Non-linear radar: the “NR” concept.

Nonlinear radar: the “NR” concept.

Dmitry Vladimirovich Semenov, postgraduate student at Bauman Moscow State Technical University
Dmitry Viktorovich Tkachev

NONLINEAR RADAR: THE “NR” CONCEPT.

Over the last 6-7 years, Russia has experienced a real boom in the development of non-linear radars (NRL). During this time, more than ten (!) different models have been presented to the consumer: Cyclone, Oktava, Lux, Onega 2, NR900E, NR900M, NR900N, Onega 3M, Envis, Perekhod, Rodnik-2, Rodnik 23, Ob, Ob 2S… The reason for such diversity is in the unique consumer qualities of NRL — it is a very effective, universal and easy-to-use search equipment. NRL is a kind of semiconductor device indicator: it allows you to detect illegally placed electronic devices of any purpose, both working and «sleeping».The high guarantee of detecting foreign electronic objects using location methods has led to the creation of a wide range of devices designed to survey premises in various situations. The proposed products have different technical characteristics of their components, which leads to differences in their final efficiency in terms of detection ability.

Specialists in the equipment survey of premises need to know the physics of the phenomena that determine the efficiency of nonlinear location, as well as the features of the main operating modes of NRL.

SEMICONDUCTORS: REAL AND FALSE.

Unlike classical (linear), in nonlinear radar information about the detected object is determined by its ability to spectrally transform the probing signal and re-reflect it at the harmonics of the probing frequency. These phenomena are possible if the object contains elements with nonlinear volt-ampere characteristics (VAC). Such elements, by the nature of their origin, can be conditionally divided into “real” and “false”.

By “real” we mean semiconductor devices of artificial origin containing a p-n junction (for example, diodes, transistors, etc.). For brevity, we will call them “real” semiconductors. It is “real” semiconductors that are the object of interest in the search.

The I-V characteristic of a p-n junction is described by an exponential function [1]. If we calculate the multiple spectral components of the current through an element with such a I-V characteristic depending on the amplitude Um of the applied sinusoidal voltage, we can show that at Um values ​​less than several tens of millivolts they will be represented mainly by the second harmonic. As Um increases, the amplitudes of higher harmonics will rapidly increase.

The paper [2] presents the results of experimental studies of nonlinear scattering of the probing signal by “real semiconductors”. It is shown that the dependence of the second harmonic power in the receiving antenna of the NRL Pr2 on the power flux density of the probing signal – Пt, at a fixed distance R, has the form shown in Fig. 1.

Fig. 1

The nature of the experimental dependence Рr2(Пt)

This graph has three regions:

1 – the region of weak interaction Рr2 ~ Пt2,
2 – the region of strong interaction Pr ~ Пt,
3 – the saturation region Pr2=const.

This dependence is a characteristic of the efficiency of nonlinear scattering. Its appearance for different “real semiconductors” is usually preserved, but the numerical values ​​are different in each case.

If we trace the physical processes occurring during the operation of the NRL, then for the region of weak interaction and the case of free space we can obtain the following expression establishing the relationship between the main parameters of the NRL, the distance and the power in the receiving antenna at the second harmonic frequency

where Pt is the power of the probing signal at the input of the transmitting antenna,
Gt – gain of the transmitting antenna,
Seff. – effective area of ​​the receiving antenna of the NRL.

The efficiency of detecting electronic devices by the presence of nonlinear elements is determined not only by the technical parameters of the equipment, but also by the properties of the object being examined — ceilings, walls, furniture, etc. The practice of using NRL has shown that responses to the harmonics of the radiation signal are created not only by special semiconductor devices, but also by various metal elements of structures that are in contact with each other.

The most typical structures that create interference are the metal frame and reinforcement of reinforced concrete buildings, metal structures of window and door frames, reinforcement of suspended ceilings, etc. The resulting nonlinear elements are detected by NLR similarly to “real” semiconductors.

Nonlinear elements formed as a result of mechanical contact of metal surfaces through a thin oxide film, by analogy, will be conventionally called “false (contact, corrosion) semiconductors. The properties of “false semiconductors” that interest us are most fully described in the work [3].

With a sufficiently small thickness of the oxide film (less than tens of A), the main mechanism of carrier transfer through the contact is the tunnel effect. At low voltages (less than 1 V) and identical metals, the contact current-voltage characteristic can be approximated by a third-degree polynomial. At a contact voltage of more than ~ 1.5 V, the current-voltage characteristic becomes steeper, and with a further increase, it becomes unstable and, in most cases, irreversible breakdown of the contact occurs. An essential feature of the current-voltage characteristic of contact semiconductors is its instability under mechanical action (change in pressure on the contact).

The specified approximation of the I-V characteristic means that the third harmonic will prevail in the spectrum of the reflected signal. The nature of the dependence of the scattered signal power (at the third harmonic) at the input of the NRL receiver — Pr3 at a fixed distance is similar to that shown in Fig. 1, with the difference that in the weak interaction region the power flux density exponent Πt is equal to three: Pr3 ~ Πt3.

• With nonlinear scattering of the probing signal, the dependence of the harmonic signal power at the input of the NRL receiver — Prn (at R = const) on the power flux density of the probing signal Πt has the form shown in Fig. 1.
• In the weak interaction region Prn ~ Πtn (n is the harmonic number).
• The current-voltage characteristics of “false” semiconductors have a significant dependence on the contact pressure force, which determines the mechanical instability of the scattering characteristics.

DETECTION OF NONLINEAR OBJECTS.

Sometimes we come across the opinion that the detection range is not an important characteristic of the NRL. The search is still conducted at a short distance (tens of centimeters) and increasing it does not make sense. This is a false argument. In fact, it should be borne in mind that search objects have a significant spread in scattering efficiency, which in some cases can be extremely small. Comparison of different devices by the maximum detection range of the same simulator is nothing more than the simplest assessment of the ability of a particular NRL to detect a “target” with a lower or higher reflectivity. Under real-world conditions and a priori uncertainty of scattering characteristics, a shorter range will result in missing an object with low nonlinear conversion efficiency. Therefore, we believe that the detection range for a given signal-to-noise ratio is the main characteristic of the NRL, determined by the combination of its technical parameters and the parameters of the nonlinear object.

Let us consider the method of comparative evaluation of the detection capability of NRLs with different characteristics. We will assume that the same nonlinear object was detected by two NRLs at the same distance. In this case, the signal-to-noise ratio at the receiver output of the first NRL is q1, and of the second — q2. Using expression (1), we can write:

where Gr1, Gr2 are the gain factors of the receiving antennas.

All NRLs can be divided into two large groups, differing in the law of modulation of the probing signal. The first group includes the so-called “pulse” NRLs (NR900E, Cyclone, Oktava, Onega…), the probing signal of which is a sequence of short (units of microseconds) radio pulses with a repetition rate of hundreds of Hz (duty cycle Qn? 1000) and peak power Ptimp. – tens … hundreds of watts.

The second group consists of the so-called “continuous” NRLs (“Ob”, Rodnik”, all NRLs of foreign manufacture), the probing signal of which is either not modulated in amplitude, or has a pulse modulation with a duty cycle of Qn? 10. In some NRLs of foreign manufacture (for example, ORION), frequency modulation of the probing signal with a deviation of ~ 1 kHz is used. This technique allows for a slight increase in the isolation between the receiver and the transmitter. The power of “continuous” NRLs ranges from tens of mW to units of W.

Using expression (2), we will consider a specific example, comparing “continuous” and “pulse” NRL. Joint tests of the NR900E and Rodnik 23 devices showed that these two devices, which are completely different in their characteristics, provide approximately the same detection range. This result was obtained for ten different simulators. The criterion for measuring the range corresponded to the lighting of the same number of segments on the bar indicators of the devices q1=q2, corresponding to the threshold powers Ppor.1 and Ppor.2. For this case, (2) can be rewritten:

Table 1 presents the NRL parameters, with the transmitter powers taking into account the attenuation in the cables, and the gain values ​​of the NR900E antennas with circular polarization are recalculated relative to a linear isotropic source, i.e. reduced by 3 dB.

Table 1.

The numerical calculation gives a coincidence of equality (3) with an accuracy of no worse than 3 dB, which fits into the overall error of the experiment and is quite sufficient for engineering assessments.

As follows from the given dependencies and calculations, the decrease of the transmitter power by N times required the increase of the receiver sensitivity by N 2 times — this is the specificity of nonlinear radar. This means that the influence of external (ether) and internal (hardware) interference on the NRL with lower power increased proportionally to N 2.

It should be explained that the lower sensitivity of the receivers of «pulse» NRL is explained by the need to expand their bandwidth, matched with the spectrum of the envelope of radio pulses. The noise coefficients of the receivers — Fш «pulse» and continuous» NRL with well-chosen parameters, providing both high potential detection ability and the ability to work in «real» conditions, due to the presence of fairly wide power and sensitivity adjustments. We consider the prevailing opinion that «pulse» NRL, unlike continuous» ones, are supposedly suitable only for work in absolutely empty — «clean» rooms, to be erroneous, since it is devoid of any physical basis, in our opinion.

According to the comparative table given in [4], the main technical characteristics of “pulse” NRLs are approximately the same, and therefore, the detection characteristics are also approximately the same. The difference in their consumer qualities, in our opinion, is determined by ergonomic parameters: ease of use, ease of control, clarity of indication, availability of additional modes, etc. The technical parameters of the Rodnik 23 product are the best among “continuous” NRLs and their values ​​are close to the maximum feasible for this class of equipment.

• One of the main characteristics of NRLs is the detection range, determined mainly by the power of the probing signal (the product Pt Gt). “Pulse NRLs are potentially superior to continuous NRLs in these parameters.”
• An attempt to compensate for a decrease in the probing signal power by N times, with the same detection range, will require an increase in the sensitivity of the NRL receiver by N 2 times, while the noise immunity will deteriorate proportionally.

SELECTION OF “FALSE” SEMICONDUCTORS.

The methods for selecting “false semiconductors” are based mainly on two physical principles:

• the difference in the levels of the second and third harmonics scattered by “real” and “false” semiconductors (for real semiconductors, the level of the second harmonic is ~ 20 dB higher than the level of the third for “false” ones, vice versa);
• instability of the current-voltage characteristics of “false semiconductors” under mechanical action.

The “fading” effect, actively advertised in [5], is, in our opinion, simply a consequence of the first principle.

In the paper [6] the principle of selection of “true” and “false” semiconductors by the difference in harmonics is criticized and an attempt is made to prove the uselessness of the third harmonic receiver in general. We share the author’s opinion that sometimes the “difference in harmonics” rule does not work or works with precision “to the contrary”. Such a situation corresponds to an extremely high value of the probability of a false alarm and requires real art from the operator, and from the equipment — to provide him with maximum information. When a “false” semiconductor is mechanically affected, due to the instability of its current-voltage characteristic, amplitude modulation of the harmonics of the probing signal reflected from it occurs. This is perceived by the ear as a crunch with a rhythm corresponding to the rhythm of the effect. We consider the ability to receive at two harmonics at once to be an important quality of the NRL, at least because receiving the third harmonic allows one to more clearly hear the characteristic modulation.In some NRL receivers, for example in Onega”, the signal from the detector output is pre-measured, in principle this is advantageous — accumulation can be used, and then for sound indication it is formed using an auxiliary generator with an amplitude proportional to the measured value. In our opinion, this is an unsuccessful solution — the operator is deprived of the opportunity to listen to the modulation. This opportunity is also unavailable to the operator of the NRL Cyclone due to the threshold nature of its output signal.

A more complex task is to select a “true” semiconductor against the background of “false” ones. There are several methods for solving this problem, but all of them assume sufficient linearity of the NRL receiver, ensuring the ability to recognize signals from different sources. In our opinion, the dynamics of the NRL receiver should be at least 20 dB, which, unfortunately, is not implemented in all models. For example, the ORION NRL will miss a “target” located near a “false” object if it operates in the continuous signal emission and FM detection mode, since the FM detector assumes amplitude limitation.

In a number of practical cases, powerful pulsed NRLs provide the ability to very effectively “fight” against “false semiconductors”. For small distances and point contacts (e.g., reinforcement, twisted wire), the energy of the emitted signal is sufficient to break through the oxide film and destroy the nonlinear contact. Continuous” NRLs do not provide such an opportunity.

• When selecting “false” semiconductors, the best results are provided by powerful pulsed NRLs, which provide the ability to simultaneously receive the second and third harmonics and visually display the levels.
• The design of the receiving path should allow the operator to listen to the characteristic modulation that occurs when the false semiconductor is mechanically affected. The dynamic range of the receiving path should be 20…30 dB or more.

ENVELOPE SEPARATION MODE.

This mode allows the NRL operator to detect electronic devices based on the modulation of the second harmonic by signals characteristic of this device.

The point is that if, for example, a low-frequency signal current a(t) due to the circuit operation flows through a diode that is part of the circuit, then the second harmonic of the probing signal scattered by this diode will also be modulated according to the law a(t).

This mode is implemented most effectively in continuous-action NRL. In pulsed NRL, the envelope extraction mode requires changing the modulation parameters of the probing signal, which leads to complication and increased cost of the transmitter. Therefore, simple models of pulsed NRL (for example, «Cyclone») do not have this mode.

In more advanced pulse NRLs (e.g., “Oktava”, NR900E) this mode is called 20 K”, which is explained by the increase in the pulse repetition rate of the probing signal to several kilohertz (the pulse repetition duty cycle in this mode decreases Q20» 100, and the peak power also decreases). The reflected signal is, as it were, samples from the modulating function a(t), made with the repetition rate of the probing pulses. In the NRL receiver, they pass through the low-pass filter, and the form of a(t) is restored. Compared to continuous NRLs, the signal-to-noise ratio loss (at the same level of the second harmonic at the receiver input) is at least Q times.

The envelope extraction mode expands the capabilities of the NRL, and in addition to the above-mentioned purpose, it is often used to select “false semiconductors.”

The ORION NRL also has a “20 K” mode (?), which is quite surprising for a “continuous” locator: switching to pulsed radiation leads to a significant increase in noise. It seems that the “20 K” mode was “spied” on one of the Russian products, not fully understood, but implemented just in case.

• The envelope extraction mode allows detecting and identifying weakly shielded electronic devices in the operating state. In continuous NRL, this mode has significantly greater efficiency than in pulsed ones.

CONCLUSION.

In this article we have outlined our vision of the physical principles underlying the operation of the NRL, the features of its operating modes, and the requirements for its main parameters. This material is a kind of concept — the concept of nonlinear radars of the «NR» series, the concept of the «IKMC-1» company on whose behalf we speak.

To the question: which NRL to choose, which is better? — we will answer: of course, «pulse», as it has a number of fundamental advantages over «continuous», making it more universal both in terms of search objects and application conditions, which we tried to talk about in this article.

Questions can be sent to stt@detektor.ru.

LISTING:

1. I.P. Stepanenko. Fundamentals of Microelectronics. M. “Soviet Radio”, 1980.
2. A.A.Gorbachev et al. Radio Engineering and Electronics. 1996. Vol. 41, No. 5, pp. 558-562.
3. V.B.Shteynshleger. Advances in Physical Sciences. 1984. Vol. 142, issue 1, pp. 131-145.
4. A.A.Khorev. Methods and Means of Information Protection. M.: Ministry of Defense of the Russian Federation, 1998. – 316 p.
5. T.Jones. Review of Nonlinear Radar Technology. Special Equipment. No. 3, 1999.
6. N.S.Vernigorov. The Principle of Detecting Objects with a Nonlinear Radar. Confidential, 5, 1998.