This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
"Never in the field of human conflict was so much owed by so many to so few" - Winston Churchill
1. Several types of radar signatures can be used to acquire information about the target aircraft. These may be divided into two families of techniques. Techniques from the first family are based on the radar reflections from the rotating parts of the aircraft i.e engine compressor, turbine, propeller, rotor etc. Such reflections have a characteristic signature and this can be used for identification. Some of the techniques of the first family are Jet Engine Modulation (JEM), Propeller Rotor Modulation (PROM) and Helicopter Rotor Modulation (HERM). While these techniques have different names, they essentially work in a similar fashion. Techniques from the second family are based on the radar returns of the 'Aircraft as a whole'. The advantages of the second family far outnumber the advantages of the first family of NCTR and hence would be more suitable for NCTR. Each of these techniques have been deliberated in the subsequent paragraphs. While analyzing the various NCTR techniques, it would be evident as to why the NCTR technology has not fully matured so far.
JET ENGINE MODULATION
2. Jet Engine Modulation (JEM) was one of first techniques to be incorporated in any fighter aircraft airborne interception radar. The earliest example of JEM is the 'Musketeer' program of USAF. A jet engine can be described as a series of propellers contained within a housing, with each of the propellers rotating about a central shaft. Using the radar, one can look into the jet engine if the target is flying towards or away from the radar (either compressor or the turbine of the engine). Since all parts are likely to be highly reflective at various radar frequencies, the electric field resulting from re-radiated energy will be extremely difficult to compute. However, it has been observed from experimental radar studies than certain key features can be extracted from signals deriving from jet engine reflections. Jet engine reflections are known to result in periodic amplitude and phase modulations upon the carrier signal. Hence the term 'Jet Engine Modulation' (JEM). For successful implementation of JEM algorithms, following assumptions are to be made:-
(a) Each engine blade acts as a homogeneous, linear, rigid antenna. Pitching and twisting of the blades is not considered.
(b) The jet engine is in the far-field of the Radar. Far field of a radar is the region from where distance from the source to the target is far enough, such that the electromagnetic wave can be considered a plane wave.
(c) The main contributions to the received target signature are derived from reflections off the engine blades at the compressor / turbine. Reflections from the housing and rotating shaft are not considered since they are along the wave propagation and will not result in any significant effects.
(d) The aspect ratio of each blade is such that the length is much greater than the width.
3. How is JEM Spectrum Generated. Like in any radar returns the reflections from the jet engine would also have a doppler shift corresponding to the relative speed of fighter / target aircraft and the radar carrier frequency. However in addition to this, the rotating blades of the engine would cause doppler shifted return to be modulated. Reflections off rotating jet engine compressor / turbine blades would be a chopped reflection of the impinging signal. The reflections are characterized by both positive and negative doppler sidebands corresponding to the blades moving toward and away from the radar respectively as can be seen in right side graph of Fig 5.1. Thus every engine stage wpould have a characteristic signature which can be used for recognition. The largest fraction of the radiation is reflected by the blades of the first rotor. A smaller portion passes along and is reflected by the second rotor. Theoretically, reflections from the subsequent stages are also included in the radar return, however reliable attribution from the subsequent stages is insignificant.
Fig 5.1 : Radar returns from Non Rotating and Rotating Objects
4. A characteristic JEM spectrum from a twin engined aircraft is shown at Fig 5.2. The central peak, the Body line, shows the reflection of the aircraft as whole. The term BCF denotes the Blade Chopping Frequency i.e the frequency corresponding to rotation of the first stage rotor over a single blade interval given by 360/NB where NB corresponds to the number of blades. As two BCF lines can be seen, it can be concluded that the aircraft has atleast two engines. Somewhat lower peaks in the spectrum under the phrase SRF are harmonics of the so called Shaft Rotation Frequency. This corresponds to a 360Â° rotation of a blade. Division of BCF by the SRF gives the number of blades. In a case wherein two different types of engines have same number of blades on the first stage, it is necessary to use features such as SRF and second stage rotor returns to resolve the ambiguities in classification. Same technique can be applied if the radar looks at the rear side of engine onto turbine blades.
Fig 5.2 : Jet Engine Modulation Spectrum
5. Drawbacks of JEM. The major drawback of this system is that it is heavily aspect dependant. The target detection is possible only within a small zone of frontal and rear sector. This may not be possible in a dynamic and dense combat environment. For successful classification using JEM signature, a large signal to noise ratio is required. With increase in range, noise level increases due to atmospheric attenuation. This would mean that JEM is suitable for classification only at relatively shorter distances. Considering the ranges of current generation BVR missiles, the ranges at which target aircraft are identified using JEM, may not be adequate for BVR combat. Also the JEM cannot be reliably interpreted in the following cases :-
(a) Aircraft with 3 or more engines where SRFs are not accurately synchronised.
(b) Engine types where the first and second stages are on different engine shafts rotating at different rates.
RADAR RANGE PROFILE
6. Single Dimension Radar Range Profile. One of the major drawbacks of JEM as brought out previously is that the engine of the target aircraft should be visible to the fighter aircraft's radar. To overcome this limitation, research scientists used the radar returns from the aircraft as whole to identify targets. These returns were termed as 'Radar Range Profiles'. A single dimension radar image of target is a plot of radar returns from the target aircraft as a function of time. This is also termed a 'Time Domain Signature'. This signature has only one dimension - Amplitude. Since different parts of the aircraft reflect differently due to varying reflectivity, each aircraft has a characteristic signature. This signature can be used to identify the type of aircraft. An example of a time domain signature is shown in Fig 5.3. The radar returns from the 'scatterers' (parts of the aircraft that give strong radar reflections) are plotted on a time scale.
Fig 5.3 : Single Dimension Radar Range Profile
7. Compared to the other types of NCTR, single dimension radar range profile can be considered most optimal. It is not limited by target aircraft aspect like the JEM. And unlike a two dimensional range profile (which would be covered subsequently), it does not require a complex radar like Inverse Synthetic Aperture Radar (ISAR). A single dimension radar range profile can be generated by the newer generation pulse Doppler radar with suitable modifications. Hence amongst NCTR techniques used globally, this technique is most commonly used. However since the target signatures are not easily discernable, a large library would be required. A one dimension radar range profile does not have features which can be related to the optical images of target. While optical images appear to us as three-dimensional images with brightness, contrast and colour of each element, range profiles only contain amplitude information from the larger target scatterers. In comparison to optical images range profiles are far more abstract. The range profiles contain information on the geometry of the aircraft for the given aspect angle and range.
8. Radar Resolution Required for Range Profiling. For obtaining a time domain signature of the target, the radar should have adequately small resolution so as to get different amplitudes from different parts of the aircraft. Radar resolution is a function of pulse width. To understand this, let us first understand pulse width. The pulse width or the pulse duration is the time interval when the radar is transmitting (Radar contains both transmitter and receiver and they work sequentially). The pulse width is to ensure that the radar emits sufficient energy to allow that the reflected pulse to be detected by the receiver. While the radar transmitter is active, the receiver input is blanked to avoid the amplifiers being damaged. Now if the distance to target is such that the transmitted pulse reaches back before radar changes over to reception mode (within pulse width), there would be no detection / signature. Therefore the minimum resolution depends on the pulse width of the radar. The radar resolution is given by the formula,
Resolution (in meters) = Â½ x Speed of EM Waves x Pulse Width.
Most Airborne Interception radars have a pulse width of the order of micro seconds. A microsecond pulse would correspond to a resolution of 150 m. This means that the profile of various parts of the aircraft within 150 m would be transmitted back in the single pulse i.e the profile would contain a single return from the aircraft as a whole. This resolution is acceptable for target detection but not for recognition. Therefore considering general size of fighter aircraft, a resolution of at least 1 m would be required for good target signature. This would correspond to a pulse width of 6.67 nanoseconds.
If the pulse width of the radar is narrowed down to nanoseconds, the detection range of the radar would be adversely affected. This poses severe limitations on the designer. For practical reasons most radars are not designed to generate very short duration pulses and support their transmission, reception and digitisation. To overcome this problem, a technique of stepping up Pulse Repetition Frequencies (PRF) linearly with time called as 'Stepped Frequency Waveforms' are used. Such techniques impose severe hardware limitations on the radar which would discussed in Chapter VII.
9. Constraints in Obtaining Range Profiles. The radar range profile obtained from the radar returns would be distorted due to a number of effects. The return signals need be reflected back towards the radar directly, it can bounce between different parts of the target aircraft. This induces additional time delay, which manifests as a different signature. Another effect, which can occur, is the 'shadowing' of one part of the target by another part. This occurs when part of the target is obscured by a large part of the same target located between it and the radar, significantly reducing the energy in the wave reaching it and being reflected back to the radar. When this occurs at a particular aspect angle, a certain part of the target may not contribute to the range profile. For example, the fuselage and wings of the target can obscure the tail at certain aspect angles to the radar resulting in an apparently shorter aircraft.
10. Two Dimension Range Profile. Instead of one dimension profile, a two dimensional radar range profile can also be generated. Purely in terms of target signatures, two dimensional profiles carry far more information than single dimension range profiles. Hence accurate recognition is possible with a smaller target signature library. However, they are more complicated, require lot of computational power and complex algorithms for generation. And therefore the radar hardware requirement and capability are much more. To generate a two dimensional signature two type of radars namely the Inverse Synthetic Aperture Radar (ISAR) and the Monopulse radars are used. The employment of monopulse techniques in radar target recognition is very much in its infancy. Hence we will discuss the more relevant ISAR in this paper.
11. Inverse Synthetic Aperture Radar. ISAR uses target's own rotational motion to generate a high resolution image in the cross range (Cross range and down range are equivalents of x and y axis in a 2D space). When this is employed with a high-range resolution waveform, a two-dimensional target signature is obtained, which can then be used for target recognition. ISAR is completely dependent upon the target having a relative rotational motion component in relation to the radar. These rotational components are integrated to form a cross range profile. Fig 5.4 shows how the target has relative rotation component due to changing aspect angle.
Fig 5.4 : Target Aircraft Rotational Component for ISAR
12. Generation of ISAR Image. The radar data is initially collected in the down range of the target. As the target rotational component sets in, a series of range profiles are obtained. The rotational motion is determined by analyzing and appropriate corrections applied. The range profile and the cross range data are merged to generate a two dimensional profile. The same has been illustrated in Fig 5.5 which shows the steps in ISAR image generation for a ship.
Fig 5.5 : Steps in ISAR Image Generation
13. The ISAR generated range profile is suitable even for human interpretation. Since lot of data is contained in the radar profile itself, the classification is easier as compared to other methods of NCTR. However as mentioned earlier, the radar hardware and computational power requirements are very high when employing ISAR for NCTR. Considering the technological developments as on date, the ISAR based NCTR techniques have only been used in experiment evaluations and have not been employed in any fighter radar.
CONSTRAINTS IN DATABASE GENERATION AND CLASSIFICATION
1. So far we have seen how NCTR works. In this as mentioned earlier, the classifier can only do its job if a good database is available. So the efficiency of entire process involved in NCTR depends on an accurate and exhaustive database. Without a good database, the entire process is rendered useless. While generating databases of own aircraft may still be possible, the data on targets of potential adversaries and neutrals, who do not wish to provide it, can only be done artificially. A good database should ideally include both civil and military targets for friendly, neutral and potential adversaries and also include variations with aspect angle, stores fits and operational conditions.
2. Factors Affecting Range Profiles. The main factors affecting radar range profiles are aspect angle of the target, range from the radar and aircraft configuration. Since the radar range profile is quite different from optical profile, it may be difficult to appreciate how significantly these factors affect the range profile. An airbus A-320, at certain angle and distance may generate the same radar signature as the F-16 at another angle and distance. And coupled with this fact is that, a wide variety of weapon configuration is possible on the fighter aircraft. Since a wide variety of stores can be carried on the fighter aircraft, database has to be generated for each of these configurations.
3. The aspect of the aircraft in 3D space, as mentioned earlier is both in terms of azimuth as well elevation. Therefore a simple database per aircraft has to consist of a number of aircraft configurations based on weapons it carries. Each of these configurations must correspond to varying aspect both in azimuth and in elevations. And profile data for various ranges for each of these should be available.
4. Now added to be above mentioned complexities are the various sources of variability for a radar range profile. Before proceeding further, it would be prudent to consider the difficulties in obtaining correct profile measurement due to these factors. The main sources of radar range profile variability are due to atmospheric disturbances and the dynamic nature of fighter aircraft. These are enumerated below :-
(a) Measurement Noise. Any radar measurement is subject to measurement noise, which is caused by both thermal noise (Johnson Noise) in the radar receiver and clutter including unwanted radar returns from birds or atmospheric effects.
(b). Translational Range Migration (TRM). Any change in distance between the radar and the aircraft causes scatterers to move within the range profile. Since all scatterers are translated by the same amount, the relative distance between two scatterers does not change. Therefore, the shape of the profile does not change due to TRM, and so the effect of TRM is a translation of the original range profile.
(c) Rotational Range Migration (RRM). If an aircraft rotates over a significant aspect angle (of the order of a few degrees) such that the outermost scatterers move from one range bin to the other, the range profiles collected during this rotation are distorted and causes variation from the predicted profiles.
(d) Speckle. The next source of variability, speckle, is also related to aircraft rotations. Speckle occurs if in a single range bin l two or more distinct scatterers are present. Then, only a slight rotation of the aircraft in aspect azimuth or elevation is enough to cause the sum of the scatter contributions to turn from constructive to destructive interference (or vice versa) within tiny changes of aspect angle. Fluctuations of range profile amplitude due to speckle are multiplicative in nature. The larger the profile amplitude, the larger the variance and hence poses problem for the classifier.
(e) Occlusion. Occlusion occurs when a scatterer is positioned such that it is not observable by the radar. An occluded scatterer does not contribute at all to the measured range profile. Hence it causes a variation in the measured range profile.
The above mentioned sources of variability essentially distort the radar range profile of the target. While some of these may be modeled, most of them cannot be predicted and hence pose problems in classification.
5. Methods of Database Generation. It is imperative to understand that unlike surface objects, the mapping of objects in air is highly complex. NCTR database can be generated using five techniques namely - Aircraft measurements under actual combat conditions, measurement under controlled conditions in turn table, measurements using scaled models, computer generated 3D modeling and electromagnetic modeling of the aircraft. Now, one can easily visualize the effort involved in generating database required for a single aircraft in 3D space. Say for example, to generate database for F-16, the aircraft has to be positioned exactly in airspace, at the exact angle and distance. This database has to be generated for various angles and distances. The various methods for database generation are as follows:-
(a) Aircraft measurements under actual combat conditions. The best and most obvious method is to measure the range profiles and cross-range characteristics of all targets of interest at all appropriate aspect angles under actual combat conditions. This is theoretically possible for available targets, but tends to be very costly and practically not possible.
(b) Measurement Under Controlled Conditions. Aircraft targets can also be measured on turntables under controlled conditions. However, the effects of ground clutter for the turntable measurements result in differences in the measured signatures. As far as azimuth aspect changes are concerned, there are no issues, but elevation aspect changes may not be possible.
(c) Measurements Using Scaled Models. The third method is to use scale models, but here unlike optical domain, merely including scaling factor would not give correct results since radar domain is quite different from visual domain. To cater for changes in scale, the signatures would have to be measured at a correspondingly scaled higher radar frequency in a specially built test facility. This again depends upon knowing the target's physical characteristics in great detail, which also includes the use of representative materials. This technique is clearly dependent upon the available physical modelling skills. Though scale modelling is less costly than making measurements on real targets, the accuracy is compromised. The key issue with the mathematical and scale modelling techniques is whether they agree with the measured data from real targets. This means that it is still necessary to measure real target signatures, so the various techniques can be compared. Checks of real and modelled measurements need to be made at regular intervals to maintain confidence in the agreement between the methods.
(d) Computer Generated 3D Modelling. Simulated range profiles are produced by using radar simulation software with Computer Aided Design (CAD) models of aircraft. Aircraft CAD models represent the geometry of aircraft as a collection of discrete elements. These CAD models are only approximations of the geometry of aircraft. The models comprise of a number of surfaces called facets. The more the facets, the more accurate the model. A sample of Fokker 100 CAD is shown at Fig 6.1. CAD models are that made by Simulation software can only approximate the process of radar scattering.
Fig 6.1 : CAD Model of Fokker 100 Aircraft
(e) Electromagnetic Modelling. The next technique is the electromagnetic modelling of the target. This is computationally intensive and depends upon the physical characteristics of the target being modelled with very high integrity. In addition, it is necessary that the dielectric constant and electrical conductivity of the aircraft materials are defined correctly. A pre-requisite for this is the provision of a detailed physical description of the target's geometry. The degree to which this mathematical model represents the real target's physical features is one of the factors that determine the accuracy of the electromagnetic model of the target. There are two ways in which electromagnetic modelling can be done. They are :-
(i) Finite Difference Time Domain Method. The Finite Difference Time Domain (FDTD) method divides the target into a three dimensional mesh. The target's values of electrical conductivity, dielectric constant and magnetic permeability are assigned to each of the respective points on the grid. The reflectivity at each grid is computed mathematically and the resultant signature determined. This is shown in Figure 6.2.
Fig 6.2 Finite Difference Time Domain Method
(ii) Method of Moments. Another technique used for modelling target signatures is the Method of Moments. It divides the surface of the target into geometrical shapes, such as triangles which are called sub-domains. It uses the Electric Field to determine the surface currents for each of the sub-domains. These currents are then used to calculate the resultant reflected electric field generated. Factors such as whether cavities or discontinuities are present in the target and the wavelength in comparison to target features are important. The material composition of the target is also significant. The amplitude and phase of electric field reflected towards radar are used to calculate target scattering characteristics A diagram of the Method of Moments is shown at Fig 6.3
Fig 6.3 : Method of Moments
6. The database building up for NCTR is a very difficult process. Even making database for own existing aircraft is an elaborate process. The database of adversary can only be estimated which may be far from correct. Despite having a good library, the NCTR predictions can still be incorrect due to the variability of radar range profile in the real world as mentioned earlier. With all these limitations, the chances of NCTR techniques incorrectly identifying a target loom large.
NCTR SYSTEM ISSUES
1. The use of radar returns for NCTR not only requires a good software, but robust hardware also. Making a radar perform for an acceptable level of target recognition performance is a challenge to the radar designer. Past and current radar systems have been specified to detect a target of a certain radar cross-section (or size), under defined environmental conditions, with a particular probability. Radars are required to initiate and maintain tracks to certain measurable accuracies. However, requirements for specifying target recognition functions are expected to be quite different than those for mere detection and ranging.
2. It is necessary to design the radar appropriate to the type of measurement that has to be performed, to provide the type of target signature required. In order to obtain signatures of high integrity, the waveform must be carefully designed and the radar must support the transmission and reception of the signal without distortion. The waveform, the associated signal processing, radar phase noise and dynamic range performance also have to be designed to minimise the effects of clutter. There must also be sufficient energy radiating from the target to ensure that the smallest contributions to the target signature, which are needed for the recognition process, are detected reliably.
3. Most radar systems are designed to support the normal target detection and tracking functions as their primary requirements. Target recognition modes would normally utilise the existing architecture of the radar's design as much as possible, with modifications for target recognition being minimised, on the grounds of cost and complexity. However this may not always be possible (depending on the vintage of radar) and incorporation of NCTR may require a complete radar change. There are design issues associated with sensitivity, dynamic range and calibration, which are generally common to most target recognition functions. These aspects are more stressing for the design of target recognition than for the conventional radar modes. These aspects are enumerated in detail as follows:-
(a) Sensitivity. The conventional radars are employed primarily for detection and tracking of targets. Therefore these radars have a lower range resolution than the target dimensions (the range resolution may be as less as 150 m since, resolution is required only for resolving two different targets). Hence all the reflected energy detected by the radar, is confined to one or at most two range gates. For target recognition purposes, the Radar Cross Section (RCS) of target elements is only a small fraction of the RCS of the whole target. As the RCS of the target elements are much smaller than that of the whole target, the range at which high resolution mode can be utilized, is reduced in comparison to that of target detection range. Now this adversely affects the range at which NCTR is useful. For example if the RCS of the smallest target element of interest for recognition is 20 dB, or one hundredth of the value of the RCS of the whole aircraft, then the range at which the same signal-to-noise ratio is obtained is the fourth root of one hundred, which is about three. This means that the target recognition mode can only be applied at one-third of the range of the target detection mode, if all other parameters are maintained. Increasing sensitivity to improve target detection range would result in increased unwanted clutter. Therefore, the key aspect of designing a high-resolution mode, is the tradeoff between detection of the target and the application of the target recognition mode.
(b) Dynamic Range. Clutter is usually a constraining factor in the design of most radar systems. Due to the presence of clutter, the threshold level of target detection has to be increased in order to eliminate clutter. The levels of reflected returns from ground and sea clutter can be many orders of magnitude larger than the desired signals from targets, which are required to be detected. Now in NCTR systems, the dynamic range requirements are even more stringent than for conventional target detection and tracking modes. Since, the high-resolution radar waveform dissects the target into smaller elements, the clutter returns instead of only competing with target as a whole, now competes with each target element. This requires in the detection signal threshold which manifests as a reduction in detection range.
(c) System Distortion. A critical issue for NCTR based radar systems, which operate over large bandwidths, is the distortion, which occurs due to non-ideal amplitude and phase characteristics of the components comprising the radar system. The actual waveform generated can undergo distortion in the transmitter, antenna and the receiver components. This results in the signal that is effectively transmitted and received by the radar deviating from the signal which was intended to be transmitted. In the absence of any method or compensating for the distortion, the range resolution can be significantly degraded.
4. Therefore the radar intended for use in NCTR application has to be appropriately designed keeping the above mentioned facts in mind. The generation of a nanosecond pulse without compromising on the range still remains an issue at large. Unless there is a way to overcome these system issues, NCTR based radars cannot be optimized.
ALTERNATIVES TO NCTR
1. As we have seen in the earlier chapters, NCTR implementation has lot of issues. We have also seen that though target detection can take place at large ranges, identification by NCTR is possible only a closer ranges. So, is there an alternative to NCTR? Before looking for alternatives, one must be clear as to what is the essential for preventing fratricides. What is essentially required in the modern day combat environment is enhanced 'Situational Awareness' (SA). In a dense BVR combat environment, foes need to be identified reliably and at the earliest to exploit the maximum range of the BVR missiles. Though this paper primarily focuses on NCTR it would be prudent to have a glance at some of other systems available which can enhance SA and thus take on the critical task of identification of friends and foes. The use of these systems is a study in itself and this chapter only aims at listing out the alternatives to NCTR.
2. Airborne Warning And Control System. AWACS is primarily an airborne radar system used for enhancing target detection ranges. It primarily comprises of radar fitted atop a transport aircraft. Since radar is limited by line of sight, any radar fitted on an aircraft would have no physical obstruction in the form of terrain and would have significantly higher detection ranges especially at low levels. Now AWACS is not just radar positioned on top of an aircraft, but a full fledged controlling station. The crew of aircraft controllers inside AWACS would have full situational awareness of the ensuing air combat environment. Inside the aircraft, there are different sections for each controller who can closely control individual fighters. The layout of various controllers workstation inside an AWACS is shown at Fig 8.1. Since the detection ranges of AWACS is significantly larger, the fighter controllers onboard AWACS aircraft would be able to discern hostile aircraft from friendlies, well beyond the range of BVR missiles. The controllers would be continuously monitoring and electronically tagging various aircraft in the combat environment, on their monitor / scope. Therefore, the chances of misidentification would be remote. And since AWACS are moving platforms, they can be easily moved into the required theatre thus giving better situational awareness to all the personnel involved.
Fig 8.1 : Workstations Inside AWACS Aircraft
3. Aerostats. Aerostats are basically moored balloons with platforms on which surveillance radars are housed. The Aerostats work in a similar fashion like that of AWACS, only difference being that the Aerostats are static. Like AWACS, since the radars are located in the air, they do not suffer from line of sight problems. The controlling of the fighter is done by controllers who are housed in cabins on ground. Like AWACS, since the air picture is available for large distances into enemy territory, the controllers would be able to discern hostile aircraft from friendly ones.
Fig 8.2 : Aerostat
4. Operational Data Links. In today battlefield / air combat scenario, a large number of sensors are available. In case these sensors are data linked, they can provide the required degree of situational awareness. The Operational Data Link (ODL) is one such measure which can be implemented to link all available sensors. The data from various sensors can flow from one platform to another. This way, information on target would available well before the target is picked by own aircraft sensor. The ODL along with sensor fusion can provide a composite picture to the operator, in this case the pilot of the fighter aircraft and help in him in establishing friends from foes. A diagram representing the ODL in a combat environment is depicted at Fig 8.3.
Fig 8.3 : Operational Data Link
5. Sensor Fusion. The modern generation aircraft are equipped with a host of sensors. Each of the sensors has a specific purpose and can operate under varying atmospheric conditions with varying capabilities. Some examples of such sensors are the EO/lR pod, SAR Pod, Optical Locator System, radar imagery etc., to name a few. While some of the sensors may have good resolution, others may have enhanced detection range. The term "Sensor Fusion" implies that data from all available sources / sensors onboard the aircraft are taken as an input to form a composite intelligent picture of the unknown target. The main advantage of such a system would be that it would be able to generate target imagery under varying conditions. The inputs for sensor fusion can even be data from external sources via a datalink.
6. Future of aerial combat lies in the use of all available sensors intelligently. Therefore, NCTR is not the only means of de conflicting friends from foes. The above mentioned technologies only a few to name. When they are connected by suitable network, the resultant is an accurate picture of the air combat environment. Most nations today are pursuing the idea of Net Centric Operations. This kind of networking would not only give a composite aerial picture, but also can give a full picture of the land as well as the sea. And this is the path ahead.
1. In today's generation, NCTR is not the only way in which an aircraft can be identified. While NCTR still remains the only non cooperative technique of identification, modern technologies like AWACS, Aerostats, Operational Data Link and Sensor Fusion can be effectively used to generate a more positive and reliable identification of hostile targets. While the exact type of hostile target may not be possible every time, but at least fratricides would positively prevented since AWACS / Aerostats has a much larger picture of the combat theatre. Also with the use of ODL and Networking, data from all possible sources would be integrated to form a composite picture which would enhance the available aerial picture.
2. While NCTR as a concept is very good, however, the hardware and the software that goes into the technique have not been able to mature so far (more than 25 years now). Since the technology is heavily dependent on neural / genetic algorithms, it is likely to take a much longer time. Meanwhile use of AWACS and ODL along with suitable procedures would be the key to effective and economic target identification in a dense air combat scenario.
3. While the idea of "Non Cooperative Target Recognition" without active participation of unidentified aircraft, per se is very good, it would be prudent to delay its implementation till technology on 'Artificial Intelligence' catches up sufficiently.