WO2010129646A1

WO2010129646A1 - A method and system of estimating signal interference in a predetermined frequency range

Info

Publication number: WO2010129646A1
Application number: PCT/US2010/033683
Authority: WO
Inventors: Edward M. Valovage; Stephen E. Mcmahon
Original assignee: Sensis Corporation
Priority date: 2009-05-05
Filing date: 2010-05-05
Publication date: 2010-11-11

Abstract

A system and method for estimating signal interference in a predetermined frequency range is disclosed. The system for estimating signal interference comprising a plurality of antennas in an antenna array, a multi-channel signal processor and a memory for storing data. The plurality of antennas in the antenna array receive a plurality of signals in the predetermined frequency range during at least one observation time period and transmit the plurality of signals to the multi-channel signal processor. An algorithm in the multi-channel processor determines a number of signals present in the at least one observation time period and estimates a signal interference in the predetermined frequency range during the at least one observation time period from the determined number of signals present.

Description

A METHOD AND SYSTEM OF ESTIMATING SIGNAL INTERFERENCE IN A PREDETERMINED FREQUENCY RANGE

FIELD OF THE INVENTION

The invention relates to a system and method for improving surveillance system performance by estimating the number of signals present and the signal interference in a predetermined frequency range.

BACKGROUND OF THE INVENTION

The 1090 MHz aviation frequency band is being used by a growing number of aircraft, applications and equipment types. Some of the types of surveillance systems using the 1090 MHz aviation frequency band include:

Air Traffic Control Radar Beacon System (ATCRBS) and Mode S transponders respond to ground-based surveillance radar as well as to Traffic Collision Avoidance System (TCAS) interrogations from other aircraft;

Automatic Dependent Surveillance - Broadcast (ADS-B) information is sent by aircraft for reporting of position and other flight information;

Ground stations transmit Traffic Information Service - Broadcast (TIS- B) and cross-link ADS-B rebroadcasts (ADS-R); and

Military applications also use 1090 MHz for similar surveillance functions, in some cases using more complex waveforms. In all of the above protocols, information is transmitted asynchronously, and some modest loss of information due to overlapping messages or garbling is accepted.

As time goes on, message traffic density is expected to increase. ADS-B is being rolled out as the primary method of surveillance for air traffic management. The 1090 MHz frequency is used by aircraft transponders responding to interrogations from ground radar and from other aircraft, aircraft transmitting Automatic Dependent Surveillance - Broadcast (ADS-B) information, and by ground stations transmitting traffic information and cross-link ADS-B rebroadcasts. Since this means more equipment transmitting more messages in the 1090 MHz band, the 1090 MHz aviation frequency band is expected to soon reach critical interference levels. The failure to correctly receive one message due to overlap or garbling from another message is commonly referred to as "self-interference" or just "interference."

In order to design modern equipment to operate on this 1090 MHz band, it is desirable to quantify the existing message density. To do this, a number of simulations based on known models have been used. However, since these models use the number and location of aircraft as basic parameters of the models and make significant assumptions about the details of the overall environment, the aforementioned parameters alone are not sufficient to determine message density. Further, the specific equipage of each aircraft including the type of transponder, number of antennas, and whether or not it is TCAS-equipped are all significant factors. Even the equipment manufacturer is important, particularly with TCAS equipment where the behavior is not consistent across manufacturers. Typically, assumptions are made using a statistical selection for the equipage and behavior details. Another factor is the number, locations, and operating modes of interrogating radars in the area. The resulting predictions of message density are therefore no better than the assumptions upon which the predictions are based. In the above protocols, there is concern that the performance loss due to overlapping messages will soon become unacceptable as message densities increase. Again, to quantify the problem, analyses and simulations have been used that make assumptions about the details of the interference environment. However, there is a strong need to measure actual RF environments in the field to verify the models. Measuring dense RF interference environments is difficult because RF receiver/decoders cannot receive and decode (and thereby count) all of the messages in the RF environment. Overlapping messages will cause garbling, thereby causing the receiver/decoders to miss some messages, which results in underestimating the measured RF density data. One known method to estimate the number of messages when they overlap is the observation of a received power envelope. Changes in the power level may indicate that a message has started or ended even if another was already in progress. These start and end estimates can be used to infer the presence of individual messages that can then be counted. While this method is simple in principle, in practice it is quite complex. One difficulty is that in the 1090 MHz frequency band, at least three different types of messages are broadcast, Mode S long, Mode S short and ATCRBS. The relevant message characteristics are presented in Table 1 below. The method described senses message start and end times from the power envelope, and infers a message type based on the durations of each type. As shown in Table 1, the ATCRBS message type is very sparse (low duty cycle) making it difficult to determine if an individual pulse is from a new message or a continuation of a message already in progress.

Table 1

Another difficulty in estimating the number of messages in this manner is that messages do not have equal power, so the instantaneous number of messages cannot be directly inferred from the power. A further difficulty is that the messages may not combine with additive power. Destructive interference can actually lower the received power of one message when a second one begins, making it difficult to identify whether a change was the start of a new message or the end of a message in progress. In one example of the power envelope process, eight hypothetical messages of equal length having varied transmission start times are shown in Fig. 1. For simplicity of the diagram, they are shown as constant power envelopes rather than a series of pulses. The plot in the very back labeled "sum" depicts the number of simultaneous messages at each point in time in the example. If an estimate of this number of messages is made as a function of time, i.e. estimating the back plot, the contributing messages can be inferred, essentially moving from back to front in the drawing.

Using a portion of the example shown in Fig. 1, four overlapping messages with different start times are added together to measure the power level and estimate the number of messages, as shown in Fig. 2(a) (which corresponds to roughly the first 40% of the example in Fig. 1). In the example shown in Fig. 2(a), each of the four overlapping messages has the same amplitude and combine additively. As shown in Fig. 2(a), a solid outline indicates the correct message envelope sum. The waveform that fills in the outline is the instantaneous power of the combined messages.

Due to the pulsed nature of the messages, it is challenging to identify the start and end times of each message since the pulses, or "on" times, of each message only line up on occasion, as shown in Fig. 2(a). Therefore, in this method messages can only be de-interleaved to provide a reasonable message count if the messages are transmitted on same frequency and with similar power levels.

Fig. 2(b) shows an example containing four overlapping messages with different start times similar to Fig. 2(a) but with slightly different frequencies for the messages, causing a more complex but realistic interference. In this example, note that the leading and trailing edges of the messages are less defined and are difficult to discern in time in Fig. 2(b). More specifically, in the portion of Fig. 2(b) where there are three messages simultaneously present in time, the first pulse actually reaching the amplitude of 3 is well into the third message at approximately time sample 6000. This is due to the fact that not only do the pulses not line up, as before, but when they do, the carrier frequencies may not be in phase. This results in the signals combining constructively and destructively at different times, even within one message. This is typically what will happen in an actual RF environment with non-coherent transmitting sources, and it points out another limitation of this method. Another difficulty in using this approach for estimating the number of messages is that ATCRBS messages are very sparse with a duty cycle from 5% to 30%, making the identification of message start and stop very difficult. Therefore, the method of observing the received power envelope cannot provide a sufficiently accurate estimate of the RF interference environment in a given frequency range. Another method for successfully counting message density in the presence of overlap is to have multiple receivers with directional antennas. The directional antennas effectively limit the azimuth coverage of each antenna and receiver, lowering the number of messages received at any one receiver and limiting the number of message overlaps, which enables decoding and more accurate counting of the messages, compared to the process of observing the received power envelopes. The directional antenna is likely to have higher gain causing more distant messages to be received, but this can be controlled by raising the detection threshold.

For example, a directional antenna having four sectors is shown in Fig. 3. The receiver is in the center with its intended coverage volume represented by the dashed line in Fig. 3. If an omni-directional victim receiver is placed in the center of the relevant volume of coverage shown in Fig. 3, the received message traffic density would be difficult to de-interleave and count due to the sheer volume. By using four sectorized antennas with one receiver assigned to each sector, the message density received at each receiver is reduced, thereby enabling de-interleaving of the messages and a more accurate count. However, one problem with this measurement technique is that the coverage area of the four sectorized receivers must overlap to some extent, which affects the message count. Even if redundant receptions are removed from across the overlapping portions of the sectors, the total coverage area may not be the same as the original omni-directional victim receiver. Another approach is to use multiple receivers that are spread through the area with their receive thresholds lowered, as shown in Fig. 4. This approach is similar to the sectorized antennas in that coverage area is reduced to lower the message density and allow a more accurate count. This method has similar problems with the overlap of the receiver coverage areas and the total coverage area may not be the same as the original omni-directional victim receiver.

Another drawback with the sectorized antenna and the distributed antenna approaches is that more receivers are needed to provide coverage to characterize a given area. Further, these approaches add uncertainty if the goal is to estimate the message density at a given antenna of greater coverage. A model would be needed to determine how the message traffic received at multiple lower-coverage receivers (e.g., sectorized antenna and the distributed antenna) can be combined to estimate the message traffic received in the single antenna with greater coverage. If the goal is to make a measurement to validate a model, however, it is not desirable to use a model within the measurement.

What is needed is a method for measuring RF message density to determine the RF interference environment that solves the limitations of the techniques described above, and may be combined with them to improve overall performance.

SUMMARY OF THE INVENTION In one embodiment of the present invention, the method for estimating signal interference in a predetermined frequency range comprises the steps of arranging a plurality of antenna elements in an antenna array, connecting the antenna array to a multi-channel signal processor, receiving a plurality of signals in the predetermined frequency range at the antenna array for at least one observation time period, transmitting the plurality of signals received to the multi-channel signal processor, determining a number of signals present in the predetermined frequency range during the at least one observation time period and estimating signal interference in the predetermined frequency range for the at least one observation time period from the determined number of signals present. In one embodiment, determining a number of signals present comprises the steps of generating a covariance matrix of the plurality of signals received for the at least one observation time period, wherein covariances in the covariance matrix are determined from the direction of arrival of the plurality of signals; and computing N eigenvalues by performing an eigendecomposition of the covariance matrix of the plurality of signals received for the at least one observation time period, where N represents a number of signals present in the predetermined frequency range for the at least one observation time period. In one embodiment, the multi-channel signal processor in the method of the present invention uses a super resolution technique with a direction finding algorithm to compute the N eigenvalues. In another embodiment, the multi-channel signal processor in the method of the present invention uses the Multiple Signal Classification (MUSIC) algorithm to compute the N eigenvalues.

In one embodiment of the method of the present invention, the covariance matrix is Ry = E {xjX_j }, where: xj is the baseband received signal samples at antenna element i, and * is the complex conjugate.

In one embodiment of the method of the present invention, estimating signal interference comprises the steps of decomposing a time history of the computed N eigenvalues to isolate each of the plurality of messages and counting each of the plurality of messages present in the predetermined frequency range for the at least one observation time period. In this embodiment, decomposing the time history comprises determining a start time and a stop time for each signal in the predetermined frequency range received by the plurality of antenna elements during the at least one observation time period and determining whether a signal is a single message or a plurality of messages by matching the start time and the stop time for each signal with lengths of known messages. In another embodiment, estimating signal interference in the predetermined frequency range is from the computed N eigenvalues. In one embodiment, the multi-channel signal processor determines a message type and a directional of arrival for each message received by the antenna array in the predetermined frequency range during the at least one observation time period. In another embodiment, the method further comprises spacing adjacent antenna elements in the antenna array by a distance λ/10 <= d <= 4λ, where λ is a center frequency in the predetermined frequency range. In yet another embodiment, adjacent antenna elements in the antenna array are spaced by a distance λ/4 <= d <= λ, where λ is a center frequency in the predetermined frequency range. In one embodiment, the method further comprises changing a duration of the at least one observation time period based on the number of signals received.

In one embodiment of the present invention, the system for estimating signal interference in a predetermined frequency range comprises a plurality of antenna elements in an antenna array, a multi-channel signal processor and a memory for storing data, wherein the plurality of antenna elements in the antenna array receive a plurality of signals in the predetermined frequency range during at least one observation time period and transmit the plurality of signals to the multi-channel signal processor, and an algorithm in the multi-channel processor determines a number of signals present in the at least one observation time period and estimates a signal interference in the predetermined frequency range during the at least one observation time period from the determined number of signals present.

In one embodiment, the algorithm generates a covariance matrix of the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period, and performs an eigendecomposition of the covariance matrix of the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period to determine N eigenvalues, where N represents a number of signals measured in the predetermined frequency range during the at least one observation time period. In another embodiment, the algorithm estimates signal interference in the predetermined frequency range from the computed N eigenvalues. In one embodiment, each of the plurality of signals comprises at least one message. In another embodiment, the multi-channel signal processor determines a start time and a stop time for each signal in the predetermined frequency range received by the plurality of antenna elements during the at least one observation time period and determines whether a signal is a single message or a plurality of messages by matching the start time and the stop time for each signal with lengths of known messages.

In one embodiment, the multi-channel signal processor determines a message type and a direction of arrival for each message received by the antenna array in the predetermined frequency range during the at least one observation time period. In one embodiment, the message type comprises one of air traffic secondary radar transponders, TCAS messages and ADS-B messages.

In another embodiment, the multi-channel signal processor estimates signal interference by decomposing a time history of the computed N eigenvalues to isolate and count each message in the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period. In yet another embodiment, the multi-channel signal processor stores the determined number of messages received and direction of arrival of each of the messages received during the at least one observation time period in the memory.

In one embodiment, the multi-channel processor uses a super-resolution direction finding algorithm to determine the number of messages and direction of arrival of each of the messages received in the predetermined frequency range during the at least one observation time period. In another embodiment, the algorithm in the multi-channel processor uses the Multiple Signal Classification (MUSIC) algorithm to determine the number of messages and direction of arrival of each of the messages received in the predetermined frequency range during the at least one observation time period. In one embodiment of the system of the present invention, the duration of the at least one observation time period is adjustable. In one embodiment, the plurality of antenna elements in the antenna array are spaced by a distance λ/10 <= d <= 4λ, where λ is a center frequency in the predetermined frequency range. In another embodiment, the plurality of antenna elements in the antenna array are spaced by a distance λ/4 <= d <= λ, where λ is a center frequency in the predetermined frequency range.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 shows an example of transmitted messages overlapping in an RF environment;

Fig. 2(a) shows the power envelope for four overlapping messages having the same power levels that are transmitted at the same frequency;

Fig. 2(b) shows the power envelope for four overlapping messages having the same power levels that are transmitted at slightly different frequencies;

Fig. 3 shows an example of a sectorized antenna approach for measuring transmitted messages and interference in an RF environment; Fig. 4 shows an example of a distributed antenna approach for measuring transmitted messages and interference in an RF environment;

Fig. 5 shows an example of a multichannel processing steps in one embodiment of the present invention;

Fig. 6 shows an example of a four antenna array and transmitted messages according to the present invention for measuring transmitted messages and interference in an RF environment;

Fig. 7 is a flowchart of the analysis steps performed by the multi-channel signal processor in one embodiment of the present invention;

Fig. 8 shows an example of the estimate of the number of signals of one embodiment of the present invention using a noise level of -90 dBm;

Fig. 9 shows an example of the estimate of the number of signals of one embodiment of the present invention using a noise level of -80 dBm;

Fig. 10 shows the accuracy of the method of the present invention for one, two, or three sources; and Fig. 11 is a perspective view of one embodiment of a four element circular antenna array of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

The present invention uses multiple closely-spaced antenna elements 10 in an antenna array 15 and a multi-channel signal processor to perform a joint, coherent analysis of the messages arriving at a single location for a predetermined frequency range. The system and method of the present invention enable determination of the number of independent messages arriving at the array of closely-spaced antenna elements 10 as a function of time. The system and method of the present invention also enable determination of the direction of arrival of each of the number of independent messages arriving at the array of closely-spaced antenna elements 10. The ability to determine a message density and estimate a signal interference for a predetermined frequency range over a predetermined time period provided by the present invention is distinctly different from prior art methods that simply place a separate receiver and message decoder at each antenna or that attempt to identify and count messages based on power envelope.

In one embodiment, the multi-channel signal processing steps, shown in Fig. 5, are performed in a multi-channel signal processor to de-interleave and count the messages present in the predetermined frequency as a function of time. The multi- channel signal processor uses a direction finding algorithm and a super resolution technique to determine the number of messages present and the direction of arrival of each of the received messages in the predetermined time period from the RF signals received at the multiple closely-spaced antenna elements 10. The algorithm uses the determined number of messages and their associated direction of arrival data to determine a message density 20 for a predetermined frequency range for the predetermined time period. The algorithm then uses the determined message density 20 to estimate the signal interference within the predetermined frequency range for the predetermined time period. In one embodiment, the multi-channel signal processor uses the Multiple Signal Classification (MUSIC) algorithm to analyze the received RF signals to determine the number of messages in the received signals and the direction of arrival of each of the messages in the present invention. The system and method of the present invention enable the determination, as a function of time, of the number of messages present without the problems and inaccuracies of trying to follow the power envelope as discussed above, thereby enabling the present invention to estimate the message density (i.e., sum plot or rearmost plot shown in Fig. 1) without the measurement problems previously discussed and illustrated in Figs. 2(a) and (b). In the present invention, the de-interleaving and counting of the messages is performed, from a rudimentary perspective, in a manner similar to the power envelope approach, but the message count estimate is far more accurate than can be provided by the power envelope approach.

For example, if the baseband received signal samples at element i is denoted ^x', the covariance estimate is the matrix

where * is the complex conjugate. This covariance estimate is taken over a predetermined observation or time period. The covariances are determined by the direction of arrival of the signals. The antenna array 15 consists of four ideal omni- directional antenna elements 10 and the geometry and spacing of the 4-element antenna array 15 is shown in Fig. 6. The length of the observation time period can be varied based on the number of messages being received over the predetermined time period. The predetermined observation period of time is an important system parameter that can be optimized to provide an accurate message count estimate. In this example, after the covariance matrix is generated, the multi-channel processor performs an eigendecomposition of the covariance matrix, as shown in Fig. 7. The eigendecomposition of the covariance matrix for the four antenna array 15 shown in Fig. 6 will result in four eigenvalues, /L₁ - A₄. If there are 3 or fewer signal sources, the smallest eigenvalue will represent noise and the larger eigenvalues will represent signals received. By knowing the noise level of the system, the number of eigenvalues above the noise indicates the number of sources. If there are N antenna elements, the number of signal sources can be determined from the eigenvalues if the number of signal sources is less than or equal to N. If there are N antenna elements, the angles of arrival of the signal sources can be determined if the number of signal sources is less than or equal to N-I. The noise threshold for this estimate is another parameter of the method of the present invention that can be adjusted.

In the example shown in Fig. 6, multiple Mode S long messages with proper pulsed behavior and random data contents are shown approaching the 4-antenna array 15. For simplicity and ease of understanding, messages having different lengths, such as Mode S short, and messages having significantly smaller lengths (i.e., duration) and/or duty cycles, such as ATCRBS messages, were not included in this example. The eight Mode S long messages were simulated with various degrees of overlap in time. In this example, the relative time of arrival, message overlap and time extent of each of the eight messages is the same as the example shown in Fig. 1.

The multiple simulated Mode S long messages are at slightly different frequencies, as shown in Table 2 below so that the coherency of the overlapping signals is realistic. Table 2 shows the frequency, power, and arrival angle of the messages in this example.

Table 2

With this realistic received signal coherency and the addition of noise in varying degrees, the estimate of the covariance matrix is realistic. The covariance matrix for this example is shown in Table 3

Table 3

The variation in frequency of the simulated Mode S long messages is similar to the degree of frequency variation that may be found in older transponder equipment. This variation in frequency is a little greater from the center frequency than the frequency variation that will generally be found in newer transponder equipment. In this example, the power levels are spread from -45 to -57 dBm, and the angles of the received signals only cover a range of 40° (-17° to +23°). This is more demanding for signal resolution, because all of these messages could fall within one beam of the sectorized measurement approach discussed earlier.

Fig. 8 shows the estimate result of the number of signals with a noise level of -90 dBm, which is well below the power level of all of the signal levels shown in Fig. 1. Fig. 9 shows the estimate result of the number of signals with a noise level of -80 dBm.

In Figs. 8 and 9, the data points "x" are the estimated number of sources as a function of time and the solid lines are the true number of sources. In this example, the output of the MUSIC algorithm for almost all of the data points correctly determines the true number of sources. The estimated number of sources can be used in conjunction with the process of identifying signal envelopes illustrated in Fig. 2 to de-interleave and count the messages.

The performance of the present invention is a significant improvement compared to looking just at amplitude or phase changes of prior art methods because the envelope is estimated more accurately in time and the instantaneous number of signals is more accurate. Note that some of the overlapping sources are as close to each other as 5°. Since the array is only 4 elements at A/2 spacing, which is an aperture of only 0.8 meter, the system has far better resolving capability than a sector antenna of a similar aperture.

The estimate result of the number of signals with the noise level increased to - 80 dBm is shown in Fig. 9. The increased noise level causes some of the estimates to bounce between the correct value and a wrong value. In one embodiment, additional logic and averaging has been added in the covariance estimation process to resolve this ambiguity. Thus, the present invention provides an accurate method for counting the messages present having more than 20 dB SNR, which is a very useful capability.

The performance of this measurement technique is dependent on many parameters. Major factors are the message type and length, the signal-to-noise ratio of multiple signals, the geometry of the array and number of elements, the numbers of overlapping signals, and the frequency spread. Fig. 10 shows the accuracy of the method of the present invention for each section where there are one, two, or three sources, plotting the fraction of correct message results. As shown in Fig. 10, the performance drops off sooner for three messages than for one or two messages, as expected. In one embodiment, the method incorporates an algorithm to infer the start and end of individual messages and estimate the density.

Fig. 11 shows one embodiment of an antenna 10 used in an antenna array 15 of the system of the present invention. The antenna is not a linearray but each corner of the antenna 10 is a separate receive element, providing 360° coverage. The antenna 10 has a cube shape and the overall dimension of the cube is approximately 6-inch on each side. Although the backward visibility is reduced for each element, a cluster of 3 or 4 sources can still be resolved if they are strong enough.

While the array shown in Fig. 11 has an overall dimension of a cube approximately 6-inch on each side, the antenna 10 can be larger and/or contain more elements to meet overall system requirements on density and resolution. The system and method of the present invention provides a very high message resolution capability using a very modest-sized array and low number of channels. The eigenanalysis approach allows n elements to resolve up to n sources without requiring that the sources be in different pre-formed beams or coverage regions, whereas in the prior art methods shown in Figs. 3 and 4, the 4 beams will only resolve the sources if there is one in each beam. An array of the present invention, such as shown in Fig. 11 will resolve sources within 5° at SNRs of 20 dB or more. At the lower SNRs, the present invention will still resolve targets that are separated by more than 5°. While the present invention has been particularly shown and described with reference to the preferred mode as illustrated in the drawing, it will be understood by one skilled in the art that various changes in detail may be effected therein without departing from the spirit and scope of the invention as defined by the claims.

Claims

What is claimed is:

Claim 1. A method of estimating signal interference in a predetermined frequency range, the method comprising the steps of: arranging a plurality of antenna elements in an antenna array; connecting the antenna array to a multi-channel signal processor; receiving a plurality of signals in the predetermined frequency range at the antenna array for at least one observation time period; transmitting the plurality of signals received to the multi-channel signal processor; determining a number of signals present in the predetermined frequency range during the at least one observation time period; and estimating signal interference in the predetermined frequency range for the at least one observation time period from the determined number of signals present.

Claim 2. The method of claim 1, wherein determining a number of signals present comprises the steps of: generating a covariance matrix of the plurality of signals received for the at least one observation time period, wherein covariances in the covariance matrix are determined from the direction of arrival of the plurality of signals; and computing N eigenvalues by performing an eigendecomposition of the covariance matrix of the plurality of signals received for the at least one observation time period, where N represents a number of signals present in the predetermined frequency range for the at least one observation time period.

Claim 3. The method of claim 2, wherein the multi-channel signal processor uses a super resolution technique with a direction finding algorithm to compute the N eigenvalues.

Claim 4. The method of claim 2, wherein the multi-channel signal processor uses the Multiple Signal Classification (MUSIC) algorithm to compute the N eigenvalues.

Claim 5. The method of claim 2, wherein the covariance matrix is RiJ = E (X₁X_j*), where: x, is the baseband received signal samples at antenna element i, and

* is the complex conjugate.

Claim 6. The method of claim 1, wherein estimating signal interference comprises the steps of decomposing a time history of the computed N eigenvalues to isolate each of the plurality of messages and counting each of the plurality of messages present in the predetermined frequency range for the at least one observation time period.

Claim 7. The method of claim 6, wherein decomposing the time history comprises determining a start time and a stop time for each signal in the predetermined frequency range received by the plurality of antenna elements during the at least one observation time period and determining whether a signal is a single message or a plurality of messages by matching the start time and the stop time for each signal with lengths of known messages.

Claim 8. The method of claim 7, wherein the multi-channel signal processor determines a message type and a directional of arrival for each message received by the antenna array in the predetermined frequency range during the at least one observation time period.

Claim 9. The method of claim 1, further comprising spacing adjacent antenna elements in the antenna array by a distance λ/10 <= d <= 4λ, where λ is a center frequency in the predetermined frequency range.

Claim 10. The method of claim 9, wherein adjacent antenna elements in the antenna array are spaced by a distance λ/4 <= d <= λ, where λ is a center frequency in the predetermined frequency range.

Claim 11. The method of claim 1, further comprising changing a duration of the at least one observation time period based on the number of signals received.

Claim 12. The method of claim 2, wherein estimating signal interference in the predetermined frequency range uses the computed N eigenvalues.

Claim 13. A system for estimating signal interference in a predetermined frequency range comprising: a plurality of antenna elements in an antenna array; a multi-channel signal processor; and a memory for storing data, wherein the plurality of antenna elements in the antenna array receive a plurality of signals in the predetermined frequency range during at least one observation time period and transmit the plurality of signals to the multi-channel signal processor, and wherein an algorithm in the multi-channel processor determines a number of signals present in the predetermined frequency range during the at least one observation time period and then estimates a signal interference in the predetermined frequency range during the at least one observation time period from the determined number of signals present.

Claim 14. The system of claim 13, wherein the algorithm generates a covariance matrix of the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period, and performs an eigendecomposition of the covariance matrix of the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period to determine N eigenvalues, where N represents a number of signals measured in the predetermined frequency range during the at least one observation time period.

Claim 15. The system of claim 13 , wherein each of the plurality of signals comprises at least one message.

Claim 16. The system of claim 15, wherein the multi-channel signal processor determines a start time and a stop time for each signal in the predetermined frequency range received by the plurality of antenna elements during the at least one observation time period and determines whether a signal is a single message or a plurality of messages by matching the start time and the stop time for each signal with lengths of known messages.

Claim 17. The system of claim 16, wherein the multi-channel signal processor determines a message type and a direction of arrival for each message received by the antenna array in the predetermined frequency range during the at least one observation time period.

Claim 18. The system of claim 17, wherein the message type comprises one of air traffic secondary radar transponders, TCAS messages and ADS-B messages.

Claim 19. The system of claim 13, wherein the multi-channel signal processor estimates signal interference by decomposing a time history of the computed N eigenvalues to isolate and count each message in the plurality of signals received by the plurality of antenna elements in the predetermined frequency range during the at least one observation time period.

Claim 20. The system of claim 19, wherein the multi-channel signal processor stores the determined number of messages received and direction of arrival of each of the messages received during the at least one observation time period in the memory.

Claim 21. The system of claim 13 , wherein the multi-channel processor uses a super-resolution direction finding algorithm to determine the number of messages and direction of arrival of each of the messages received in the predetermined frequency range during the at least one observation time period.

Claim 22. The system of claim 13 , wherein the algorithm in the multi-channel processor uses the Multiple Signal Classification (MUSIC) algorithm to determine the number of messages and direction of arrival of each of the messages received in the predetermined frequency range during the at least one observation time period.

Claim 23. The system of claim 13, wherein a duration of the at least one observation time period is adjustable.

Claim 24. The system of claim 13, wherein the plurality of antenna elements in the antenna array are spaced by a distance λ/10 <= d <= 4λ, where λ is a center frequency in the predetermined frequency range.

Claim 25. The system of claim 24, wherein the plurality of antenna elements in the antenna array are spaced by a distance λ/4 <== d <= λ, where λis a center frequency in the predetermined frequency range.

Claim 26. The system of claim 14, wherein the algorithm estimates signal interference in the predetermined frequency range from the computed N eigenvalues.