Direct Reception

[33] Vertical polarization omnidirectional coverage is provided using discone antennas at each site. This allows for local reception of signals where little directivity is required. A combination of log periodic and yagi antennas are used for coverage of likely regions of coherent scatter from E region irregularities. These antennas are typically organized to provide sector coverage to the north, northwest, and northeast of each site. Satellite beacon coverage is provided by turnstile antennas that cover 150, 240, 400, and 1066 MHz frequencies to allow support of CERTO type beacons [Bernhardt and Siefring, 2006] such as those carried on satellites such as DMSP, Transit, RADCAL, and COSMIC. These beacon antennas are not yet uniformly deployed at all sites. In some cases, low frequency active antennas appropriate for the AM radio band (0.5 to 2 MHz) have also been used in support of specific propagation experiments.

2.3.2 Multistatic Magnetic Aspect Angle Coverage

[34] For the observation of E region irregularities using meter scale radio waves, the scattering geometry is strongly constrained to be nearly perpendicular to the Earth's magnetic field [Fejer and Kelley, 1980] with an attenuation on the order of 10 dB per degree from perpendicular at UHF [Foster et al., 1992] with somewhat higher attenuation close to perpendicular and lower attenuation at greater aspect angles. Evaluation of the bistatic magnetic aspect angle is done using the IGRF 2010 magnetic field model [Showstack, 2010].

[35] Figure 1 shows a multistatic evaluation of the magnetic aspect angle geometry by combining all possible E region irregularity scattering paths between array sites and realistic transmitter selections for each site. The coverage is shown as a function of absolute degrees from the perpendicular aspect angle. It is constructed by evaluating the scattering paths between all array sites, limited by the E region horizon at 110 km altitude, and for magnetic aspect angles which meet the perpendicularity condition within 2° (i.e. corresponding to approximately 20 to 30 dB attenuation; similar to the maximum expected signal to clutter ratios for FM radio based passive radar). In each case, the best possible scattering geometry is selected. We approximate here by ignoring the variation in scattering wave vector, flow angle effects, and any atmospheric and ionospheric refraction or attenuation. In practice, the refraction effects are not particularly significant at the VHF operating frequencies most common to ISIS Array experiments. We also simplify by assuming the use of uniform antenna patterns and transmitter signals received local to each ISIS site. In practice, the observable region for a specific antenna will be convolved with the antenna pattern used for reception in a specific direction of observation. Wide spatial coverage is achieved even with a fairly small number of sites. In practice, transmitters used for observation of E region irregularities can be widely distributed (e.g., FM radio stations), and this spatial diversity provides both additional coverage and the opportunity to apply interferometric techniques [Meyer and Sahr, 2004] with a single receive site.

Figure 1.

Multistatic magnetic aspect angle coverage for the entire ISIS Array when a realistic combination of transmit signals is evaluated over potential E region scattering paths between the sites. For the computation, sites receiving scattered signals are shown as triangles, sites intercepting transmitter signals are noted with a cross, and actual transmitter locations are indicated using inverted triangles. Aspect angle is shown as the absolute value of degrees from perpendicular to the Earth's magnetic field.

3 The MIDAS-Mobile Platform

[36] The ISIS Array is constructed using the Millstone Data Acquisition System - Mobile (MIDAS-M) platform. This is the current generation of software radar platform implemented at MIT Haystack Observatory and is the second generation of software radar based platform we have implemented [Grydeland et al., 2005].

[37] MIDAS-M is designed to directly manipulate voltage level representations of RF signal bandwidth using general purpose computers with instrument functionality implemented in software. The high level architecture is shown in Figure 2. The MIDAS-M platform consists of an analog system, a digital system, and in the case of an active radar, a timing system (which is not discussed in detail here). Voltages are aligned coherently between MIDAS-M units which allows for multistatic radio and radar applications. One implementation of the MIDAS-M system is shown as a photo in Figure 3.

Figure 2.

A MIDAS-M system consists of an analog system and a digital system which are typically constructed in transport racks (i.e., 'cubes').

Figure 3.

Photo of a deployed MIDAS-M system at Siena College in Loudonville, NY. This photo is of node four (004) out of seven (007). The analog cube is only partially visible.

3.1 MIDAS-M UHF Tuner

[38] The MIDAS-M UHF Radar Receiver is a high integration receiver for down-converting UHF signals from RF to an IF appropriate for digitization by the digital receivers. A photo of the tuner is shown in Figure 4. The tuner is primarily designed for the down conversion of signals in the 150 to 1500 MHz range with an external image reject filter. A bypass input allows direct digitization of signals below 150 MHz by the digital receivers. The UHF tuner is appropriate for radar applications through the provision of trigger controlled blanking capabilities. External time reference inputs for GPS derived PPS and a 10 MHz reference allow the unit to provide highly accurate mode changes (100 ns level alignment between sites). The tuner incorporates ethernet and serial via USB control with a ucLinux operating system running on an Altera FPGA and Nios2 processor. Web based status and control of tuning, modes, and configuration are provided. Tuner specifications are presented in Table 2.

Table 2. Specifications for the MIDAS-M UHF Tuner
Inputsdual per tuner with bypass
Frequency range10 kHz to 1.5 GHz
RF input level−120 to −30 dBm (tuner)
Gain range−4 to 25 dB
Noise figure<8 dB
IF frequency126 MHz
IF bandwidth32 MHz
Onboard PLL LO range240 to 760 MHz
Spur free dynamic range70 to 90 dB
Blanking isolation>100 dB
Noise diode level>−143 dBm / Hz
ComputingAltera NIOS II on FPGA
Memory16 MB SDRAM
Storage8 MB FLASH
Operating systemucLinux
CoherenceExternal 10 MHz reference
SynchronizationGPS PPS, External trigger
Figure 4.

Photo of the MIDAS-M UHF tuner used by the ISIS Array.

3.2 Digital Receivers

[39] The MIDAS-M platform uses two Mercury Computing Echotek GC314-FS digital receivers per system. Three analog inputs are digitized using 14 bit A/D converters at 72 MHz, and the RF band can be digitally channelized using Texas Instruments GC4016 digital down converters into up to four narrowband signals per input. Output signals are typically 16 bits with additional dynamic range produced from the processing gain associated with oversampling and filtering. Many flexible tradeoffs of bandwidth, channel count, oversampling and filter configuration are possible with the system. The MIDAS-M platform can channelize up to 24 narrowband signals simultaneously from the six analog inputs. The frequency centers of the signals can be selected anywhere within the overall digitized bandwidth (i.e., within a 36 MHz window). Bandwidths of order 100 kHz with 24 channels or up to 2 MHz with fewer channels have been sustained. Wider bandwidths of up to 8 MHz bandwidths have been demonstrated using burst sampling of limited time intervals. These RF bandwidths are far lower than was expected from the card specifications (i.e., even given I/O rate limitations) and it is unlikely that we will resolve these performance limitations. Newer generation digital receivers are rapidly becoming easy to implement and this will result in next generation systems that offer much larger reliably sampled bandwidths.

[40] The complexity of modern digital receivers is significant, and the software implementation process for the ISIS Array proved to be a great impediment. The complexity of dealing with so many RF signal channels from multiple receiver sites (i.e., up to 192 for the current ISIS Array deployment) requires significant software and good planning. The effort involved in software development should not be under-estimated by researchers implementing systems of a similar or larger scope.

4 Array Techniques and Examples

[41] We have implemented a series of signal processing and analysis techniques for use with the ISIS Array as an initial demonstration of the experimental flexibility made possible by a coherent software radio array. These capabilities are available for scientific observations, and we provide discussion of the implementations and examples of the data in the following sections. The list of techniques discussed here are by no means exhaustive, and the array can be configured and software processing developed to enable a wider range of measurements relevant to the study of the atmosphere, ionosphere, and near space environment.

4.1 Spectral Monitoring

[42] Spectral monitoring is a basic approach to radio signal analysis, which is widely applied for many applications. Monitoring of wide spectral bands can be done using off-the-shelf equipment, although precisely calibrated monitoring can be challenging [Rogers and Bowman, 2008]. Spectral monitoring of selected radio bandwidths can be useful for many Geospace applications both as emissions from natural radio phenomena [LaBelle et al., 1994] and in the propagation of man-made signals [Helliwell et al.1973].

[43] The ISIS Array is suitable for relatively narrowband applications of spectral monitoring over moderate durations. For example, the result of monitoring from an hour long observation of the AM radio band is shown in Figure 5. These data were taken with 4 MHz sampling rate, a digital filtering bandwidth of 3.2 MHz, and a limited duty cycle of (∼6%). The resulting voltages can be analyzed with a range of spectral resolutions (in this case 122 Hz). Calibration of the data is relative to the A/D converter voltage reference and is not absolute.

Figure 5.

Spectral monitoring of the AM radio band at Dartmouth College (005) over an hour period.

[44] Spectral monitoring using the ISIS Array can also be applied to observation of satellite beacon signals. Such signals are a useful probe of ionospheric variation and structure, and with appropriate measurement of dual and multiple frequency beacons, ionospheric electron density can be accurately measured along the propagation path [Bernhardt and Siefring, 2006]. An example of the raw spectral information associated with an overflight of the RADCAL satellite, which carries 150 and 400 MHz beacons, is shown in Figure 6.

Figure 6.

Spectral monitoring of an overflight of the RADCAL satellite at MIT Haystack Observatory using a turnstile antenna and showing both the (top) 150 MHz and (bottom) 400 MHz signals generated by the transmitter beacon on the satellite.

4.2 Passive Radar

[45] Passive radar is a technique for observing targets of interest using radio signals already present in the environment. By intercepting both the signal radiated by a transmitter and the scatter from targets of interest, it is possible to make traditional radar measurements. The range, doppler shift, and bearing of a target can all be determined with an accuracy that depends on the nature of the illuminating waveform and the characteristics of the passive radar system [Griffiths and Baker, 2005; Howland, 1999]. In most cases, the signals used by a passive radar are generated by non-cooperative illuminators such as FM radio or television stations. Such illuminators are present in most of the world, radiate significant power, and FM radio stations in particular transmit signals well suited to radar applications and the observation of geophysical targets Sahr and Lind [1997]. The technique can also be applied to signals from a cooperative illuminator such as an active radar system. In this case, it is similar to more traditional forms of bistatic or multistatic radar.

[46] In brief, a passive radar must record the signal of interest with one receiver while excluding it from a second receiver that then detects the weak signal from the target. The dynamic range between the transmitter signal level and that of the scatterers can be greater than 130 dB [Lanterman, 1999]. This dynamic range can be achieved using high performance receivers combined with appropriate antenna patterns. In some cases it is appropriate to use adaptive filtering to cancel unwanted signal energy. Another possibility is to use a multistatic configuration and appropriate local topography to achieve the required dynamic range.

[47] For the ISIS Array, we have implemented a basic passive radar processing capability which allows Grid batch processing with machine level parallelization for multiple signal pairs. Some sites have been operated in quasi-realtime modes, but the current emphasis is on measurement campaigns to burst capture geomagnetic events. The structure of the signal processing is shown in Figure 7. The core of which is simply the ambiguity function calculation

display math(1)

which we have previously discussed [Sahr and Lind, 1997] in detail. Where inline image is the target correlation function as a function of range, r is the delay variable, τ the lag variable, and the summation is over time t. The reference transmitter signal is given by xt and the remote detected signal is given by yt, anddenotes the complex conjugate.

Figure 7.

Diagram of the typical passive radar signal processing used for ionospheric observations.

4.2.1 Reference Signal Filtering

[48] To improve the performance of the FM radio passive radar measurements, we have implemented a reference signal bandwidth threshold filter. This filter functions by zeroing the amplitude of the transmitter reference signal when the instantaneous bandwidth of the reference falls below a threshold value. This has the effect of eliminating energy which is of narrow bandwidth and that often results in poor ambiguity performance.

[49] For a complex valued input reference signal of the form x(n)=a(n)+jb(n), the filter is implemented using a two point instantaneous bandwidth approximation of the form

display math(2)

where this approximation is discussed in detail in Barnes [1992] and in a more general manner in Ristic [1996]. This estimation equation is then applied to an input reference data vector to obtain the instantaneous bandwidth of the reference signal, which is then used as a threshold of the form

display math(3)

where εis a very small value compared to the mean signal value and γ is a filtering threshold, which is determined empirically. The application of this filter results in a reference signal where bandwidths lower than the threshold now have sample values which contribute no net energy to the ambiguity calculation. A comparison of ambiguity function performance for a real world FM radio signal is shown in Figure 8, before and after application of the filter.

Figure 8.

The ambiguity function of an FM radio station for an interval of poor signal quality is shown (top) before and (bottom) after reference filtering. The plot shows the signal to clutter ratio (SCR) normalized to the peak of the ambiguity response.

[50] This bandwidth based filter essentially selects short time intervals which have a good ambiguity function in range. This is not the only potential choice for reference signal filtering. For example, an approach which we have not yet tried is to compute a reference self ambiguity for short time intervals and then create a threshold based on integrated sidelobe levels.

[51] The cost of reference signal filtering of this type is a reduction in the number of samples which contribute to the estimation of target parameters. The exact number of samples thrown out is determined by the specific nature of the radio signal used as a reference waveform and the selected filtering threshold.

[52] In practice, for FM radio stations with poor ambiguity performance, of order 50% of the samples can be eliminated by a reasonable choice of γ. However, the resulting improvement in signal usability is significant. The filtering can transform reference stations, which routinely broadcast otherwise unsuitable signals into useful reference signals for practical observations.

4.2.2 Illuminator Availability

[53] Potential passive radar illuminators are widely available in most parts of the world. The illumination of potential scatterers is constant, frequency diverse, and of high availability. Spatial diversity of the transmitters can be used for interferometry in some applications with antennas at a single receiver site. Interference is also omnipresent on many frequencies and occasionally very strong for nearby transmitters. The recent advent of in-band on-channel (IBOC) digital radio broadcasts has created a challenge for FM passive radar observations. By occupying local channels that do not contain a transmitter, these signals act to raise the noise floor for receiving scatter from more distant transmitters. This essentially jams the channel, and it is necessary to use some form of adaptive selectivity to eliminate the interference. Fortunately, not all channels are currently occupied, but the increasing ubiquity of such transmissions points toward the need for array receivers with greater aperture and space time adaptive processing that filters the interfering signals. This is similar to the requirements which have been discussed for aircraft tracking applications [Howland et al., 2005; Malanowski and Kulpa, 2008].

4.2.3 Ground Clutter

[54] Ground clutter is a very common signature in passive radar data taken with the ISIS Array. The exact characteristics of the ground clutter vary with frequency, site, antenna orientation, the surrounding terrain, and propagation effects. Bistatic observations between two ISIS sites show appropriate time delays, while monostatic observations often eliminate the direct path and ground clutter signatures to improve system dynamic range. Figure 9 shows an example of ground clutter for the Manastash Ridge ISIS site for both FM radio and HDTV signals to provide a feel for the typical observations. In these data the strongest scatterer is nearby Mount Rainer, which is a large volcano visible to both the transmitter and remote site. The Northeast ISIS sites show significantly weaker ground clutter signatures due to the local topography having much lower elevations. The ground clutter is variable with time, and it seems likely that a component of the clutter is related to atmospheric effects on propagation. Extraction of geophysical parameters from such observations has yet to be analyzed or attempted.

Figure 9.

ISIS Array observations of ground clutter using (top) passive radar for an FM signal and (bottom) a 1 MHz sub-band of an HDTV signal. The strong ground clutter signature is due to the Cascade mountains and Mount Rainer, which are visible to both the reference transmitters intercepted at the UW and the remote Manastash Ridge site. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

4.2.4 Aircraft

[55] Aircraft are the most common class of non-geophysical targets observable with the ISIS Array using passive radar techniques. Observation of aircraft is a useful gauge of system sensitivity and is used to monitor correct array operation in the absence of geophysical targets. The utility of passive radar systems for air defense and surveillance applications is now widely discussed. Commercial systems to perform such observations and the required analysis have been available for government customers (e.g., Lockheed's Silent Sentry system). The ISIS Array site configurations are not optimized for aircraft detection or tracking applications. However, an example of such data using FM radio signals for a single site configuration is shown in Figure 10. The single site example uses a particularly favorable transmitter located to the south of the Haystack ISIS Array site. In this and other single site examples, the ground clutter signature has been removed using an RLS (recursive least squares) adaptive filter similar to that discussed in detail by other authors [Griffiths and Baker, 2005; Cardinali et al.2007].

Figure 10.

ISIS Array observation of aircraft using FM radio passive radar for a monostatic geometry using front to back FM log periodic antennas. Adaptive RLS filtering has been applied to remove the direct path signal and local ground clutter. The strongest target shows significant sidelobe structure due to the ambiguity of the FM signal. The transmitter was located to the south, and the aircraft were to the north of the reception site. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

4.2.5 Meteor Trails

[56] Relatively modest radars allow observation of meteor trails and the determination of meteor radiants, altitudes, and count rates. Modern meteor radar systems provide measurements of neutral winds at altitudes of 90 to 120 km by estimation of the Doppler content of meteor trails [Hasebe et al., 1997]. Such observations are more optimally made at HF frequencies where the scattering cross-sections are large enough to allow for high count rates and rapid estimation of geophysical parameters. Because the VHF scattering cross section of meteor trails is still quite large, the trails may be observed using transmitters of opportunity as sources.

[57] Systems which detect forward scattering path enhancements have been used for automated meteor counting for a number of years [Wislez, 1995], and detailed analysis of FM forward scatter time behavior has been reported [Yrjola and Jenniskens, 1998]. However, such systems are not coherent, do not provide range and doppler estimates, and have a limited ability to distinguish meteors from other types of targets.

[58] Figure 11 shows ISIS Array observations of a typical meteor trail as observed using FM radio based passive radar using a single site passive radar configuration. More extensive observations have been attempted during meteor showers with somewhat mixed results. It is apparent that the trails observed on backscatter paths using the ISIS Array are dominated by large events which are relatively rare. Observations of FM radio forward scatter in very radio quiet locations do show that this technique should be very promising when applied using the next generation of low frequency digital radio telescope arrays [Bowman et al., 2007].

Figure 11.

An example of observation of a meteor trail using FM radio passive radar for a monostatic configuration at MIT Haystack Observatory with 1 s time resolution. Adaptive RLS filtering has been applied to remove the direct path signal and local ground clutter. In this example, a 50 kW ERP (effective radiated power) reference transmitter on 105.1 MHz located near Providence, RI (WWLI) was used. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

4.2.6 E Region Irregularities

[59] Coherent backscatter from E region irregularities has been extensively investigated using medium and large-size radars [Haldoupis, 1989; Sahr and Fejer, 1996]. It is well understood that magnetic field-aligned ionospheric irregularities form in the E region (near 110 km altitude) when the ambient electric field exceeds ∼20 mV/m for mid to high geomagnetic latitudes. The irregularities are generated at metric wavelengths due to two stream and gradient drift instabilities, and simulations have recently provided great progress in understanding irregularity generation, evolution, and dissipation processes [Oppenheim et al., 2008]. Recent demonstration of linearity relationships between average backscattered power and the driving electric field at UHF [Foster and Erickson, 2000] and HF wavelengths [Hysell et al., 2009] raise the potential for monitoring of fine scale electric field structure during geomagnetically active conditions.

[60] The ISIS Array has been applied to make observations of E region irregularities with a focus on addressing two specific experimental issues of existing data sets. First, the combination of monostatic radar systems with the aspect angle sensitivity of the observations has led to limited spatial coverage at VHF and UHF wavelengths. This is a significant limit for understanding the mesoscale context of these irregularities for which the generation conditions are in some cases clearly widely distributed. To make matters more challenging, the irregularities and their driving electric fields are often highly structured on fine spatial scales, which can only be resolved with radars having resolution on the order of at most a few kilometers. Second, the observations to date have been largely monochromatic with a single radar operating frequency being available. There were some notable early exceptions where simultaneous experiments were done on several frequencies [Flood, 1960; Moorcroft, 1966; Fialer, 1974], but these observations were so early in the development of the field that the radar performance levels were quite limited by modern standards. This limited access to the k-spectrum of the irregularity energy distribution makes the larger context of the turbulent evolution of the scattering plasma hard to quantify experimentally. Recent work has begun to address this limitation with Noble et al. [1987] providing such comparison at several HF and low VHF wavelengths in a modified ionosphere, Milan et al. [2003] examining multiple HF frequencies, and Hysell et al. [2007] performing a comparison of HF and UHF frequencies at Jicamarca.

[61] Figure 12 shows an early detection of E region irregularities during engineering testing of the ISIS nodes deployed to replace the previous Manastash Ridge Radar system hardware. A first marginal detection was made somewhat earlier on 10 November 2007 but was extraordinarily weak. After this initial pair of events, an extended solar minimum occurred [Russell et al., 2010], and this resulted in a unexpectedly long period until the conditions were again right for observing mid-latitude irregularities with the Array.

Figure 12.

Irregularities from 14 December 2006 were observed using the nodes at UW and Manastash Ridge in parallel with the use of the prior system hardware. The irregularities were strong enough for a clear detection over a brief interval. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

4.2.7 The Geomagnetic Event of 3 August 2010

[62] On 3 August 2010, we used the ISIS Array to observe E region irregularities associated with a moderately strong geomagnetic storm. The event was captured using a multistatic configuration of the ISIS Array consisting of all sites, but those in Ithaca, NY and the one at NRAO Green Bank, WV. The systems at UW and Manastash Ridge Observatory did not observe E region irregularities during the event.

[63] At MIT Haystack Observatory, a set of four reference transmitters was intercepted on (89.7, 94.1, 96.1, and 107.3 MHz) along with a remote signal from the Siena site (103.5 MHz). At Siena College, data were collected using three log periodic antennas covering sectors to the Northeast, North, and Northwest using a single remote signal (89.7 MHz) looking for scattered signals and a single reference transmitter (103.5 MHz). At Dartmouth College, a single FM log periodic antenna pointing slightly West of North (∼16°) was used to examine scattered signals for reference signals from both Siena (103.5 MHz) and Haystack (89.7, 96.1, 107.3 MHz). Even for this relatively small array, the number of signals and scattering paths available were reasonably large. Stations were selected based on their estimated sensitivity determined by the signal to clutter ratios typically seen for aircraft. The composite aspect sensitivity of the experimental configuration in the Northeast United States is a subset of the overall array coverage with scattering paths that were in general near backscatter at the ranges where irregularities were detected.

[64] Detection of the irregularities was only made on two scattering paths during the course of the event. The first involved the reference signal on 103.5 MHz which was intercepted using the Siena site and a scattered signal observation at the Dartmouth site with an example of the range doppler data in Figure 13. Irregularities were also observed at a somewhat later time interval by the Siena site using 89.7 MHz as the reference signal from the Haystack site with an example of these data shown in Figure 14. The irregularities were only detected in the Northwest pointing antenna with only very faint evidence of a signal in the North pointing antenna, which is clearly due to energy entering in the sidelobes.

Figure 13.

A brief interval of E region irregularities was detected by the ISIS Array on 3 August 2010. The FM passive radar technique allows for the observation of the range, doppler, and time evolution of the irregularities. A cross-ambiguity range doppler plot of the Siena to Dartmouth scattering path is shown with signal to clutter ratio normalized to the median clutter floor. E region irregularities are clearly detected by the array on this scattering path. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

Figure 14.

The E region irregularities as observed on the Haystack to Siena scattering path from the 3 August 2010 event. The scatter was strongly detected on the Northwest pointing antenna (shown above), as very weak crosstalk in the North pointing antenna and with no detection in the Northeast pointing antenna. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

[65] Other scattering paths with sufficient sensitivity to detect the irregularities did not observe them, but these correspond to paths that were farther East and South in the array. This is an indication (along with the ranges involved) that the irregularities were first detected far to the Northeast and then later were located to the Northwest of the array. This is also consistent with changes in the observed doppler shifts. Some potential ambiguity exists because the irregularities could be detected outside the main lobe of the antenna if they fell in a sidelobe and were of particularly high cross section. Future scientific experiments will employ additional antennas, use the Millstone Hill UHF radar system to localize the irregularities, and employ interferometry to fully break any ambiguity in the scatter signal's angle of arrival.

[66] The data we have presented here show that the overall sensitivity of the system is sufficient for the monitoring and study of E region irregularities in the plasmasphere boundary layer region of the mid-latitude ionosphere. The detection of the irregularities on multiple scattering paths (and their absence on others) also points to the potential of larger arrays which could provide much wider mesoscale coverage in space and additional diversity in the number of available transmitters. A detailed scientific analysis of the August 3, 2010 event and subsequent observations will be documented elsewhere.

4.2.8 Single Site Observation of E Region Irregularities

[67] During the geomagnetic storm of 9 March 2012, the ISIS Array was used in a configuration which combined multistatic observations with single site observation of several favorable transmitters (97.5 and 105.1 MHz FM) at MIT Haystack Observatory. During this moderately strong Geomagnetic event, the system at MIT Haystack Observatory was used to observe E region irregularities using a single site passive radar configuration that employs RLS adaptive filtering to remove the direct signal and local ground clutter. An example of simultaneous observation of E region irregularities on two frequencies is shown in Figure 15. These observations point to the potential of single-site passive radar for Geospace applications. To our knowledge this is the first time that such a single-site direct reception and detection capability has been demonstrated for the observation of E region irregularities. Although a prior experiment (unpublished) to use FM radio signals scattered by Mt. Rainier as a reference signal for single site passive radar was also successful.

Figure 15.

One interval from an E region irregularity event on 9 March 2012 as observed using the ISIS Array. Two frequencies are shown (97.5 and 105.1 MHz), and the data shown were collected using two FM log periodic antennas at the Haystack site in a front to back configuration (i.e., single-site passive radar). Adaptive RLS filtering has been applied to remove the direct path signal and local ground clutter. The cross-ambiguity range doppler plots are normalized to the median clutter floor of the individual frequencies. In this example 50 kW ERP reference transmitters on 105.1 MHz located near Providence, RI (WWLI) and 97.5 MHz near Dover, NH (WOKQ) were used. This plot shows the signal to clutter ratio (SCR) normalized to the median clutter floor of the cross-correlated signals.

5 Summary

[68] We have discussed the design and implementation of the ISIS Array (Intercepted Signals for Ionospheric Science) and presented a range of observations using the network of receivers. Specific examples have been shown that demonstrate the implementation of a set of radio and radar techniques with an emphasis on passive radar. This instrument array gives an initial demonstration of the capabilities provided by a coherent software radio array that is designed for observation of phenomena in the Geospace environment. Going forward we intend to use the system for monitoring of the plasmasphere boundary layer as well as measurement and scientific study of other ionospheric phenomena.

[69] Networks of coherent software radio systems are capable of capturing signals over widely distributed spatial regions. These signals can be analyzed in a software defined manner that allows for highly flexible and precise control of the experimental process. This analysis can include many radio science techniques of proven scientific value. Distributed software radio arrays can provide simultaneous wide spatial coverage and high resolution. Large scale arrays of this type promise to provide a powerful means of observing and monitoring the Geospace environment.


[70] We would like to acknowledge the extensive help of particular individuals in the development, deployment and testing of the ISIS Array hardware and software systems. At MIT Haystack Observatory, these individuals include Will Rogers, Chris Farrell, Steve Holmberg, and Bill Rideout. Additionally we wish to thank Wes Swartz, Alan Weatherwax, Jim Labelle, as well as the students at the University of Washington, Dartmouth and Siena College, who helped greatly with the installations at their respective sites. We also thank the 2007 and 2010 MIT Haystack Observatory REU students who made significant contributions to enabling the satellite beacon observations made with the ISIS Array system. This work has been supported by the Air Force Office of Scientific Research under the DoD DURIP program and the National Science Foundation under AGS 0856093 and AGS 0733510. We gratefully acknowledge the support that has been provided for this project.