Over the Horizon Sky-wave Radar: Coordinate Registration by Sea-land Transitions Identification

Transcription

1 Progress In Electromagnetics Research Symposium Proceedings, Moscow, Russia, August 18 21, Over the Horizon Sky-wave Radar: Coordinate Registration by Sea-land Transitions Identification F. Cuccoli 1, L. Facheris 2, D. Giuli 2, and F. Sermi 1 1 U.O. CNIT, c/o Dipartimento di Elettronica e Telecomunicazioni, Università di Firenze Via di Santa Marta, 3, Firenze, Italy 2 Dipartimento di Elettronica e Telecomunicazioni, Università di Firenze via di Santa Marta, 3, Firenze, Italy Abstract In this paper, we propose an approach to the problem of real time coordinate registration for Over the Horizon Sky Wave signals. The approach is based on the a priori knowledge of the displacement of the sea-land transitions within the radar coverage area, namely, takes advantage of the geo-morphological structure of the surveillance area, employing it to build a binary mask to be used as a geographic reference for the received radar echo. The georeferencing algorithm, based on the maximization of the cross-correlation between the received radar echo and the binary clutter signatures, is outlined in order to point out the minimum requirements in terms of received signal-to-noise ratio and differential sea-land backscattering coefficient. 1. INTRODUCTION Despite it is based on a well known technology, the Over The Horizon Sky Wave Radar (OTHR-SW), developed during World War II and largely employed during the Cold War, is attracting today much interest, thanks to the great steps forward made in the last years by the signal processing and data storage techniques. OTHR systems are HF-band sensors that, via surface propagation (Ground-Wave) or ionospheric reflection (Sky-Wave), cover an area that is located well beyond the horizon line, that is the natural operational limit for most ground-based radars. In particular, OTHR-SW are the only ground-based sensors with a surveillance area comparable in surface with that of satellite constellations or airborne radar networks. This is achieved by exploiting the propagation characteristics of the Ionosphere (see Fig. 1 for a simplified sketch, where are also defined some parameters used in the following). Nevertheless, the employment of the Ionosphere as part of the Tx-Rx channel implies a substantial uncertainty in the actual ray path. In fact, the non-homogeneous structure and time-dependent behavior of the Ionosphere has an important impact on the OTHR-SW Coordinate Registration (CR) procedures of current common use. The consequent uncertainty propagates to every radar measurement and the estimated position of an eventually detected target may result heavily inaccurate. So far, several different approaches to the CR task for OTHR-SW systems have been developed. Most of them require external sources of information arranged within the surveillance area, such as ionosonde networks [1, 2], beacons or transponders [3], passive receivers [4], additional radar systems [5], etc. Often, the employment of these methods in an operative radar context is limited by one of the following factors: the construction/maintenance costs growth due to the introduction of external systems; Figure 1: OTHR simplified operative scenario scheme.

2 22 PIERS Proceedings, Moscow, Russia, August 18 21, 2009 the need to arrange those systems within the surveillance area, that generally is a restricted access zone; the sporadic nature of additional information availability. The OTHR-SW real time CR approach proposed in this paper does not ask for any information source external to the radar, but exploits the a priori knowledge of the geo-morphological structure of the surveillance area in order to georeference in real time the received echoes. In the following sections we describe the CR method we developed together with the main hypotheses on the OTHR-SW simplified scenario and with some relevant simulation results. 2. THE PROPOSED REAL TIME CR METHOD Although the OTHR transmission channel is non-homogeneous and time-dependent and therefore extremely hard to model (eventually requiring continuous ionospheric soundings), the geomorphological structure of the surveillance area can be easily and profitably employed as static geographic reference. An ad hoc procedure (not described here) to generate a static Binary Clutter Mask (BCM ) must be applied before starting the real time echo processing, with the objective to mark the sea-land transitions within the radar coverage area. When the radar is gathering data pertinent to a given azimuth θ, the corresponding binary clutter profile is extracted from the BCM. An ensemble of equivalent ionospheric heights is then defined and for each of them the corresponding radar footprint projection is evaluated. Such ensemble is centered around a first guess value of the equivalent ionospheric height as well as provided by seasonal ionospheric statistics over the geographic areas under test. The projections of the radar footprints on the binary clutter profile provide a set of binary clutter signatures (BCS). Every BCS is cross-correlated with the received radar echo and the maximum (the best match ) is then selected (dashed line in the example of Fig. 4). The chosen BCS unambiguously 1 corresponds to a certain value of the elevation vector. This value is selected to be the estimate of the actual ionospheric equivalent reflection height (h eq ) and consequently of the radar footprint position relative to the received echo. 3. MODEL HYPOTHESES AND SIMULATION SCENARIOS In order to prove the reliability of the method and to determine possible restrictions to its employment in an operative radar system schedule, we designed a simplified model of the complex OTHR-SW scenario. In this section, we outline the most significant hypotheses assumed for this purpose, explaining how they affect the CR procedure applicability in an actual radar context. Figure 2: Radar footprint and clutter area (radar patch). c = light speed; τ = pulse length; β = take-off angle; δi = pulse projection; θ = azimuthal angle; δθ = radar beam azimuthal span angle; Ac = clutter area. The radar system on which the proposed CR method is developed is a monostatic OTHR-SW that transmits a vertically polarized unmodulated pulse of length τ, with a take-off angle β and a beam azimuthal span angle δθ, as shown in Fig In order to avoid the possibility of similar clutter signatures on the same azimuth, it is necessary to employ the proposed CR procedure on favourable (in terms of sea-land transitions dislocation) binary clutter profiles.

3 Progress In Electromagnetics Research Symposium Proceedings, Moscow, Russia, August 18 21, The Earth magnetic field effects on ray propagation are neglected, while the spherical geometry of the planet and the ionospheric layers is substituted by its projection on parallel planes. Note that, since for a given radar position the former aspect appears time-independent and the latter is just geometric, they can both be taken into account by a constant multiplicative coefficient. The propagation path (that includes the ionospheric reflection) is the scenario s aspect that requires more restrictive assumptions in order to fit in a simplified geometrical model. It s assumed that a stable single-hop ionospheric path is available (hypothesis that appears to be correct at Europeans latitudes) with mirror ray reflection, identical for transmitted signal and received echo. At this stage the iono-scattering effect is neglected 2, i.e., the Ionosphere is modeled as a stratified medium, with no geographic gradient. Figure 3: OTHR simulated scenario hypotheses. Figure 4: Snapshot of a simulation result. On the left, the main radar parameters: sea & land backscattering coefficients (σ m and σ t ); actual & estimated ionospheric reflection height (h eq and h est ); clutter-to-noise ratio (CNR). On the right, plot of amplitude versus down range (echo pertinent to a single radar pulse). The surface clutter is considered binary (that implies sharp-cut sea-land transitions). Such hypothesis is confirmed by the marked difference between sea and land backscattering coefficients σ m and σ t [8, 9]. A spatially correlated Gaussian clutter is assumed [6, 7]. 2 Generally multipath distorts the surface backscattered signature by superimposing signals scattered at different ground ranges in one radar range bin [2]. This effect has not yet been taken into account in the model.

4 24 PIERS Proceedings, Moscow, Russia, August 18 21, 2009 The system noise for an OTHR-SW originates predominantly from the external, divided into three main components: atmospheric, cosmic and man-made. We represented the overall noise as an additive Gaussian process with assigned (in accordance with the radar site location) mean value, statistically independent from clutter. The radar footprint, intersection between an elliptical base cone (the radar beam) and an oblate spheroid (the Earth surface), is geometrically approximated with an ellipse, while we suppose the clutter area limited in range by the radar pulse length projection on the Earth surface (see Fig. 2). Fig. 3 summarizes schematically the assumptions made for the basic scenario components. 4. RESULTS AND CONCLUSIONS Simulations have been performed under different geographic contexts, in terms of number and relative location within the radar footprint of sea-land transitions, intensity of backscattering coefficients, size of the clutter area, etc. In the following, we discuss the most significant results providing a performance analysis of the proposed CR method. Figure 4 shows a simulation result carried out under the hypotheses outlined in the previous section, with a plot of normalized amplitude versus down-range distance of the received radar echo (continuous line), the selected (cross-correlation maxima) binary clutter profile (dashed line) and the clutter area profile during the radar footprint scan (dash-dot line). The results of simulation trials in different scenarios suggest that in case of Clutter-to-Noise Ratio (CNR) higher than 3 db (CNR = db in this test) a minimum differential sea-land backscattering coefficient ( σ = σ m σ t ) of 7 db/km 2 is required to guarantee an error smaller than 3 km in the estimate of the actual ionospheric equivalent reflection height (h eq ). Such requirement is promising for the feasibility of the proposed CR method. In fact, even if the bibliography about real HF sea-land backscattering coefficients is scarce, in [9] we find the following statement: Extensive observations [... ] indicated that, averaged over a wide area, sea clutter power levels were about an order of magnitude higher than those from an area of similar size in the central United States.. Note that in the proposed simulation a ionospheric reflection height bias of hundred meters leads to an error in the radar footprint range position estimation of about one kilometer. This value is quite satisfactory if compared with the OTHR-SW down-range resolution (dr-res) that, with a pulse length τ = 0.1 s and a take-off angle β = 13, corresponds to about 15.4 km. In conclusion, the simulation results we presented here suggest that the proposed method, based on a priori knowledge of the sea-land transition dislocation within the surveillance area, can be actually exploited for CR procedures in an OTHR-SW operative context. The next step in this OTH radar context is the development of a specific simulation tool allowing an easier management of the simplified scenario hypotheses and of the radar parameters. It will be possible to improve the overall model with more complex and realistic hypotheses, in order to test the CR method in scenarios closer to the real ones. In particular we are interested in the development of a complete ionospheric model accounting of non uniform space-time structure. ACKNOWLEDGMENT The authors would like to thank Luca Capannesi for his technical support. REFERENCES 1. Krolik, J. L. and R. H. Anderson, Maximum likelihood coordinate registration for Over-The- Horizon Radar, IEEE Transactions on Signal Processing, Vol. 45, No. 4, , Wheadon, N. S., J. C. Whitehouse, J. D. Milson, and R. N. Herring, Ionospheric modelling and target coordinate registration for HF Sky-Wave Radars, Sixth International Conference on HF Radio Systems and Techniques, , Jul Weijers, B. and D. Choi, OTH-B coordinate registration experiment using an HF beacon, IEEE International Radar Conference, 49 52, Frazer, G. J., Forward-based receiver augmentation for OTHR, IEEE Radar Conference, , Apr Torrez, W. C. and W. J. Yssel, Over-The-Horizon Radar surveillance sensor fusion for enhanced coordinate registration, IEEE Proceedings on Information, Decision and Control, , Feb Ward, K. D., R. J. A. Tough, and S. Watts, Sea clutter: Scattering, the K distribution and radar performance, The Institution of Engineering and Technology, 2006.

5 Progress In Electromagnetics Research Symposium Proceedings, Moscow, Russia, August 18 21, Khan, R. H., Ocean-clutter model for high-frequency radar, IEEE Journal of Oceanic Engineering, Vol. 16, No. 2, , Barrick, D. E., Remote sensing of sea state by radar, OCEANS, Vol. 4, , Sep Skolnik, M., Radar Handbook, 2nd Edition, Chap. 24, McGraw-Hill, 1990.