GB2160686A - Identification of ships - Google Patents

Abstract

Apparatus for determining ship's identity provides a radar "finger print" for a ship. A number of radar receiver channels with closely spaced overlapping intermediate frequencies fL01, fL0Z... classify the ship radar r.f. frequency into one of a number of groups. The summed integrated video outputs are threshold detected and entered into a shift register, the contents of which are analysed to determine the pulse repetition frequency of the ship's radar. The individual receiver outputs are also digitised and analysed to determine the ship radar antenna rotation rate. Sufficient differences exist between different makes and models, also between nominally identical radars, to enable a large number of unique combinations of characteristics to be distinguished and allocated to known ship identities. Thereafter a ship can be identified simply by observing its radar 'finger print'. <IMAGE>

Description

SPECIFICATION Identification of ships This invention relates to an apparatus and method for determining identity of ships to enable them to be tracked, e.g. by coastguards, without requiring either a continuous flow of information from the ships or new equipment to be fitted onboard the ships.

Apart from strategic reasons there are a number of situations in which it is beneficial to know the identity of a particular vessel in a particular position: (a) There is a communication originated ashore destined for a ship, the location of which is unknown. Currently traffic lists are transmitted from coast radio stations near where the ship is believed to be. If there is knowledge of that ship's position there will be a higher probability of the communication reaching the ship more rapidly with less spectrum wastage.

(b) If there is knowledge of a ship's position, directional antenna systems may be used at the coast radio station with higher communication probability in fringe areas.

(c) A shore authority like the Coast Guard observes on its radar that a particular target is steering into danger (a sandbank, a cable laying operation). Without knowing the target identity it cannot easily warn the target about the danger. Today's messages are general: "Vessel 3 miles SW of SANDETTIE light vessel heading NE you are standing into dan ger (d) A vessel of unknown identity is contravening IMO traffic flow regulations. The authorities want to fine him and thus need to know his identity.

(e) Vessel 'A' observes a potential collision conflict with target vessel 'B' and desires to communicate with it on VHF but does not know its name or call sign.

Identification of a particular radar target in aviation is generally by SSR type transponders or by sending an instruction "aircraft so and so do 90 turn". At sea the carriage of transponders is under consideration but problems of unfitted ships or non-working systems are still a long way from being resolved.

Most ships carry radar, most larger ships carry 2 or more. Generally they are left on whether the navigator studies it or not. The radar emissions contain a "finger print" in the form of a particular carrier frequency (magnetron characteristics), a particular pulse repetition frequency (PRF) and pulse width (selected radar range dependent), and a particular antenna rotation rate. Where there is a choice of PRF and pulse width (generally the case) the amount of choice is limited to 2 or 3 values.

The spread of values for any one parameter is limited but by examining all parameters the probability of several vessels having the same combination is low. Thus an examination of the spectrum transmitted provides a kind of fingerprint. Similarly the ship's engine and the propeller transmit an acoustic spectrum which helps to define a unique combination of characteristics as is well known to submariners.

Visual clues also have a role to play, number of decks, funnels, masts and navigation lights.

The present invention is concerned only with the radar finger print.

According to the present invention there is provided apparatus for determining identification of ships the apparatus comprising a radar antenna, a plurality of radar receivers each receiving signals from the antenna, said receivers each having a different one of a range of closely spaced intermediate frequencies, each receiver having individual means for integrating the receiver output, the apparatus also including means for comparing the integrated outputs whereby radar signals received by the antenna can be classified into different frequency groups and different PRF groups.

An embodiment of the invention is now described with reference to the accompanying drawings, in which: Figure 1 illustrates a receiver apparatus for determining ships identity, and Figure 2 illustrates receiver gain characteristics.

The apparatus of Fig. 1 consists of a radar scanning antenna feeding a number, e.g. 17, of individual radar receiver channels. Each channel includes an IF stage fed with a different local oscillator frequency f,O. The conventional marine radar band in the 3cm band covers the frequency range 9320-9500 MHz. Most of the energy of a radar pulse with 0.1 ,us width is contained in a band 20 MHz wide. Thus if 1 7 receivers are connected in parallel each with an IF 20 MHz wide centred 10 MHz above the previous one then the integrated outputs of those receivers provide an indication of carrier frequency.If the receivers have a triangular gain response within the passband (Fig. 2) (rather than the more common flat response) then a comparison of the integrated outputs provides a resolution of about 5 MHz thus dividing radars into one of 36 groups. Magnetron frequencies vary typically by 0.25 MHz per degree C and by about 10 MHz over 5000 hours.

To determine the PRF of a radar each channel provides a video output, the outputs from all the channels being fed to a summing network. The summed video outputs are passed through a threshold detector and thence to a 4 kilobit shift register. If the register is clocked at 1 MHz, it will hold 4 ms of data or 4 transmitter pulses from the radar if it operates at 1 kHz PRF. (Commonly there will be between 8 and 24 pulses from a radar for each antenna rotation). Thus we have a resolution of about 1 in 4000. If it were assumed PRFs when nominally 1 kHz, varied by + 2% from ship to ship it would mean dividing ships into one of 80 groups. (2% of 4000 pus = 80 its clocked at a rate of 1 per its).

Clearly an alternative to a 4 kbit shift register could be a simple computer store which would enable a more leisurely analysis of the data when the ship's radar scanner points away from the receiving antenna.

Another alternative is to perform a separate analysis from each of the 7 receivers thus being more easily able to cope with a multiplicity of simultaneous radars.

To determine the antenna rotation rate it would be necessary to 'listen' to the transmissions over several antenna rotations and to determine the level of received signal. Each channel feeds a sample and hold circuit followed by an analog-to-digital converter the output of which is buffered in a register. The data thus gathered is fed to a computer, which also generates the clock signals for each channel. The computer is programmed to analyse the antenna rotation rate, which may be derived from a study of the way the received signal level varies with the polar diagram of the ships rotating antenna. The maximum signal can generally be assumed to coincide with the centre of the main beam.

The time of maximum may be determined by averaging the points before and after the maximum when the signal level crosses a slightly lower threshold. Time between successive maxima provides rotation rate. It is probably easy to determine the interval between maxima to + 2 pulses which amounts to typically 2 in [1000 (PRF) x 60 (secs)/22 (RPM)] which is 0.40%. Greater accuracy would be possible by averaging over several pulses within each main antenna lobe. However allowance must be made for variations in wind and in onboard power supply frequency and possibly + 2.0% is all that could be achieved in terms of calculated rotations per minute describing a particular ship's radar.

Suppose for each nominal rotation rate we allowed 5 possibilities. Since there is a wide range of nominal rates (Decca 125/1 50 = 23, Seafarer Seascan 111 = 20, Marconi Predicta = 24, AEI 655 series = 22, Selenia = 22 or 27, Sperry Mk 1 6A = 18 or 2 Sperry Mk 127 = 1 5 or 30 RPM) it should be possible to classify rotation rates in at least 20 groups.

From these considerations we have 36 groups of carrier frequencies, 80 groups of PRF and 20 groups of antenna rotation rates, giving a possible 57,600 combinations. Thus there is a high probability of being able to identify a vessel in terms of its 'fingerprint'.

There are a number of occasions when a ship communicates by radio and has to state its identity in order to facilitate the setting up of the radio call. Some of these occasions are: 1. Crew and passengers telephone home.

2. Commander--owner periodic contact.

3. Ship contacts pilots 24H/1H before ETA.

4. Ship contacts arrival port and agent.

5. Coast radio station has traffic lists.

6. Ship to ship communication.

7. Channel Navigation Information System.

8. Officers request bunkers and stores.

It follows that there exists already the environment in which a ship communicates that identity several times while in transit. The first time a vessel approaching a coast makes a radio call within VHF/radar range a shore station complex does DF on the transmissions.

Thus identity of a target in a particular location becomes known. The shore station then examines the radar signature of the target at that position. This information is now stored and made available to all other interested parties. Thus an "international look-up table" is built up providing electromagnetic fingerprints of identity. If a ship wishes to call another but unknown ship in a confluence zone it asks the nearest coast station to identify the unknown ship's fingerprint and so its call sign.

This kind of procedure would enable tracking of targets even where there is a discontinuity in radar coverage. For instance the Esbjerg-Harwich ferry is out of coverage of the Harwich radar for most of the crossing but when again it enters the coverage area its radar signature will be recognised from previous visits.

A further extension of the scheme would be for a port from which a vessel leaves to check the fingerprint of the vessel and to transmit details to interested parties en route.

Some points which should be borne in mind when designing the apparatus of Fig. 1 and writing the software for the computer are: (a) Some ships carrying several radars, generally one at X-band and one at S-band.

This will increase the uniqueness of the vessel. If there are two radars operating in the same band it will slightly increase the complexity of extracting the two PRFs.

(b) Sometimes the ships that neglect to inform coastal authorities about their presence are the very ships whose radar does not function properly. However malfunction in receiver or display system does not likely affect the emitted characteristic. Thus the probability of 'detection' is better than the probability of onboard radar working.

(c) A potential problem exists in that the PRF is sometimes altered when the navigator switches maximum range on the onboard set.

(d) One could envisage a scheme in which a radar supplier notifies IMO or some similar body when fitting a new radar. Such notification could include the principal radiated char acteristics.

(e) The spectrum of the transmitted signal has line components spaced apart by the PRF and an envelope controlled by the shape of the pulse. The pulse characteristics could therefore be determined if receivers with some narrow bandwidth were employed and the integrated outputs of all the receivers were compared.

(f) In the above it is assumed that there is only one transmitting ship within the beamwidth of the decoding shore station. If there are several it will clearly require more sophisticated software. Two alternatives are possible not relying on software; one is to delay a study of received signal until the relative geometry is such that the various targets are no longer in line. The other would be a corelation technique utilising two shore stations since the various vessels could not possibly be in line for both shore stations.

Claims (8)

1. Apparatus for determining identification of ships, the apparatus comprising a radar antenna, a plurality of radar receivers each receiving signals from the antenna, said receivers each having a different one of a range of closely spaced intermediate frequencies, each receiver having individual means for integrating the receiver output, the apparatus also including means for comparing the integrated outputs whereby radar signals received by the antenna can be classified into different frequency groups and different PRF groups.

2. Apparatus according to claim 1 wherein each radar receiver has a triangular gain response.

3. Apparatus according to claim 1 or 2 further including means for summing the video outputs from all the radar receivers, a threshold detector through which the summed outputs are passed and means for storing the output of the threshold detector.

4. Apparatus according to claim 3 wherein the storage means comprises a clocked shift register.

5. Apparatus according to any preceding claim further including clocked sample and hold means in each receiver to which the integrated output is applied, analog-to-digital conversion means to digitise the sampled and held signals, means for storing the digitised signals, and means for analysing the stored digitised signals to determine the rotation rate of an antenna from which radar signals have been transmitted.

6. Apparatus for determining identification of ships substantially as described with reference to the accompanying drawings.

7. Apparatus for detecting, analysing and classifying the transmission characteristics of a ship's radar and correlating the classified characteristics with known ship's identity.

8. A method of determining a ship's identity including the steps of ascertaining and classifying the radio frequency of radar signals emitted by the ship, ascertaining and classifying the antenna rotation rate of the ship's radar, and correlating the particular combination of classified information with known ship's identity.