Communication Receiver

This invention relates to burst communication receivers and more particularly to communication receivers which operate asynchronously on a plurality of messages which are received nearly simultaneously from multiple transmitters.

Receivers used in burst communications systems where it is necessary to exchange data between communicators (such as an interrogator and a transponder) may be subjected to interference from enemy spoofing, jamming, exploitation and deception. Receivers which have been used in the past for IFF purposes encounter difficulties when attempting to decipher multiple, overlapping messages from an interrogator or transponder. Additionally, an enemy might attempt to generate some or all possible interrogation signals (or transponder replies) to deceive or exploit a friendly IFF system. Prior art IFF systems were easy to jam and required that the transmitter power be very large compared to the jamming signals applied to the receiver. Receivers operating in an IFF system (either in a transponder or an interrogator) must be able to operate simultaneously on relatively short messages, that is, detect the message content as it arrives without the benefit of knowing when it is going to arrive.

Therefore it is an object of the present invention to provide a receiver which may operate asynchronously on relatively short messages.

Another object of the present invention is to provide a receiver which is able to receive data substantially simultaneously from multiple communicators such that all messages are received.

Another object of the present invention is to provide a receiver system which limits the communication sequence to only the communicators (interrogators and transponders) of interest, thereby reducing the degradation of the system through mutual interference.

Another object of the invention is to provide a receiver capable of operating in hostile (enemy) environments where large numbers of communicators are active.

Another object of the present invention is to provide a receiver system which incorporates a unique means for determining a valid message (interrogation or reply) thereby preventing deception and exploitation of the system.

Another object of the present invention is to provide a receiver system which allows messages to be of less power than a jamming signal yet still detect such valid messages.

Another object of the present invention is to detect valid interrogation and reply messages even when the jamming power levels are equal to or higher than the power levels of the interrogation and reply messages.

Another object of the present invention is to provide an IFF system which has improved beam sharpening capability to reduce the number of overlapping replies to an interrogator.

A still further object of the present invention is to provide a receiver system which is reduced in cost and increased in terms of reliability.

Other objects and features of the invention will become more readily understood from the following detailed description and appended claims when read in conjunction with the accompanying drawings, in which like reference numerals designate like parts throughout the figures thereof, and in which:

FIG. 1 is an identification-friend-or-foe system in which the receiver according to the present invention can be used.

FIG. 2 is a block diagram of the portion of the communication receiver according to the present invention.

FIG. 3 is an illustration of a typical transmitted interrogator word or message and a control signal.

FIG. 4 is the timing sequence illustrating the theory of operation of a 2T convolver.

FIG. 5 illustrates waveforms at various points in the receiver system of FIG. 2.

FIG. 6 illustrates the sum and difference pattern of a typical transmitter antenna.

FIG. 7 illustrates a more detailed block diagram of the preamble matched filter illustrated in FIG. 2.

FIGS. 8 and 9 represent waveforms at various points in the receiver of FIG. 2.

Referring now to FIG. 1, there is illustrated an identification-friend-or-foe (IFF) system 10 which incorporates the receiver according to the present invention. An airplane 12 includes an interrogator 14; the interrogator 14 includes both a transmitter 16, a receiver 18 and a common antenna 20. The pilot in airplane 12 can actuate interrogator 14 to transmit a multibit coded message signal 22 to other aircraft 24, 26 and 28 within a predetermined location within the antenna pattern of aircraft 12. Airplanes 24, 26 and 28 each have transponders 30, 32 and 34 included therein. Transponder 30 includes a transmitter 36, a receiver 38 and an antenna 40 while transponder 32 also includes a transmitter 42, receiver 44, and antenna 46 and transponder 34 includes transmitter 48, receiver 50 and antenna 52. Assuming that each of the transponders 30-34 are located within the antenna pattern of antenna 20 in aircraft 12, then transponders 30-34 must detect and decode in receivers 38, 44 and 50 respectively, the signal 22 and reply with a coded message from transmitters 36, 52 and 48 that the interrogator receiver 18 must detect and decode. It is common to have multiple aircrafts, such as additional aircraft 54 in the area, said aircraft having on board an interrogator 56 including a transmitter 58, receiver 60 and antenna 62. This interrogator 56 may also be activated to transmit a coded message 64 which may also be detected by transponders 30-34. Receivers 38, 44 and 50 are operating in environments where messages will be arriving from several other system users (such as interrogators 16 and 56) nearly simultaneously. In addition, none of the receivers will know accurately when messages will arrive although each will know the coding sequences to be expected. Also, when used in an IFF system 10, hostile forces or the enemy will attempt to jam system 10 by repeating all or part of the coded message in an attempt bo make the transponders 30-34 reply to those false messages and thereby not be able to reply to a valid message received during this time interval.

Although the receiver according to the present invention will be described with respect to an IFF airborne system 10, it will be recognized by those skilled in the art that other applications as well as ground based systems may incorporate the receiver to be described hereinbelow. The receiver to be described herein can be used both in the interrogator as well as the transponder portion of the system.

In other words, the receiver according to the present invention is used in a burst communication system where it is necessary to exchange data between communicators (the interrogators and transponders) where speed and security from enemy spoofing, jamming and exploitation are primary concerns. The probability of jamming suggests that spread spectrum techniques be used to achieve signal versus jamming improvements with respect to data detection and synchronization. The spread spectrum technique employed is modulation of a CW carrier. As mentioned earlier, it is necessary that both the interrogator and transponder generating the waveform know the proper modulation sequence or code so that the signal versus jamming improvements can be obtained in the receiver. To prevent an enemy from exploiting the IFF system by also knowing the proper code from past observations, the code is selectively changed at short intervals (known as code validity intervals).

In IFF system 10 illustrated in FIG. 1, because of the large number of communicators utilizing such system, the transponders and interrogators must be able to communicate rapidly which impacts the receivers included therein in the following manner. Firstly, the receivers, such as the receiver 18 and 38, must be able to operate synchronously on relatively short messages, that is, the receivers must be capable of detecting the symbols (or codes) as they arrive without the benefit of knowing when they are going to arrive or the benefit of long code search times. This type of requirement exists when the sender is interested in communicating with many receivers in very short time periods. Secondly, the receivers must be able to receive data almost simultaneously from multiple senders so that all messages are detected. Thirdly, the communication sequence must be limited to only the senders and receivers of interest to each party since without this limitation, senders will receive return communications from many receivers which will block the reception of all messages through mutual interference.

FIG. 2 illustrates a receiver 70 according to the present invention. For purposes of discussion, it will be assumed that this receiver is designed for the transponder in IFF system 10. The transponder antenna and r.f. to i.f. down conversion circuitry is not shown in FIG. 2. A multibit coded message r.f. signal is input to receiver system 70 and downconverted to an i.f. signal 72 compatible with SAW convolver operation. A first convolver 74 and a second convolver 76 have each of their input ports 78 and 80 respectively connected to receive such i.f. signal 72. The reference port 82 of convolver 74 is connected to receive a reference A signal while reference port 84 is connected to receive reference B signal. The general configuration of the convolver is shown in the insert 86. The output port 88 of convolver 74 and the output port 90 of convolver 76 are connected to video processing circuitry means 92 and 94, respectively. Video processing circuitry means converts the i.f. correlated input signal from output ports 88 and 90 to a video correlated signal. Video processing circuit means 92 may be comprised of a log detector 96 coupled by conductor 98 to peak detector and stretch circuit 100. The output of such peak detector and stretch circuit 100 is coupled by way of conductor 102 to sample-and-hold circuit 104.

In similar manner, the output present on output port 90 is full correlated i.f. bit amplitude signal which is coupled by way of conductor 106 as an input to log detector 108. The output of log detector 108 is, in turn, coupled by conductor 110 to peak detector and stretch circuit 112. The output of peak detector and stretch circuit 112 is coupled by way of conductor 114 to sample-and-hold circuit 116. The outputs from sample-and-hold circuits 104 and 116 are coupled by way of conductors 118 and 120 to analog/digital converters 122 and 124, respectively.

It will be recognized by those skilled in the art that more than 2 convolvers can be used according to the present invention (e.g., 4, 8, etc.) without departing from the spirit and scope of the invention. With such an increase of convolvers, for example 4, the symbol length T would convey 2 data bits of information; with 8 convolvers, 1 symbol would convey 3 data bits of information, etc. The advantage of increasing the number of convolvers used to decrease the time necessary to transmit the information (with however, the attendant disadvantage of increased hardware). For purposes of the discussion herein, it is assumed that 2 convolvers are used so that symbol and bit are used interchangably.

The outputs 126 and 128 from A/D converters 122 and 124 are divided such that one path, namely conductors 126' and 128' are connected to synchronization or preamble detection circuitry 130 and sidelobe suppression circuitry 132. If the message includes time stagger information (rather than preamble information), then the synchronization circuit 130 may be a shift register which is one message length long with taps coming off the register going to a m/n detector where n is less than or equal to the number of bits in the message and m is a threshold level less than n. The output from the m/n detector inputs memory means 162 (to be described later). It will be assumed therein that the incoming message includes preamble information for purposes of describing the system operation. "Preamble" information as used herein means synchronization information that either precedes or follows the data information. Preamble detection circuitry 130 and sidelobe suppression circuitry 132 are, in the preferred embodiment, matched filters that are designed to detect particular symbol sequences in the receiver A and B channels. One such sequence is the preamble signal that preceeds the message data and establishes message time of arrival for the receiver 70. The other sequence is the sidelobe suppression or control signal that is transmitted substantially simultaneously with the preamble signal. The relative magnitude between the preamble detection circuitry output and the sidelobe suppression circuitry output is a measure of whether the message was received by the transponder receiver 70 in the main bean of an interrogator (such as interrogator 14 in FIG. 1) or in the sidelobe pattern. A voltage from the SLS margin circuit 134 is added to the output of the sidelobe suppression circuit 132 which is compared via conductor 138 to the preamble output on conductor 140. This comparison takes place in digital comparator 142. If the preamble output on conductor 140 is larger than the output on conductor 138, then receiver 70 will continue to process the message. The voltage added to the output of sidelobe suppression circuit 132 on conductor 136 in SLS margin circuit 134 can be varied to sharpen or broaden the response of receiver 70. If the preamble signal on conductor 140 is larger than the sidelobe suppression signal (on conductor 138), then the preamble signal is compared to the average signal output on conductor 144 generated in average value circuit 146. If the preamble output on conductor 148 is larger than the average output of the preamble generated on conductor 144 by a certain voltage magnitude, namely the magnitude of voltage generated in threshold circuit 150, then receiver 70 has detected a valid preamble within the desired response region and a timing or FIFO clock circuit 154 initiates a clock or timing signal. This procedure is repeated for each incoming preamble and if a valid preamble is detected, the first in first out (FIFO) clock circuit 154 will actuate different timing signals. If N timing signals are required, then clock generator 154 may generate N different timing signals having N different phases.

The other path from channel A and channel B on conductors 126 and 128 are coupled to the input of a digital comparator 156 which generates an output designated binary "1" when the channel A signal from A/D converter 122 exceeds the magnitude of the signal from channel B A/D converter 124 on conductor 128. When the channel B signal from A/D converter 124 exceeds the magnitude of the signal from channel A A/D converter 122, the digital comparator 156 generates an output designated as binary "0". The output 158 from comparator 156 is connected to a data bus 160 which sequentially stores the data from line 158 in data bus 160. When the receiver 70 detects a valid preamble and initiates a clock signal 1 from FIFO clock circuit 154, this clock signal 1 will enable a storage means 162 comprised of a plurality of shift registers connected in parallel to data bus 160. The FIFO clock output timing signals 1, 2 . . . N are connected as clock inputs to the plurality of shift registers FIFO 1, FIFO 2 . . . FIFO N, respectively, in storage means 162. As mentioned previously, when the receiver has detected a valid preamble and a clock is started, this enables the remaining data portion of the coded signal to be clocked into one of the parallel bank of shift registers labeled FIFO 1, . . . FIFO N. As subsequent preambles are detected, different clock signals from FIFO clock circuit 154 are actuated thereby allowing the remaining portion of the data signal to be stored in one of the parallel shift registers FIFO for later processing.

As mentioned earlier, the interrogator receiver is identical to the transponder receiver 70 (with the exception that the SLS circuitry 132 and 134 as well as comparator 142 are not required). The data messages stored in storage means 162 in the parallel storage registers may be read out as desired on data bus 164.

FIG. 3 illustrates a typical message 170 transmitted by an interrogator 14 (FIG. 1) and includes a preamble section 172 and a data section 174. The preamble section is comprised of a predetermined number of bits (15 being used for illustration). Receiver 70 in FIG. 2 must detect the unique distribution of ones and zeros in the preamble in order to verify that the remaining portion of the data 174 is to be processed. A control signal 176 (identified in FIG. 3 as the ISLI Interrogator Side Lobe Inhibit) is transmitted substantially simultaneously with the preamble; the control signal 176 is transmitted, however, on the delta pattern of the interrogator antenna. It will be noted from FIG. 3 that the control signal code 176 is orthogonal to the preamble code 172.

In the preferred embodiment, the r.f. message received by the receiver 70 (FIG. 2) is down-converted to an i.f. frequency compatible with convolvers 74 and 76. In the preferred embodiment, convolvers 74 and 76 are surface acoustic wave (SAW) convolvers. SAW convolvers 74 and 76 perform the programmed matched filter function to despread the coded waveforms (the data representation of which is illustrated in FIG. 3). A matched filter has an impulse response equal to the time inverse of the impulse response of the desired received signal and thus has the characteristic of providing the maximum achievable signal-to-noise improvement on the received signal. In hostile environments, such as IFF system 10, the noise level will actually be generated by a jamming signal from an enemy, so signal-to-jammer improvement is obtained. Programmability is achieved by changing the impulse response of the filter. In a SAW convolver, the impulse response is the reference signals (namely reference signal A on reference port 82 (FIG. 2) and reference signal B on reference port 84 of convolver 76) which can be changed by electronically generating different waveforms. Thus, the SAW convolvers 74 and 76 can be used in spread spectrum systems to achieve signal-to-noise jammer improvement and be programmable to prevent spoofing and exploitation.

SAW convolvers 74 and 76 also provide time compression of the received signal energy. The sender tries to maximize the energy content of the transmitted data symbol to maximize the probability that the receiver can detect the symbol. Since the energy is equal to the peak power multiplied by the symbol length in time (for uniform amplitude signals), there are two choices: increase the power or increase the time length. In practical systems, power is constrained below some upper limit so typically symbol lengths are increased to achieve the desired energy content. However, as symbol lengths are increased, the ability of a receiver to detect all messages in a dense environment decreases due to mutual interference. It is a property of convolvers 74 and 76 acting as matched filters to compress the symbol energy into a pulse equal in width approximately to the inverse of the symbol bandwidth. For example, if the bandwidth were 100 MHz, the compressed pulse would occupy only about 10 ns of time regardless of the original transmitted length of the pulse. In this case, a second symbol could arrive at the input of receiver 70 from a second sender or interrogator only 10 ns ahead or behind the first symbol and still be detected by receiver 70. This greatly increases the data handling capability of receiver 70, thus allowing it to detect all messages as they are received.

The process used to adjust the bandwidth of the symbol is the spread spectrum technique. A pulse of CW carrier 10 .mu.sec long has a bandwidth equal to about 1/10 .mu.sec or 0.1 MHz. However, if the phase of the carrier is altered (say every 10 ns), then the bandwidth of the symbol will be 1/10 ns or 100 MHz. There would be a thousand such phase changes in the total symbol length. The sequence of these phase changes is the code that must be duplicated, time inverted and input to the convolves 74 and 76 as the reference waveforms (reference A and reference B) to provide the matched filter function.

A SAW convolver uses the non-linearities in the elastic constants for propagation of surface waves in a substrate (such as a LiNBO.sub.3 substrate) to form the convolution of the input signal with a reference signal. The general configuration of a SAW convolver is shown in insert 86 in FIG. 2. The efficiency of the non-linear interaction is proportional to the power density in the device. Therefore, acoustic signals, representing the input signal and reference waveforms, generated by interdigital transducers are compressed into a narrow band in the interaction region of the device. The compressed acoustic beams are directed into a thin film metallic waveguide 166 which confines the acoustic energy and serves as the integrating electrode for the convolver output. A more detailed description of a convolver suitable for the present application is entitled "Wideband Elastic Convolvers" by H. Gautier and C. Maerfeld, 1980 IEEE Ultrasonics Symposium, pages 30-36, 0090-5607/80/0000-0030.

The binary orthogonal key data demodulation is accomplished using two convolvers 74 and 76. The use of these convolvers provides one additional function, namely, the asynchronous correlation of the received symbol. In other words, regardless of when the input signal (symbol) arrives at receiver 70, convolvers 74 and 76 acting as matched filters will provide maximum signal-to-jammer improvement. There is no lengthy code search time required to synchronize to the incoming signal.

FIG. 4 illustrates the operation of convolver 74 to achieve synchronization and correlation of the incoming signal. An input signal consisting of, for example, five data bits (symbols) is input to the input port 78 of convolver 74 from the left and can be pictured as moving to the right as time progresses. At the instant in time illustrated in FIG. 4, bit 1 is approximately half way into convolver 74. Each bit is T units long and consists of the same code sequence. The reference waveforms are also T units long and generated from the same code sequence (but actually time reversed) enter the convolver 74 on reference port 82 from the right and travel to the left as time progresses. As is illustrated in FIG. 4, it can be seen that each reference symbol is spaced T units behind the end of the previous symbol, i.e., the referenece signal A has a 50% duty signal. The time line shown under convolver 74 illustrates the convolver output as time progresses where zero is the time depicted in FIG. 4. The numbered vertical arrows represent the compressed pulses which occur when a reference symbol and data bit overlap exactly while in convolver 74. It can be seen that in a period of time greater than T but less than 2T the first reference symbol and the first data bit will correlate while both are completely within the convolver. This correlation pulse is labeled 1. One-half T units later, the first reference and bit 2 will correlate while both are completely within the convolver. Hence, arrow 2 is at the same height as arrow 1. As the first reference is partially out of the convolver (moving from right to left), it will align with part of bit 3 that is entering the convolver from the left, yielding a partial correlation illustrated by the shorter arrow 3. A similar series of correlation pulses is produced as the second reference travels through convolver 74. It should be noted that on the time line each data bit has one and only one full correlation. These represent the compressed pulses that will be detected by the remainder of receiver 70. It should also be noted that full correlations only occur when a reference symbol is completely within convolver 74. Partial correlations which represent spurious, unwanted signals occur when a reference is partially in convolver 74. It is this timing relationship between the position of the reference and the desired output pulses that allows receiver 70 to sort out the full correlations from the partial correlations. In essence, receiver 70 only looks at the output when a reference symbol is completely within the convolver. The length of 2T for convolver 74 and a 50% duty cycle for the reference signal is the only combination which will produce one and only one full correlation on a data bit that arrives at any time. Accordingly, convolvers 74 and 76 perform the functions or are used to achieve receiver signal-to-jammer improvement, programmability, time compression of the received signal to decrease interference effects and asynchronous correlation to insure rapid detection of received messages. Typical values for T would be 10 .mu.sec such that the convolver would have a convolution interval of 2T or 20 .mu.sec.

As mentioned previously, the two convolvers 74 and 76 in FIG. 2 utilize binary orthogonal keying (BOK) data modulation. That is, a data bit "1" is communicated by transmitting a code which is correlated in upper convolver 74 and a data bit "0" is correlated in lower convolver 76. Data is determined by comparing the two convolver outputs to determine the larger. This comparison takes place in digital comparator 156 (FIG. 2). Typically, these data bits are on the order or 10 ns which make it extremely difficult to compare the magnitude of the two pulses. In order to enable comparison of the magnitudes of the output signals from correlators 74 and 76, each of said output signals is transmitted through video processing circuits 92 and 94 which are identical and comprise, in channel A, a log detector 96, a peak detector and stretch circuit 100 and a sample-and-hold circuit 104. The log detector 96 removes the i.f. carrier and reduces the dynamic range of the signal while the peak detecting and stretch circuit 100 and sample-and-hold circuit 104 reduce the pulse bandwidth of the signal in order to make it compatible with state-of-the-art A/D converters, such as A/D converter 122.

FIG. 5 illustrates the bandwith reduction accomplished through the use of the log detector 96, the peak detector 100 and sample-and-hold circuit 104. Waveforms 180, 182 and 184 are illustrative of the output from the convolver 74, the log detector 96 and the sample-and-hold circuit 104, respectively. For purposes of illustration, we have assumed a bit rate of 10 .mu.sec and a sampling rate of 20 MHz at a 50% duty cycle. Waveform 180 illustrates an i.f. correlated signal having both full correlation peaks 186 and partial correlation peaks 188. A 50% duty cycle is used since it has been shown that full correlations occur at a 50% duty cycle (see FIG. 4). Waveform 182 illustrates the conversion of the i.f. signal 180 to a video signal 182 having a reduced dynamic range. Waveform 184 illustrates the output from sample-and-hold circuit 104 and with a 20 MHz sampling rate, 50 samples of the convolver output are achieved every 10 .mu.sec. Not shown for reasons of clarity is the action of the peak detector circuit 100. As mentioned previously the width of the pulses 186 in signal 180 is approximately 10 ns in the example. After log detection in circuit 96 the pulses are still very short. The peak detector 100 determines the maximum voltage of the pulses and maintains the value until the value is sampled by the sample-and-hold circuit 104. This action, in essence, stretches the pulses to 50 ns (1/20 MHz). Accordingly, it is now possible, after going through A/D converters 122 and 124 to readily compare the magnitude of the signals in digital comparator 156.

Knowing when to look for data (called message synchronization) in a burst communication system is solved in one of two ways. Either a preamble is included at the beginning of a message to cue the receiver to receipt of data at a later time or the data is staggered in time about uniform time spacing in such a manner that the detection of data bits in that stagger pattern is the clue that a message has been received. Either technique is approximately equivalent. If the need did not exist to be able to limit the number of communication links to only partners of interest in IFF system 10, then either method of message synchronization would be acceptable. However, in an IFF system where it is necessary to communicate rapidly with partners in an environment where there are a large number of communicators and without limiting the number of communication links, interference effects would prevent the receipt of desired signals, thus delaying or preventing data transfer.

Under this condition, message synchronization is best accomplished with the use of a preamble. The receiver according to the present invention is capable of limiting the number of communication links through beam sharpening. To understand, it is necessary to review how a message is transmitted through space. The interrogator 14 (FIG. 1) who initiates the communication action transmits a preamble to the message over an antenna pattern known as a sum pattern and similarly transmits a control signal over a difference pattern.

FIG. 6 illustrates a typical sum pattern 190 and difference pattern 192. In an IFF system where there are a large number of transponders, there may be many transponders that can detect the preamble transmitted in the sum pattern 190 (also known as the main beam). Since these transponder receivers would typically process the received message and reply approximately in unison, a larger interference problem would exist at the receiver of the interrogator. In order to prevent this, it is desired to limit the transponder receivers (such as receiver 44 in aircraft 26 in FIG. 1) that reply to only those in the center portion of the sum pattern 190. This is called beam sharpening. Observing the delta pattern in FIG. 6, it will be noted that everywhere inside a portion of the main beam (namely between points 194 and 196), the difference pattern amplitude is less than the sum pattern amplitude and everywhere outside the points 194 and 196, the difference pattern amplitude is greater than the sum pattern amplitude. Beam sharpening is accomplished by comparing the amplitude of the received (sum pattern) signal to that of the control signal received on the difference pattern and replying only if the amplitude difference between the two exceeds some predetermined level. State another way and referring again to FIG. 6, if the sum pattern exceeds the difference pattern by an amplitude of x, the antenna beamwidth is 7 degrees. If the sum pattern exceeds the difference pattern by a greater amplitude y, then the antenna beamwidth is further reduced to 2 degrees.

In typical implementations of beam sharpening, a single pulse is transmitted on the main beam followed at some interval by a single pulse transmitted on the difference beam. The receiver detects the pulses and compares the amplitudes. This approach fails in environments where jamming is present because the jamming energy can cover the lower level control signal, thereby preventing an accurate comparison and thus preventing a reply. This can occur even when the jammer is not strong enough to prevent receipt of the message. This fault is often corrected by transmitting greater power on the control signal, but this requires more expensive transmitters and, in effect, reduces maximum range. This beam sharpening technique can also fail when a multipath is present such that the main beam can arrive twice, once coincident with the control signal after taking a longer path to the receiver. A third common failure mode is when an interferring signal arrives coincident with the control signal, a situation likely in dense environments.

In the system according to the present invention an interrogator transmitter transmits a preamble consisting of multiple symbols alternating in a known pattern between the "1" and "0" codes on the main beam and simultaneously transmitting the orthogonal code on the difference pattern, i.e., if a "1" is transmitted on the main beam, a "0" is transmitted on the difference pattern. These symbols are detected and correlated in filters matched to the respective "1" and "0" code patterns. In receiver 70 of FIG. 2, the preamble matched filter 130 detects and correlates the preamble on the sum pattern while the sidelobe suppression matched filter 132 detects and correlates the orthogonal control (ISLI) code on the difference pattern. Amplitude information is preserved during both correlation processes.

Referring again to FIGS. 2 and 6, the output of the preamble match filter 130 on conductor 140 is compared to the output of the sidelobe suppression match filter 132 on conductor 136 and if the amplitude of the preamble exceeds that of the sidelobe suppression by a predetermined amount which is inserted by SLS margin circuit 134 (for example an amplitude y in FIG. 6), then receiver 70 is assured that it is within the main beam (the 2 degree beamwidth) of the interrogator transmitter antenna. The comparator 142 will generate a signal which actuates clock circuitry 154 causing the data information following the preamble to be inserted in storage FIFO registers 162.

The advantage of this process is that additional signal-to-jammer improvement is obtained by matched filtering the preamble and control signals so that accurate amplitude comparisons can be made up to the point where enemy jammer energy destroys the ability to detect the message. In other words, if receiver 70 can detect the message, then it an also perform the beam sharpening process. This is accomplished without increasing the power on the control (ISLI) signal.

This technique also virtually eliminates the multipath problem because the signals are transmitted simultaneously and so multipath would appear as a second set of later signals which would not interfer with the reception of the true signals. The problem of an inteferring signal arriving simultaneously with the preamble and control signals is greatly reduced because of the large number of symbols in this signal. A single random interference or even several would not degrade the processing gain significantly.

With time stagger synchronization, additional symbols are added to the message and transmitted over the sum and difference antenna patterns similarly to the preamble and ISLI signals except these additional symbols are not used for synchronization functions. To achieve the same degree of beam sharpening available in the preferred embodiment disclosed herein, it would be necessary to determine amplitude information of the additional symbols. To do this would require determining amplitude information of all symbols since synchronization and therefore position of the added symbols is not known until the entire message has been received and stored in memory. In the time stagger receiver, amplitude information is typically not determined. The added cost and complexity to determine and store all of the required amplitude information is the reason that beam sharpening is best accomplished with a preamble. A receiver concept can be designed where the added symbols are added to the end of the message where they could be located after message synchronization and then processed as described herein to achieve beam sharpening to the same degree. In other words, the receiver according to the present invention will achieve synchronization regardless of the form of synchronization signals, either time stagger or preamble information.

The simultaneous transmission of the preamble and control (ISLI) signal is done primarily to prevent an enemy from exploiting the system by separating the preamble and control signal. Typically this requires the use of a second transmitter, one for the main beam and one for the control signal. With the beam sharpening technique incorporated in receiver 70, it is possible to adjust the number of preamble symbols so as to have one transmitter and divide the power during the preamble portion of the message and then regain the signal-to-jammer improvement by matched filtering the preamble and control signals.

The receipt of the preamble cues receiver 70 to look for data arriving at known time intervals following the preamble. For example, every ten microseconds the output of convolvers 74 and 76 can be compared to determine if a "1" or a "0" has been transmitted. These decisions can be stored in a memory for processing when the total message has arrived.

Receiver 70 can process many messages which arrive almost simultaneously. The arriving pulses after compression can be approximately 10 ns wide. These pulses are stretched to typically 50 ns and then decisions are made very 50 ns. Therefore, messages that arrive offset from each other by only 50 ns can be processed totally independently, almost as though the other message did not exist. As mentioned previously in discussion of FIG. 4, data symbols entering the receiver 70 10 .mu.sec apart leave the convolver 74 5 .mu.sec apart. Therefore data from the same message will occupy two of the 50 ns samples taken during the 10 .mu.sec interval. Since there are two hundred 50 ns intervals in a 10 .mu.sec interval, the receiver system 70 could conceivably handle 100 overlapping messages (this would mean that storage means 162 in FIG. 70 would require 100 FIFO registers, namely n=100 and clock circuit 154 would be required to generate 100 timing signals).

FIG. 7 illustrates a more detailed block diagram of the preamble matched filter 130. The preamble filter 130 provides a second stage of matched filtering of input signal 72. The preamble detection circuit 130 is used for detection of the interrogator preamble and to provide synchronization for data detection. In the preferred embodiment, the preamble detection is performed entirely digitally. The analog outputs from the video processing circuit means 92 and 94 are A/D converted in converters 122 and 124 and coupled to a digital differencing circuit 198. The B-channel signal is delayed (approximately one time slot, see waveform 224, FIG. 9) with respect to the A-channel signal and then the difference is formed. In this way, convolution peaks in the A and B channels do not cancel each other, but any common-mode noise or jamming cancels out. The bipolar output (see waveform 226, FIG. 9) of digital differencing circuit 198 is coupled by conductor 200 to a preamble amplitude correlator 202; correlator 202 is a 15-tap transversal filter. Fifteen taps are required to correspond to the 15 bits of preamble information (illustrated in FIGS. 3 and 8). As noted herein above, with a 20 MHz sampling rate, approximately 100 samples of amplitude are taken for each bit of information, thereby requiring 15.times.100=1500 words of storage in correlator 202. Fifteen taps 204 are taken from correlator 202 and multiplied by the tap weights 206 which are are the sign of the stored signals of the preamble code (in other words, a +1 corresponds to a "1" and a -1 corresponds to a "0"). In other words, the tap weights are matched to the preamble code. The outputs of the taps are summed in summer 208. The preamble amplitude correlator 202 output peaks when the incoming preamble is aligned with the appropriate time slot. The sidelobe suppression matched filter 132 operates in the same manner as preamble match filter 130 but is used to process the control (or ISLI) signal. Also in a similar manner, the tap weights are matched to the control signal code.

FIGS. 8 and 9 illustrate additional waveforms present at various points in receiver 70. The preamble 172 and control (or ISLI) code 176 are assumed to be the same as illustrated in FIG. 3. Waveforms 210 and 212 illustrate the portion of the outputs of convolvers 74 and 76 due to the preamble code 172. Waveforms 214 and 216 illustrate the portion of the outputs of the convolvers 74 and 76 due to the control (or ISLI) code 176. Waveform 218 represents the composite of waveforms 210 and 214 while waveform 220 represents the composite of waveforms 212 and 216.

FIG. 9 illustrates waveform 222 (the B convolver output) delayed 1 sample length (versus waveform 220, FIG. 8). Waveform 224 represents the bipolar output of digital difference circuit 198. Waveform 226 represents the output waveform from preamble detection circuit 130 while waveform 228 represents the output waveform of sidelobe suppression detection circuit 132.

Although the present invention has been shown and illustrated in terms of a specific apparatus, it will be apparent that changes or modifications can be made without departing from the spirit and scope of the invention as defined by the appended claims.