HK1136876A - Method for reducing multipath propagation effects during the processing of replies in mode "s" - Google Patents

Abstract

The invention relates to a method for processing replies from targets interrogated by a surveillance radar, involving the Mode-S interrogation of all the targets present in the receiving lobe of the radar. The inventive method comprises the following steps: the different targets present in the receiving lobe of the radar are interrogated at least once; the set of Mode-S reply signals received for the lobe are collected after each interrogation; and a processing operation is performed for each target, comprising the detection of replies followed by the detection of errors and, if necessary, the correction of said errors and the extraction of the corresponding blips. The method is characterised in that the aforementioned processing operation involving the detection and determination of signal quality is performed by forming a synthetic message with the set of replies to each interrogation for each target, establishing the value and the quality of each bit of the message and detecting and correcting errors using said synthetic message. The three variables, Σ, Δ and monopulse, of all of the failed replies from the same target are used to form the synthetic message. Said method can be used in a very polluted electromagnetic environment where existing methods are insufficient. The same target may be asked a question again for reasons other than a reply failure.

Description

Method for reducing multipath propagation effect in processing of 'S' mode reply

The present invention relates to a method of reducing the effect of multipath propagation (propagation) when processing "S" mode replies from targets such as aircraft by means of secondary surveillance radar (commonly known as SSR).

In some radar signal echo reception situations, this reception may be disturbed by spurious signals due to strong multipaths in different directions. In these cases, currently known S-mode signal processing operations cannot accurately handle S-mode replies. This can result in unacceptable missed aircraft detections.

It is recalled here that the S-mode principle is to selectively interrogate the craft by using the monopulse information in sequence, substantially (virtual) positively (calculating CRC, i.e. cyclic redundancy code) to "locate" and "decode" the messages transmitted by the airborne transponder in a single interrogation (interrogation) within the lobe (lobe). Algorithms have been developed for this purpose, and all the emphasis has therefore been on the processing of replies.

The S-mode standard (ICAO standard, annex 10) is characterized by:

the main objective is to perform detection and localization of the vehicle in a single interrogation (three-dimensional (3D): position, distance, altitude). Thus, the european aviation security organization (EUROCONTROL), such as STNA in france, has defined a metric: in addition to the detection probability, there is also the number of interrogations per radar antenna rotation and per aircraft. It can therefore be seen that the way in which it is obtained is very important (efficiency index) in addition to the traditional radar performance.

S-mode replies (see simple example in fig. 1) are structurally much longer (64 μ S or 120 μ S) and much denser than SSR replies (21 μ S) and are therefore more sensitive to multipath: the separation between the two S-mode pulses is 500ns or 1 μ S, while the SSR recovery is in the order of 1 μ S, 2.45 μ S, 3.45 μ S. Thus, the S mode is much higher than the SSR mode for the probability that the multipath of a reply interferes with the same reply pulse.

The data exchange between the ground and the onboard systems must be reliable: according to the EUROCONTROL standard, the S-mode site specification requires an error rate of 10-7. For this purpose, the standard has provided an error correction code (24-bit CRC) enabling detection of whether a message is disturbed. The code is designed to be a reply to garbling between conventional two-level replies (21 mus in duration, distributed over 21 mus, with no more than 14 450ns pulses disturbed in each reply, or 8 pulses disturbed on average).

In fact, to meet the security requirements of exchanging data, a maximum of 10 bits with an interval of less than 24 μ S in the S-mode message are corrected. Thus, SSR replies having a greater number of pulses than the average (the code has more than 6 pulses in addition to the possible 12 pulses), can interfere with S-mode messages of more than 10 bits, thus making the S-mode reply uncorrectable (see fig. 2).

Signal processing, i.e. decoding the S-mode reply, also marks the bits (1 bit for 1 μ S) of the message that may be erroneous (of poor quality). According to the principles of the S-mode standard, only these marked bits can be error corrected by an error correction code.

This principle is implemented in a particular propagation environment, and is fully operable in the face of "aliasing" of the interrogation (spurious signals resulting from the interrogation) due to the selectivity of the interrogation. In the presence of strong multipaths, this is by definition left "sticky" to the reply, the principle no longer holds: each reply is analyzed and discarded independently of the other.

In fact, with known methods, for each multipath, the received S-mode reply will systematically self-interfere by:

- "in-line" reflection (on the antenna axis)

And/or "transverse" (slightly off axis from the antenna).

The processing of the S-mode signal is optimized with the processing of the reply for each lobe, so that both the decoding and the correction of the message are performed on a single reply. When failing, a new query is automatically sent and the signal processing module (hereinafter abbreviated TS) uses the new reply again. In the presence of multipath, the failure is repeated. As long as the target is in the reception lobe of the radar and the reply cannot be decoded, a new interrogation is generated. Thus, when multipath is strong, the number of selective interrogations for the disturbed object may end up being equal to the number of interrogations generated by the secondary processing in the non-selective mode. However, since the decoding decision is made on every reply, it fails globally over the entire lobe.

The prior art devices have focused primarily on signal processing (TS) to best perform decoding and quality distribution functions, thereby enhancing the error detection method and error correction efficiency in the S-mode standard by the encoding employed and the required error correction rate.

For each reply, TS makes use of the information available at the output of the receiver connected to the radar antenna, that is to say:

-SUM (SUM) and Difference (DIFFERENCE) path (pathway),

phase information representing the shift (mismatch) of an object in the lobe (this information is referred to as "monopulse" hereinafter)

Three exemplary cases of interference of S-mode reply are described below with reference to fig. 3 to 5:

-is caused by the SSR products,

-caused by a synchronous S-mode reply,

multipath due to S-mode reply.

The correction principles described above cannot withstand certain extreme situations occurring at different radar locations, of which three exemplary (non-limiting) situations may be:

the situation in fig. 3 occurs in northern europe: in an environment that is disturbed by multiple asynchronous secondary replies (called "artifacts"), selective interrogation from one S-mode to the next associated reply may be disturbed by SSR artifact replies each time, which have much higher power than the S-mode replies (in the case where the desired target is located at a large distance 470km, and where the target produces artifacts close to the radar of interest, but replies to long-range radars). The product is asynchronous, which causes errors on different bits of the S-mode reply one after the other. Fig. 3 shows four non-interfering S-mode reply preamble pulses at the beginning of the timeline. What follows are data bits (represented by bit 1 through bit 56 in the figure), the first of which is disturbed by spurious pulses from full-code SSR replies (represented by shading in the figure and having a larger amplitude than the desired bit) that arrive asynchronously with respect to the desired pulse. The spurious pulses are likely to backfill the inter-symbol spaces and may also overlap the forward pulses more or less.

Fig. 4 depicts the second case. Interference is caused by simultaneous parasitic returns resulting from interrogations of the same radar from different aircraft. In the S-mode reply acquisition phase ("All calls") in an environment full of targets such as the northern european air corridor (air corridor), the expected S-mode replies interfere synchronously with each other. The error bit rate depends on the mutual overlap rate of the S-mode replies. From cycle to cycle, the error bits may not always be the same due to collisions (beats) between signals of different frequencies. Thus, in the case of fig. 4, when the radar begins to receive S-mode replies from the first vehicle, replies from the second vehicle arrive starting from the second data bit of the first reply. The four sync pulses are such that the first one is located between the second and third pulses of the first reply, while the remaining three overlap bit 3 to bit 7 of the first reply in a different manner because the respective distances of the sync pulses are different from those of the data pulses. The second recovered data pulse then overlaps the first recovered data pulse.

Fig. 5 is the case of multipath propagation for the same reply. When a TS decodes bits poorly in the presence of strong multipaths, the bits may be distributed anywhere in the reply because multipath, in general, can interfere with all bits of the message. In fact, since the multipaths are repeated identical replies and can be offset in time by a duration of about 3 μ s, the badly decoded bits depend on the message itself and on the collision of the signals in the receiver (direct reply and reply from multipath), distorting the pulses at the output of the receiver. Thus, a TS utilizing received power may incorrectly locate a pulse, incorrectly allocate power to the latter, and thus incorrectly decode a reply. Now, the error detection code principle cannot be used to correct errors that are spaced more than 24 μ s apart. From one S-mode reply to the next, the error bits are different because the distortion of the pulse caused by the collision (waiting) between the forward and reflected waves depends on the pitch difference (linear difference), which varies significantly (10ms) from one cycle to the next.

In the new application market of S-mode radar, the need to detect targets based on a small number of S-mode interrogations is increasing, even exceeding and above the need for good efficiency indicators, because:

the rotational speed of the antenna of the surveillance radar is increased: typically one revolution in 4 seconds for a range of 470 km. Thus, the illumination time on the target is reduced and, therefore, the probability of re-interrogation in case of failure is more limited.

S-mode data processing requires illumination time on the target, thus reducing the possible number of cycles to re-interrogate in case of failure in a previous attempt. Military radars require additional interrogation in specific military modes (1 and 2), thus further reducing the number of cycles in S mode.

Applicant's processing of replies from secondary radars has gone through two main stages of development since 1990, as shown in the block diagram of fig. 6 (at the top and in the middle, respectively), which is an inventive solution (illustrated graphically at the bottom of the figure). The two-level radars implementing the three different extraction techniques all have three main levels, corresponding to the three main steps of the spot (blip) extraction process and represented in the figure by the same columns: a radio frequency processing stage 1, a signal processing stage 2 (hereinafter SP) and a data processing stage 3 (hereinafter DP). Stage 1 is the same for all three implementations. Stage 1 essentially comprises a radar antenna 4, a receiver 5 and an interrogator 6. For stage 2, a number of successive quality detection and determination processing operations are represented graphically, located at the output of the receiver for each of the three methods, arranged one after the other downwards, corresponding to successive interrogations.

Two known techniques are:

"reply to processing and correlator r.p.c.". This is a two-stage extractor developed between 1992 and 1999, to which a number of patents have been applied, which relates only to innovative SSR (secondary surveillance radar) processing, characterized by a powerful identification capability based on analysis of the morphology of the signal received on the sigma channel. The secondary processing principle is based on systematic interrogation of all targets present in the lobe at a frequency of 12 replies (6 for a-mode and 6 for C-mode) for each target in the lobe. The management of its main functions is as follows (see fig. 6):

o space-time management (GST): managed by elements 7 (pacing) and 8(GST), management is very simple since interleaved a-mode and C-mode queries are systematically involved in the sequence.

Signal processing (TS):

■, based on an analysis of the morphology of the signal received on the sigma channel, to detect and decode the SSR reply,

■ determine the quality of the construction based on an analysis of the sigma and delta/sigma information.

Data processing-TD- (9) performs extraction of the light spots based on:

■ is used to detect the number of detections of each or all patterns of the light spot,

■ mode A/mode C code generation by analyzing the codes obtained in each mode in relation to their quality and based on an estimate of each pulse of the code using a flag (flag) that gives the risk of back-aliasing, i.e. that can potentially be decoded correctly.

The boxes in the figure give the complexity of the various main modules described above:

● spatio-temporal management (GST): the complexity of the method is low, and the method is simple,

● Signal processing (TS): the average complexity of the process is calculated,

● data processing (TD): average complexity.

"interrogator and reply process" i.r.p. This was the secondary extractor developed between 1999 and 2005. Several patents have been filed for inventive S-mode signal processing that obtains strong recognition capabilities based on pulse histograms defined by analysis of the morphology of the received signal over sigma and delta, while additional patents have been filed for inventive S-mode data processing in sequences of S-mode selective interrogation. The principle of S-mode processing is based on selective interrogation of each target in a lobe at two frequencies of reply for each target in the lobe:

o space-time management (GST): managed by element 10 (beam cadence in S-mode) and element 11 (GST in S-mode): this management is very complex, since the order is determined by both the operator-selected primary order, for which all selective interrogations are to fall in real time, and the arrangement of the listening windows, associated with the replies expected from the selected targets (50 targets per lobe),

signal processing (TS): is very complicated

● which detect S-mode pulses based on an analysis of the morphology of the signals received on the sigma and delta channels and a pulse histogram,

● which determines the quality of construction of each pulse based on histograms of sigma, delta and delta/sigma pulses,

● which detect replies based on the detected pulses,

●, which decodes the reply message based on the detected pulse and the associated quality of each bit of the message,

● it performs (in 12) calculation of the message error summary indicator independently for each pulse and, if necessary, it corrects the message based on the quality associated with each pulse.

The data processing (TD) simply (in 13) processes the correlation of replies to targets that have been isolated by the SP and performs calculations on the general characteristics of the targets (power, bearing, distance).

The boxes in the figure give the complexity of the various main modules described above:

● spatio-temporal management (GST): the complexity of the process is high and the process is complicated,

● Signal processing (TS): the complexity of the process is high and the process is complicated,

● data processing (TD): low complexity.

Currently, the SP determines a reference value for each detected reply according to three conventional variables (sum, difference and "single pulse"), and the maximum number of samples that coincide with the values for the three said variables, these samples being referred to below as "coincident samples". This also indicates the overall quality of the recovery: the higher the maximum number of consistent samples becomes, the clearer the overall quality becomes (without interference).

The decoding of each bit and the determination of the quality (uncertainty of its value) are related to the position of one or more pulses in the period of the bit, and the values related to three of the variables of the one or more pulses are related to the values of the replies to the three variables.

The block diagram shown in fig. 7 is exemplary of a detailed view of several bits of a message. The difficulty of decoding certain bits when a message is interfered by multiple multipaths is shown:

the first row is a simplified representation of the signal received at the input of the receiver:

■ desired reply signal

■ weaker multipath offset by 500ns

■ second weaker multipath offset by 800ns

The second row shows the signal at the output of the receiver for the sum or difference path, the receiver defining the presence of a pulse and its value with TS. The dashed line depicts the power of the reply calculated over all pulses in the position belonging to the reply.

The bottom of the graph gives the possible results of TS involving the determination of the value of the pulse.

The subject of the invention is a method for processing the "S" mode replies from an aircraft interrogated by radar, which is able to reduce the multipath propagation effects considerably, and which is able to take into account each reply, whatever the parasitic multipath, and to maintain the probability of contract (Pd) of detection, even in poor propagation conditions.

The processing method according to the invention is a method of processing replies from targets interrogated by a surveillance radar which interrogates in S-mode, independently for each target present in the reception lobe of the radar, whereby the different targets present in the reception lobe of the radar SSR are interrogated at least once, all the reply signals received by this lobe for each target are collected after each interrogation, the reply detection processing operation is performed in a signal processing module, and the errors are detected and corrected, and the corresponding light spots are extracted, characterized in that, in the event of decoding failure of the signal processing operation, said signal quality detection and determination processing operation is performed by a data processing module which forms a composite message (value and quality of each bit of the S-mode message) using a stored set of replies for each interrogation of the same targets, the value and quality of each bit of each message is determined and error detection and correction is performed using the composite message.

According to a feature of the invention, for each S-mode reply, namely:

not decoded (TS was not successfully corrected),

-not decodable (too many bits of poor quality in the error message need to be corrected), the signal processing module (TS) transmits:

the values of the replies to the three output variables of the radar receiver (sum, difference and in single pulse), and the maximum number of samples coinciding with said values of these three variables,

-the values of the three variables and the associated number of samples, for each pulse potentially located in each bit of the reply message, thus giving a pulse value quality indicator,

-information indicating a reply detection situation.

According to another feature of the invention, once the signal processing module determines that at least two received replies are not decoded or not decodable, the data processing module reconstructs the replies by performing the following steps:

-determining an estimate for each bit of the per S-mode message based on an analysis of information from potential pulses of all received reply bits, said information for each output variable from the radar receiver: the number of samples that are consistent in the recovery, the recovery quality flag, the value and number of samples per pulse,

decoding each bit of the message, typically based on the value of the estimate and several pulses occurring per bit, obtaining the position of the pulse with the maximum value of the estimate and assigning a quality to each bit,

-detecting errors with a new decoding of the message, and

attempting to correct the bits of the message, if necessary, by exploiting the new quality of each bit, as performed by TS.

The invention thus makes use of all replies received in real time and thus enables reconstruction of a more accurate reply and determination of a better quality of the bits of the message by making use of the instability of the interference of the reply to the reply, by employing the best pulse in each of the sum and difference paths, and by making use of the "one-shot" information to determine the quality. The probability of obtaining a correct "reconstructed" reply directly is thus increased, and the probability of the best possible correction based on CRC can be increased, if necessary, due to the more correct bit quality. The real-time analysis of the received replies, which cannot be corrected independently of each other, and the real-time analysis of a group of received replies at each pulse level, makes it possible to stop interrogating a given target immediately after the correction has ended, and thus to reduce significantly (by a factor of 2 to 3) the number of S-mode selective interrogations, and therefore to be able to process other targets in more lobes.

According to another feature of the invention, for radars operating at rotational speeds higher than about 4s per revolution, the ALL Call (AC) and Roll Call (RC) periods are eliminated to best allocate radar time according to the natural characteristics of the targets present in the radar lobe. S-mode interrogation is then advantageously arranged by taking into account the SSR spot present in the lobe.

The invention will be better understood from a reading of the detailed description of an embodiment, described by way of non-limiting example and with reference to the accompanying drawings, in which:

FIG. 1, as mentioned above, is a time chart showing an exemplary S-mode reply,

FIG. 2, as mentioned above, is a time chart showing an example of interference of an S-mode reply,

FIGS. 3 to 5, as mentioned above, are time charts showing three characteristic examples of different disturbances of the S-mode reply,

FIG. 6 is a block diagram in three parts, the first two of which have been described previously, at the top and in the middle of the figure, relating to the method of the prior art, and the third at the bottom of the figure, diagrammatically representing the main steps of the method of the invention.

FIG. 7 is a time diagram of the part of the S-mode reply that illustrates the problems encountered by the prior art correction method, an

Fig. 8 is a block diagram of an apparatus implementing the invention, corresponding to the "S-mode beam extractor" module part of fig. 6.

The bottom part of fig. 6, which relates to the method of the invention, shows a number of consecutive quality detection and determination processing sequences (the number of the latter being associated with consecutive detection failures) at the signal processing stage 2. Processing operations are performed to detect replies on the sigma and delta channels and, based on the values of the sigma, delta and delta/sigma histograms, to determine the quality of the replies for the sigma and delta channels, delta/sigma information ("single pulses"). The results of these processing operations are sent to the error detection and error correction circuit 15 and, in the event of a failure, to the S-pattern extractor 14 at the same time. The order of interrogation is controlled by a cadence device 16 and an S-mode GST device 17.

The processing operation according to the present invention when multiple replies cannot be decoded due to interference is described below with reference to fig. 8. The signal processing (TS) stage 2 sends the following message information (18.1 to 18.N) for each generated selective reply (up to 5 to 10 per lobe) (the replies 1 to N are identical due to the challenge and therefore always identical as soon as a failure occurs):

number of bits present in the message (consisting of 56 or 112 bits):

● for the sigma channel,

● for the delta channel,

quality of these identical bits:

● for the sigma channel,

● for the delta channel,

● for delta/sigma single pulse information,

overall characteristics of recovery (19.1 to 19. N): sigma, delta/sigma histogram results, values, maximum number of samples that are consistent with a desired valueQuantity, "confusion" (interference) index, and the like.

This information is stored in a shift register that can be cycled back (20.1 to 20.N) to re-read the data without losing the data. For each potential pulse of the message (i.e. for its theoretical position plus or minus the tolerance allowed for that position), all the aforementioned information for each variable (sum, difference and single pulse) of all the received replies is used to determine the estimate (21). The following information is deduced for each digit of the message:

the value of the bit after the pulse position with the maximum value determined by the estimator,

-a quality associated with each bit, typically according to a value assigned by the estimator and the number of pulses (22) occurring per bit.

The procedure then continues conventionally, as is done at the signal processing stage, by detecting errors using a new composite message generated by a set of replies available from the lobes (23). If necessary, the bits of the message can be corrected by using the new quality of each bit (this is also the same function as the TS). Finally, a decoded message (24) of three of the variables is obtained.

To increase the probability of success, the message can be decoded independently as follows:

on two variables and on a single pulse

On two variance differences and a single pulse

On all three variables sum, difference and single pulse

Thus exploiting the non-stationary nature of the interference, for example if the multiple channels are azimuthally offset in the lobe.

If none of the three decodes and corrections are successful, a new query must be made. If the obtained reply is again not correctable by the TS, the TD process described above is repeated completely based on three replies and continues until successful.

The success of the new decoding by the TD enables a reduction in the number of selective interrogations required, and thus the use of duty-cycle limited transmitters, which avoids saturation by the load when the antenna lobe illuminates a large number of aircraft. This also enables more aircraft to be scheduled for subsequent "call by call" periods and the S mode aircraft to be picked up faster (probability of picking up a reply PR of e.g. PR 1) when the radar station is turned on (PR 0.5 at present).

Thus, since the message decoding process operates with all received replies, it provides the best way to exploit the instability of the interference that results in a failure to decode each reply individually.

Thus, the method of the present invention significantly increases the chance of correctly decoding a message because the observed aliasing and multipath are unstable from reply to reply and therefore do not always fail to decode the same bits of the message. The method of the invention thus enables decoding of new reply messages with a predictably better quality than decoding each received reply individually, without generating any additional inquiries.

The process of the present invention is applicable to highly interfered electromagnetic environments for which the prior art methods are not applicable: the same target may be asked again for reasons other than reply failure.

According to a feature of the invention, if the real-time processing provides sufficient power, it can be applied as and when required in the lobe, thus enabling only the necessary number of selective interrogations to be made when the aforementioned laterally exploited replies are able to produce correct or correctable information. This then enables a much larger number of different targets to be processed. Furthermore, the received replies can be utilized in non-real time at the end of the lobe without increasing the computational power, thus having an additional opportunity to decode the message if one of the replies from the lobe is not decoded.

This message decoding processing operation makes use of all the replies received, so as to make the best use of the instability of the disturbances that cause the decoding of each reply to fail individually, such as "confounding" disturbances caused by SSR asynchronous replies and by the multipaths from reply to reply that are not consistent due to the displacement of the vehicle over distance (the cadence difference changes, so that the collision between the forward and reflected waves generates different signals, irrespective of the probability of change of the reflectors).

Furthermore, the initial application of the S-mode protocol to IFF military radars showed that the conventional sequence proposed by FAA, EUROCONTROL or STANAG has not been applicable, where military radars rotate faster (1, 2 or 4 seconds/rev) and need to maintain compatibility with the conventional SIF mode (mode 1 and mode 2) and with the demand from civilian radars capable of "data connection" at high rotational speeds (4 sec/rev). In fact, for high-speed radars, the strict order based on "full call" (AC) and "roll call" (RC) periods limits the number of SIF queries in the lobe, as well as the periods allocated for selective S-mode transmission for data links.

Since the IFF antenna is not an electronically scanned antenna, the illumination time of the target is directly related to the rotation speed of the radar. The strictly allocated time periods for the different protocols (AC for SIF and RC for S mode) make it impossible for the radar to adapt to the natural characteristics and quality of SSR/SIF or S mode targets appearing in the lobe.

According to an advantageous application of the invention, it is proposed that the AC and RC periods are not strictly allocated anymore. Thus, by mixing the two protocols, the radar time can be optimized according to the targets (SIF or S mode) present in the lobe and according to the type of interrogation required (SIF or S mode data link). To avoid confusion between the S-mode query and the SIF reply, or between the S-mode reply and the SIF reply, the distribution of the S-mode query takes into account the predetermined locations of the SIF replies in a similar manner to the distribution made between the S-mode replies themselves in the RC period.

However, in areas full of space vehicles, the probability of an S-mode reply being interfered (synchronous or asynchronous) by one or more SSR replies is quite high. It is generally not possible in practice to insert an S-mode process (query or reply) in an SSR synchronization process (query or reply) because SSR synchronization processes depend on the distance and azimuth distribution of the spacecraft, can form locally dense reply blocks, making insertion impossible, and essentially overlapping S-mode replies with synchronized SSR replies.

In this case, the present invention provides a method to exploit the non-stability of these overlapping cases and thus decode the S-mode reply, thus enabling the elimination of AC and RC periods, thereby enabling optimization of radar time according to the natural characteristics of the target.

Claims (10)

1. A method of processing replies from targets interrogated by a surveillance radar which interrogates in S mode, independently for each target present in the reception lobe of the radar, whereby the different targets present in the reception lobe of the radar SSR are interrogated at least once, all the reply signals received by this lobe for each target are collected after each interrogation, reply detection processing operations are carried out in a signal processing module (TS) and errors are detected and corrected, and the corresponding light points are extracted,

characterized in that said signal quality detection and determination processing operations are performed by a data processing module (TD) which forms a composite message using a stored set of replies for each challenge for the same target, determines the value and quality of each bit of each message, and performs error detection and correction using the composite message, in case of a decoding failure of the signal processing operation.

2. The method of claim 1 wherein the composite message includes a value and quality of each bit of the S-mode message.

3. A method according to claim 1 or 2, characterized in that for each S-mode reply that is not decoded after the signal processing operation has not successfully performed a correction attempt, or that is not decodable due to the excessive number of poor quality bits in the error message for which a correction is attempted, the signal processing module (TS) transmits:

the values of the replies to the three output variables of the radar receiver (sum, difference and monopulse), and the maximum number of samples coinciding with said values of the three variables,

-the values of the three variables and the number of associated samples, which values and number are for each pulse potentially located in each bit of the reply message, thus giving a pulse value quality indicator,

-information indicating a reply detection situation.

4. Method according to any of the preceding claims, wherein once the signal processing module determines that at least two received replies are not decoded or not decodable, the data processing module (TD) reconstructs the replies by performing the following steps:

-determining an estimate for each bit of the S-mode message based on analyzing information from potential pulses of all received reply bits, said information for each output variable from the radar receiver: the number of coincident samples in the reply, the reply quality flag, the value and number of samples per pulse, each bit of the message is typically decoded based on the value of the estimate and the number of pulses that occur per bit, the position of the pulse with the maximum value of the estimate is obtained, and a quality is assigned to each bit,

-detecting errors with a new decoding of the message.

5. Method according to any of the preceding claims, characterized in that when it is possible and necessary to correct a message, the data processing module tries to correct the bits of the message with the new quality of each bit.

6. Method according to any of claims 3 to 5, characterized in that to increase the probability of success, the decoding of a message is performed independently as follows:

on two variables and on a single pulse,

-on both the variance difference and the single pulse,

on all three variables sum, difference and single pulse

To best utilize the characteristics of the interference (amplitude, offset, lack of temporal stability, etc.) since one of three attempts at the S-mode message is considered correctly decoded.

7. A method according to any one of the preceding claims, characterised by reducing the number of selective interrogations by utilising those already received in real time in each lobe when sufficient real time computational power is available.

8. A method according to any one of claims 1 to 6, characterised by utilising the received replies in non-real time at the end of a lobe to enable the message to be decoded when one of the replies from that lobe is not decoded.

9. Method according to any of the preceding claims, characterized in that for a radar operating at a rotational speed higher than about 4s per revolution, the full call (AC) and calling-by-calling (RC) periods are eliminated, to best allocate radar time according to the natural characteristics of the targets present in the radar lobe.

10. Method according to claim 9, characterized in that the S-mode interrogation is arranged by considering a set of SSR/SIF spots present in the lobe.