Introduction
Australia has had almost 40 years of operational experience with skywave Over-The-Horizon Radar (OTHR) systems. The early Australian OTHR experiments and prototype developments were conducted by the Australian Defence Science and Technology Group (DSTG), and their challenges, political frustrations and performance highlights are well described by Sinnott [1]. Through the support of BAE Systems Australia, these early activities culminated in the development of an operational radar in the early 1980s, capable of detecting aircraft and later ships. The radar was named Jindalee, an indigenous word meaning “bare bones”. This prototype radar was located in central Australia near Alice Springs, Northern Territory. This radar has subsequently become known as the Jindalee Facility Alice Springs (JFAS) to distinguish it from the two subsequent Australian OTHR installations.
The JFAS radar incorporated the following features, many of which remain common to current generation skywave OTHRs:
Propagation using the ionospheric E and F layers.
Transmit and receive site separation of approximately 100 km to allow transmission of continuous waveforms.
1000–3000 km range coverage, and 90 degree azimuthal coverage.
A frequency agile, high-powered electronically steerable transmit array capable of transmitting several hundred kW in the 5–28 MHz HF frequency band.
A receive array some 3 km long, featuring 462 pairs of dual monopole antennas, and 32 receivers deployed in a receiver-per-overlapped-subarray configuration.
Deliberately mismatched receive antennas to balance elevation performance against external noise [2].
Receive beamforming to produce multiple simultaneous “finger-beams” within the transmit footprint.
The ability to choose between “full radar” and two “half-radar” modes, balancing coverage against sensitivity [3].
“Step-scan” operations, allowing observations of multiple geographic areas with a 30 second revisit rate.
An ionospheric sounder network making real-time measurements to augment a climatological ionospheric model.
A separate cognitive Frequency Management System (FMS) used to provide propagation advice from the highly variable ionosphere [4].
A Clear Channel Advice system that provides candidate frequencies which will not interfere with existing users.
Significant Signal Processing capabilities to extract the weak target signals against the Earth surface backscatter and noise floor, enable track-while-scan, and provide radar-to-ground coordinate registration.
By the mid 1980s the radar was declared ready to be operated by the No. 1 Radar Surveillance Unit (1RSU, renamed the No.1 Remote Sensor Unit in 2015) of the Royal Australian Air Force (RAAF).
JFAS soon proved its worth as an invaluable asset to surveil thousands of square kilometres of ocean directly north of the Australian coast, demonstrating a picture of both air traffic and ocean vessels entering a sector of Australian airspace or coastal waters. Consequently, in 1986 the Australian Government announced the development of two new OTHRs to augment JFAS. The contract tender was won by Telstra, who handled early development until RLM, a joint venture of Lockheed Martin Australia and the Tenix Group, assumed full responsibility for the project in 1997.
By 1995 the specified system operation of the two new OTHRs was outlined by Cameron [3], with full operational capability delivered in 2003. Named the Jindalee Operation Radar Network (JORN), the new radars were separated across the width of Australia. Radar 1 (or R1) was sited at Longreach, Queensland, and Radar 2 (R2) at Laverton, Western Australia. The radar locations were chosen to complement each other with regards to range coverage, ionospheric propagation path diversity (hours and area of coverage), and the observability of targets flying tangential to either radar. The Laverton radar features two perpendicularly oriented “aspects”, providing 180 degree coverage, while the Longreach radar features a single aspect providing 90 degree coverage (see Fig. 1). The JORN radars incorporated an extended frequency range (5–32 MHz), and were equipped with a receiver-per-antenna, providing more flexibility and full control over spatial discrimination in comparison to the receiver-per-overlapped-sub array used in JFAS.
Location of JORN radar sites, Battlespace Surveillance Centre (BSC), and radar coverage: Longreach (R1, green); Laverton (R2, blue); Alice Springs (R3, red).
The JORN capability is managed and operated by the RAAF from the Battlespace Surveillance Centre (BSC) at RAAF Base Edinburgh, South Australia, and is acquired and sustained by the Capability Acquisition and Sustainment Group (CASG).
JORN continues to grow through ongoing innovation, driven by DSTG research, and a phased acquisition approach for transition to capability, driven by industry through phases of “Defence Acquisition Project AIR2025” (hereafter JORN Phase 3–6). Some of these innovations are captured firstly by briefly describing the prior JORN Phase 5 Enhancements in Section II, and then the in-progress JORN Phase 6 Mid-life Upgrade in Section III. Aspects of system development are discussed in Section IV and potential future developments in Section V.
JORN Phase 5
With the initial JFAS prototype radar development retrospectively referred to as JORN Phase 1 and Phase 2, the design and building of JORN was known as JORN Phase 3 and 4. During the JORN Phase 3–4 period, DSTG and BAE Systems continued to improve the performance and function of JFAS, and in many respects the older JFAS had become the more capable radar. This research and development, and the need to rationalize maintenance and operations across the systems, culminated in a new Defence Project “JORN Phase 5,” which was delivered by BAE Systems and Lockheed Martin Australia. JORN Phase 5 fully incorporated JFAS into JORN as Radar 3 (R3), and introduced JFAS improvements and new capabilities into JORN [5], including:
Reduction of signal processing losses in Doppler processing [6], [7] achieving twice the coverage capability (based on DATEX [8]).
Enhanced signal processing capacity to improve range-depth coverage, particularly for the stacked half-radar range mode described in [3].
Stare mode operations, where the radar observes the same geographic area to provide high revisit rates, allowing improved measurement of manoeuvring targets.
A cognitive signal processing suite incorporating new advanced signal processing capabilities [9]–[13].
Two-dimensional Numerical Ray Tracing to improve radar-to-ground coordinate registration accuracy [14].
JORN Phase 5 also integrated JORN and JFAS human-machine interfaces (HMIs) to provide a consistent look-and-feel for operators and reduce training overheads.
JORN Phase 6
During the JORN Phase 5 timeframe, the evolving Australian Defence capability needs and gaps were once more reviewed, with a determination that JORN was a vital asset fulfilling a long-term wide-area aircraft and surface vessel surveillance role. The in-progress acquisition project “JORN Phase 6” (Mid-life Upgrade) was created with objectives to extend JORN life-of-type to beyond 2040, together with an improvement in aircraft detection sensitivity based on DSTG R&D performed during the JORN Phase 5 period. The JORN Phase 6 project also needed to address obsolescence issues and component commonality across the three radar sites (noting some JORN components dated back to the original JFAS operational demonstrator of the mid 1980s).
The JORN Phase 6 upgrade encompasses all system components except the antennas and cabling of the radars and ionospheric sounders. The extant FMS receive array, cabling and processing will also be maintained to continue the long-term synoptic FMS data collection [15], which for JFAS now extends over three solar cycles. The following subsections describe the most significant system component upgrades and resulting capability enhancements.
A. Sampling Bandwidth
The current JORN receivers are of a heterodyne design, with analog narrowband filtering of the received signal prior to digitization. The development of receivers capable of high speed sampling with sufficient dynamic range and linearity enables the replacement of heterodyne receivers with broadband devices that digitize the entire HF bandwidth, relying on digital down-conversion techniques to extract the bandwidths of interest.
The digitization of the entire HF band greatly increases the utility of the radar system, since there is no longer the need to tune to a particular carrier frequency - because multiple down-conversions can be performed on the same data stream, and multiple bandwidths can be monitored concurrently using the same receive aperture. Accordingly, the Australian HF community refers to this capability as “Common Aperture” operation.
Common Aperture operation allows considerably increased functionality compared to the extant JORN radars, allowing simultaneous use of the full main receiver array for:
“Half-radar transmit” observations, reducing the sensitivity loss compared to “full radar” observations from the current 9 dB to 6 dB, a 3 dB improvement.
Passive and active channel evaluation, whereby candidate frequency channels can be evaluated using the full main receiver array gain and the operational signal processing algorithms. Passive channel evaluation assesses channel noise characteristics, while active channel evaluation assesses clutter characteristics using FMS “mini-radar” transmissions.
Potential bistatic operation, through passive reception of transmissions from other JORN radars.
High azimuth resolution FMS operations, thereby enhancing environmental condition assessment compared to the extant small FMS receive array.
Real-time diagnostics by using frequencies within the three-octave 5-45 MHz receiver bandwidth, but outside the bandwidths used by the radar at the time.
Common Aperture FMS backscatter sounder ionograms allow improved parameter advice and ionospheric modeling, while the enhanced FMS mini-radar observations allow the measurement of land-sea maps [16], aiding radar-to-ground coordinate registration.
Despite its many advantages, Common Aperture operation is difficult to implement in the HF large signal environment since any out-of-band signals are no longer rejected by analog filters. This in turn places exceptionally high-performance requirements on the receivers to ensure that they can operate in such a wide spectrum environment.
As so-called Government Furnished Material, DSTG delivered a Common Aperture receiver that broke the nexus between noise figure and linearity. Using a novel architecture, BAE Systems further developed the JORN Phase 6 Common Aperture receivers to exceed the specification. With the need to produce several thousand receivers over all three radars, a large emphasis was placed on manufacturability.
B. Noise Figure
The HF spectrum in which an OTHR operates is congested and inherently noisy due to anthropogenic, atmospheric (e.g. lightning), solar and galactic noise [17], [18]. This external noise excludes internal instrumentation noise and strong signals from users (such as communications, broadcasters and radars), and is often referred to as “ambient” or “background” noise. Using the aforementioned long-term JFAS synoptic database [15], a Design Noise Reference (DNR) was defined for JORN, with the JORN system required to be “externally noise limited” - i.e. exhibit system noise that is never greater than the external environment. The DNR is based on the knowledge that the optimal frequency for an OTHR is dependent on the time of day, with lower frequencies predominant during night-time operation due to the lower ionospheric F2-layer electron densities. Coincidentally, higher external noise levels are observed at night due to the collapse of the ionospheric D-layer, which during the day causes non-deviative absorption of HF radiowave signals [19]. This is the principle behind the deliberately mismatched receive antenna [2], which also allows the cost-effective preservation of elevation performance.
Since the total system noise is the sum of external and internal noise, and the “externally noise limited” specification requires only that internal noise be no greater than external noise, the total system noise is on occasion doubled. Thus for JORN Phase 6 the definition has been refined to produce a receiver design that features an exceptionally low noise figure over the entire dynamic range over which the receiver will be required to operate. It is expected that the reduction in total system noise will increase detection of low Radar Cross-Section aircraft, as well as simplifying the gain management algorithms required to remain externally noise limited over the full dynamic range.
C. Large Signal Environment
A demanding problem when designing and testing an OTHR is that the ionospheric conditions vary according to diurnal, seasonal and solar sunspot cycles, and geographic location. Each of these cycles influence the performance requirements over which an OTHR receiver must operate, creating large variations in the dynamic range that is necessary to endure the incidence and intensity of the full HF spectrum. Although performance verification over a diurnal cycle is possible, testing over an annual cycle becomes difficult, and testing over a solar cycle is intractable for any sensible test program. Hence for design purposes it was necessary to predict expected dynamic range requirements for the receiver.
JORN Phase 6 has addressed
this problem by using the same multi-decade synoptic database used to
determine the noise environment discussed above. The database also
contains observations of the strong signal spectrum recorded at 2 kHz
resolution. Fig. 2
shows the probability distribution of the total signal power at the
receiver's input (blue) over a full solar cycle. To determine dynamic
range requirements, the expected noise power at the receiver input must
be estimated. This may be derived from the external noise energy
database at the radar operational frequency. Assuming the radar may
operate on any clear channel across the HF spectrum, we examine the full
HF external noise spectra that were concurrent to the occurrence of
each signal power level. Overlaid in red is the lower ventile of the
external noise energy (shown as “external noise factor” with reference
to −204 dBJ). Due to small sample statistics, the ventile statistic has
high variance for the low occurrence signal power levels
Example of measurements of the probability distribution of the large signal and noise environment over a solar cycle at a JORN site.
To verify receiver performance over the full solar cycle, DSTG developed a Receiver Evaluation Environment that allows the reconstruction of an informed synthetic RF environment over the full operating range of the receivers (see Fig. 3). Due to linearity limitations of commercially available digital-to-analog devices compared to analog-to-digital, multiple waveform generators are used to reconstruct the full receiver bandwidth. Thus, the full performance envelope of the receivers was bench tested during hardware development. By adding other synthetic signals, secondary performance aspects such as the receiver's intermodulation and cross modulation performance were also measured.
D. Timing
JORN Phase 6 incorporates custom technology based on cryogenically cooled Sapphire Crystal Oscillators (CSOs) (see Fig. 4) developed by Australian company QuantX Labs [20], [21]. CSOs have been demonstrated to provide long-term timing stability and levels of phase noise traditionally seen only in national standards laboratory settings. This high-fidelity timing and frequency source removes prior limitations to other dependent transmitter and receiver system components.
E. Waveform Generation
Like the receiver, the JORN Phase 6 waveform generation is based on direct digital conversion, discarding traditional fixed and variable frequency oscillators. The direct digital approach is enabling exceptional levels of signal purity.
F. High-Power Amplifier Replacement
Although not part of the original JORN Phase 6 upgrade proposal, increasing maintenance and sustainment costs and the recent development of high-fidelity high-power amplifiers (HPAs) by Australian company Schach RF have motivated the decision to replace the HPAs at all JORN sites.
G. Wide Transmit Capability
In environments where the radar is clutter limited (e.g. low radial velocity surface vessel detection against a sea clutter background) the transmit system may have excess power available. This power may be exploited by adapting the transmitter phased array weights to spread the transmitter footprint over a wider azimuthal span, without losing any detection performance, since the radar is not noise limited. A consequence is of course a required increase in receive site information and communications technology (ICT) capacity to process additional receive finger-beams.
Receiver evaluation environment used to generate a synthetic HF environment to measure receiver intermodulation and cross modulation performance. The racks contain 32 waveform generators, which produce signals spanning the 5 to 32 MHz operating range of the JORN radars.
Laboratory testing of the cryogenically cooled sapphire crystal oscillators. Right: Sapphire crystal sample.
H. Signal Processing
Computer power has grown exponentially since the original JORN specification and becomes an enabler for significant signal processing improvements. As an immobile land-based asset, OTHR has no significant weight, space or power constraints. Moreover, as a wide-area early warning capability for air and maritime surveillance there are low demands in terms of processing latency.
The signal processing components that will be incorporated into JORN Phase 6 include:
Simultaneous processing of radar, active, and passive channel evaluator dwells, including half- and wide-transmit radar dwells.
A reduction of conventional range and beam tapering losses through auto-regressive bandwidth [22] and antenna array extrapolation, thereby enhancing range and azimuth resolution.
Improved tracking for stare-mode operations.
Three-dimensional Numerical Ray Tracing through the ionosphere, thereby improving the modelling of geomagnetic-field induced “out-of-plane” ray-path deviations [23].
Increasing the number of ionospheric propagation modes ray-traced through the ionosphere, including high-ray modes [24].
I. Ionospheric Sounders and Transponder Upgrades
JORN Phase 6 will replace the transmitter and receiver hardware of all ionospheric sounders [25]–[28] and transponders. Multichannel receiver technology has enabled a significant increase in oblique incidence pathways [27], and additional sounder sites will extend sounder geographical coverage. This will increase the real-time ionospheric model (RTIM) fidelity and improve radar-to-ground coordinate registration accuracy.
J. Open Architecture
Like many Defence Projects of the 1990s, JORN was developed using the mandated software language Ada. The Australian experience with Ada realised the need to enable more efficient software control and modification. For this reason the JORN Phase 6 software is being developed:
In C++, allowing the project to draw from a far greater pool of qualified software engineers.
Using a Data Distribution Service (DDS) to provide a modular and scalable framework of open system architecture to allow easier addition of software modules, and increase the reliability and performance of the radar.
Reversing the previous Ada mindset whereby an operator request was denied unless explicitly allowed, to one where, unless it runs the risk of safety or damage to equipment, the operator's request will be honoured.
K. Enhanced Radar Site Commonality
JORN Phase 6 introduces significant JFAS technical and facilities upgrades designed to enhance radar commonality introduced in JORN Phase 5. The former includes the upgrade of the receiving system from a receiver-per-overlapped-subarray to a receiver-per-antenna system. The latter includes upgrades to the operational and accommodation facilities to a similar standard as the Longreach and Laverton radars, allowing fly-in fly-out crewing in line with those radars.
L. BSC Refresh
The Battle space Surveillance Centre will undergo a complete ICT refresh. This will incorporate new operator interfaces and improved versions of the extant displays to enhance ease of use.
System Development
JORN Phase 6 System Design has adopted an “edges-in” development approach, developing “top-down” specifications concurrently with “bottom-up” hardware development. The simultaneous development of system requirements and hardware implementability has uncovered many details in the system design that would otherwise not have been uncovered until the testing phase. Like almost any large project the start has been slower than expected, however it is believed efficiencies will be realized as lessons are learnt early.
By the time that the Critical Design Review is scheduled, a comprehensive set of test results based on real-world data will be available.
Future Development
The Australian Government has recently announced the expansion of JORN to provide wide area surveillance of Australia's eastern approaches [29]. Although timescales have not been announced, multiple location and configuration options are currently under consideration by DSTG, each with its own associated benefit and cost. Modeling studies using the techniques described in [30], [31], are currently underway to inform the final design decision.
The continual funding and progression of Australian OTHR innovations have and are being captured within JORN through its acquisition phases. Continued R&D looks to further evolve and expand capability of the JORN system and may include:
Antenna design improvements, such as “grail” transmit antennas (see Fig. 5), which may reduce aeolian-induced phase noise associated with log-periodic dipole curtain arrays [2], and thereby improve the detection of slow-moving targets.
Regular overdense sparse receiving arrays (ROSA), providing enhanced system sensitivity through improved rejection of anthropogenic and galactic noise [32], [33].
Mode Selective Radar, employing multiple-input multiple-output (MIMO) techniques to allow the selection of individual ionospheric propagation modes, thereby improving target detection in and around disturbed clutter, and reducing mode association problems to improve tracking performance [34].
In addition to skywave OTHR operation, HF radars can operate in line-of-sight (LOS), SkyLOS (skywave transmission, LOS reception), and multi-static combinations of these modes. HF and lower VHF LOS radars are ideally suited to high-speed weapon (HSW) detection as the sizes of HSWs are comparable to the radar wavelength, placing them in the resonant scatter region where RCS enhancements can be observed at certain aspect angles [35]. HF and lower VHF LOS radars have demonstrated detection of resident space objects in low earth orbit [36]–[37], given the development of signal processing capabilities that tolerate the high speed (up to 8 kms−1) and acceleration that are required for HSW detection. SkyLOS radar provides a low-cost means of augmenting a skywave over-the horizon radar, obviating the need for ionospheric measurement and modeling [38]. DSTG continues to develop HF LOS and SkyLOS mode capabilities for HSW detection aimed at future inclusion into JORN Phase 6 or subsequent JORN upgrades.
Another area of active DSTG research informing future development is artificial intelligence (AI) (e.g. [39]), primarily motivated by the need to increase automation to reduce the workload on JORN operators associated with a “data deluge” created with the JORN Phase 6 enhanced signal processing capabilities. The JORN Phase 6 open system architecture will allow faster and improved incorporation of new software capabilities such as AI, allowing rapid provision of these capabilities to operators.
Conclusion
The Jindalee Operational Radar Network had its foundations as a long-range wide-area surveillance operational demonstrator in the 1980s. Australia has enacted a sustained R&D program to provide innovations that have significantly improved and expanded Australia's sovereign OTHR capability. This transition of innovation to capability continues with the “AIR2025 Phase 6 JORN Mid-life Upgrade”. JORN Phase 6 is aimed at establishing life-of-type extension beyond 2040, and will deliver substantial capability enhancements by taking advantage of the significant improvements in digital processing capabilities.
The “Common Aperture” receiver, enabled by full HF band direct-digital receiver technology, allows multiple simultaneous reception channels for multiple functions, including support for cognitive and adaptive radar. The reduction in hardware required has allowed JORN Phase 6 to concentrate effort on a relatively small number of components, allowing those components to achieve exceptional performance.
ACKNOWLEDGMENT
The authors thank Mr Brett Northey and Dr Trevor Harris for figure contributions, Dr Gordon Frazer for photographs, and colleagues from BAE Systems and ISSD, DSTG for their insightful comments.
No comments:
Post a Comment