Method, observation device, system and computer program for characterizing radio frequency interference caused by multiple sources

Abstract

The present disclosure relates to characterizing radio frequency interference caused by multiple sources. At least one observation device is used to successively scan multiple radio frequency bands at different spatial locations and at different times to perform interference measurements (I2) in the multiple radio frequency bands, providing observations at successive times of the current location of the observation device and of the received interference power in the observed frequency bands. For each interference source and each observation, a likelihood probability is calculated (S1) that attributes an observation to an interference source. This likelihood probability calculation is used to assign each of the observations to an interference source (S2). More specifically, the likelihood probability calculation (S1) is based on both the frequency observations and the interference power observations.

Description

[Technical Field]

[0001] TECHNICAL FIELD This disclosure relates to monitoring a wireless environment. [Background technology]

[0002] Recently, many wireless communication devices operate in public bands, and therefore, they must deal with the coexistence of interference. Characterization of interference is useful for monitoring the radio environment and / or managing radio resources. In particular, there is a need for characterization of wireless interference. Summary of the Invention [Problem to be solved by the invention]

[0003] Interference monitoring in a wireless environment where multiple interference sources are present typically requires the analysis of observations obtained from several devices. The problem becomes somewhat more complex when the sources move during the observation. The problem is usually to identify the interference as separate sources and then characterize the following properties of each source: Operating frequency band, Geolocation, · Activation rate.

[0004] This disclosure aims to improve the situation. [Means for solving the problem]

[0005] To that end, the present disclosure provides a method for characterizing radio frequency interference caused by multiple sources, comprising: using at least one observation device that sequentially scans a plurality of radio frequency bands at different spatial locations and at different times to perform interference measurements in said plurality of radio frequency bands; Current observation frequency, The current location of the observation device, and the received interference power in the observed frequency band, providing observations at successive time points of For each interference source and each observation, calculating a likelihood probability of attributing an observation to an interference source; assigning each of said observations to one interference source using said likelihood probability calculation; For each interference source assigned one of the observations, estimating a location and frequency occupancy of the interference source; Including, More specifically, we propose a method in which the likelihood probability calculation described above is based on both frequency observations and interference power observations.

[0006] It is therefore proposed to calculate likelihood probabilities using both frequency observations and, here in particular, power observations, the power observations being related to the distance between the interference source and the observing device.

[0007] For example, the likelihood probability calculation can involve simultaneous Bayesian inference of frequency and interference power observations. Alternatively, Euclidean distance calculation techniques can be used.

[0008] In one embodiment, a database storing previous detections can be used. Typically, in one embodiment, the likelihood probability calculations can use such a database storing, for each of the interference sources, at least data on the location of the interference source and data on the occupancy of at least two radio frequency bands by the interference source.

[0009] Typically, these stored data can be derived from previous observations, so that the data on the location of an interfering source can be data on the source's past location, and the data on the occupancy of an interfering source can typically be data on the source's past frequency occupancy.

[0010] The database mentioned above can be updated after each assignment of an observation to one interference source, taking into account the next iteration of the method.

[0011] Thus, the current determination of multiple interference sources can be facilitated by the use of a database that stores data of previous identifications of interferers (location / frequency band(s) of interference). Conversely, after the determination of an interferer, the contents of the database are preferably updated so that it can be efficiently used for the next future iteration of the method.

[0012] In a particular embodiment, for each interference source, the database includes: The historical average position, and a probability map of past discretized positions, At least one of the following can be provided.

[0013] Thus, the database can provide the historical average position of a source, and more specifically, this average can be inverse weighted as a function of the date of determination of this source.

[0014] For example, depending on the confidence of the current determination of the interference source, the past average position can be used in the context of the "hard decision" embodiment specified below, and the probability map can be used in the context of the "source sampling" embodiment also described below.

[0015] Furthermore, for each interference source, the database - historical average radio frequency band occupancy, and - Probability maps of past radio frequency range occupancy, At least one of the following can be provided.

[0016] The database may store, for each interference source, occurrence data of previous observations already assigned to this interference source.

[0017] This knowledge of previous observations associated with a source can also be useful in calculating the probability that the current observation is associated (or not associated) with this same source.

[0018] For a current observation, a likelihood probability of one of the interference sources associated with the current observation is calculated based on a comparison between data in the database and data for the given observation, the comparison being: the degree of similarity between the level of radio frequency power measured in the current observation and the closest level of radio frequency power corresponding to this source given the past location of the source in said database relative to the current location of the observing device; the similarity between the radio frequency band occupied by the current observation and the radio frequency band occupied by the source according to said database; This includes the determination of:

[0019] Thus, likelihood probabilities are calculated based on two items that ultimately can correspond to the same source. the current position of this source relative to the observation device given by the radio frequency power of the signal received and measured by the observation device, and The frequency band(s) in which such a source is active.

[0020] In the "hard decision" embodiment described above, the assignment of at least one observation to an interference source is sorting the likelihood probabilities calculated for each interference source to attribute the at least one observation to the interference source; selecting the highest calculated likelihood probability and assigning the at least one observation to the interference source having the highest calculated likelihood probability; may include:

[0021] In this embodiment, the likelihood probability is, for example, the distance between the observation device and the source in the radio frequency power domain (which is then usually linked to the spatial domain); the distance between the frequency band occupied by said source according to the database and the frequency band in which interference is measured according to said observations; can be calculated based on the Euclidean distance calculation.

[0022] In a "source sampling" embodiment, the assignment of at least one observation to an interference source is sampling each source allocation based on an associated calculated likelihood probability, where the initial random draw is followed by calculation iterations until successive samples converge to the associated calculated likelihood probability; Includes.

[0023] For example, with a view to selecting between a "hard decision" embodiment or a "source sampling" embodiment, each likelihood probability can be compared to a threshold: - if the calculated likelihood probability is above the threshold, selecting the highest calculated likelihood probability and assigning the at least one observation to the interference source with the highest calculated likelihood probability (hard decision embodiment); If the calculated likelihood probability is above the threshold, the source allocations are sampled until convergence (source sampling embodiment).

[0024] For the observation device, in one embodiment, the observation device can be moved at successive known positions while the interference source is assumed to be in a fixed position.

[0025] For example, the observation device may be installed on a moving vehicle such as a train with a known track, or may simply be equipped with a GPS to know its continuous current position.

[0026] Alternatively, the scanning of multiple radio frequency bands described above can be performed by at least three observation devices having known spatial positions (e.g., fixed positions), where the spatial positions are not aligned and source location determination can be performed by triangulation across the three observation devices.

[0027] The present disclosure may also be directed to an observation device comprising computer circuitry for carrying out the methods presented above.

[0028] The present disclosure is also directed to a system comprising at least three observation devices for carrying out the method.

[0029] The present disclosure is also directed to a computer program comprising instructions that, when executed by a computer, cause the computer to perform the method.

[0030] The present disclosure is also directed to a non-transitory computer storage medium storing instruction code for such a computer program. [Brief explanation of the drawings]

[0031] [Figure 1] FIG. 1 illustrates the main steps of the interference analysis process. [Figure 2a] FIG. 1 illustrates the configuration of a mobile observation device in the spatial domain. [Figure 2b] FIG. 1 illustrates the configuration of a mobile observation device in the frequency domain. [Figure 3a] FIG. 1 illustrates a configuration of multiple observation devices in the spatial domain. [Figure 3b] FIG. 1 illustrates a configuration of multiple observation devices in the frequency domain. [Figure 4a] FIG. 2 illustrates interference activity in the frequency domain of interference sources. [Figure 4b] FIG. 1 illustrates the true respective positions in the spatial domain. [Figure 5a] 1A and 1B illustrate a mobile device's observations in the frequency domain via frequency hopping and in the spatial domain via received power measurements, respectively. [Figure 5b]1A and 1B illustrate a mobile device's observations in the frequency domain via frequency hopping and in the spatial domain via received power measurements, respectively. [Figure 6a] FIG. 6 shows the results of the implementation of the method with a mobile device and five interference sources active in different channels (five different colors or grey levels in FIG. 6a). [Figure 6b] FIG. 10 illustrates the results of implementing the method with a mobile device and five interference sources active during different time periods. [Figure 6c] FIG. 10 illustrates the results of implementing the method with a mobile device and five interference sources for different positions of the mobile device. [Figure 6d] FIG. 10 shows the results of the implementation of the method with a mobile device and five interference sources with their respective positions finally determined. [Figure 7] FIG. 1 shows an interference analysis device for carrying out the method described above. DETAILED DESCRIPTION OF THE INVENTION

[0032] Further details and advantages of the present disclosure will be understood on reading the following description of embodiments given below as examples and will become apparent from the associated drawings, in which:

[0033] 1, given current observations I2 (mandatory) of the environment including interferers of the communication system and an optional database I1 (the database can be built from previous observations in previous steps of the method), the method proposes to analyze the interference in terms of frequency usage, geolocation characteristics and / or time usage. The interference analysis device performs observations at multiple frequencies, locations and times.

[0034] The first input of each processing step is an interference database. For each isolated current interference source, the database provides the following information in step I1: Location information that is one of the following: - the average position in the space under consideration, or - a probability map of the discretized positions in the space considered, Frequency information, which can be one of the following: -average frequency, or - a probability map of the frequency range considered, Related Observations: -Previous observations already assigned to the current interference source.

[0035] The "current observations" mentioned above may consist of the following in step I2: - the position of the observation device (the above-mentioned interferometric analysis device), -received interference power, and - Observed frequency channel index.

[0036] In practice, the interference analysis device scans a given frequency band over a number of radio frequency channels (for example, 16 channels in the example of the figures discussed below), each channel having an appropriate index.

[0037] Next, step S1 involves a membership calculation: using a database of interfering sources, the likelihood of the current observation of each source is calculated based on Bayesian inference. The calculation measures: the similarity between the level of power given by the observation and the closest level of power of the source given by the position of this source in the database, and -The degree of similarity between the frequency observation and the source frequency database. This probability reflects how strong the connection is between the observation and the source.

[0038] Next, step S2 involves observation classification, where observations are assigned to sources using the probabilities calculated in the previous step, step 1. This step S2 can be performed according to several embodiments: A "hard decision" embodiment in which the source to be allocated is determined to be the one with the highest probability, and the decision is made once; "Source sampling" embodiment, where source decisions are sampled based on their estimated probabilities. This technique requires iteration to allow the samples to converge to the estimated probabilities. "Hybrid method" embodiments in which both of the above methods are employed to balance the speed of hard decision methods with the stability of source sampling methods. Typically, a probability threshold can be defined, above which hard decisions can be made, below which source sampling is used rather.

[0039] Next, step S3 involves database updating: once an observation is associated with a source, the database of this source evolves according to the observation and is therefore updated accordingly.

[0040] Finally, in step O1, the newly updated database can be used in the next iteration of the process.

[0041] Details of each step of the general method are described below.

[0042] In an interfered radio environment, it is assumed that the interference is generated by one or several radio sources, one source being distinguished from another by its geometric location and / or operating frequency band.

[0043] To be able to monitor the environment, interference observations are required, for which the observations can be obtained by any of the following devices: - One device with observation capabilities at multiple frequencies and multiple locations: this device can be a wideband or frequency-hopping transceiver mounted on a moving vehicle (train, car, boat, etc.). An example of this setup is shown in Figure 2a, which shows the variations in the spatial domain, and in Figure 2b, which shows the variations in the frequency domain. - Or multiple devices with multiple operating frequencies (possibly wideband or frequency-hopping transceivers): these devices are placed in at least three non-aligned positions to allow triangulation in geolocation. For increased accuracy, four or more positions are preferred. Examples of device placement in space and frequency are shown in Figures 3a and 3b, respectively.

[0044] To the observing device, the interference transmissions are completely random since there is no signaling exchange with the interference source. In another example, the device observes the interference blindly without knowing the interference location. Therefore, the observed data can be considered as a mixed signal.

[0045] Hereafter, the following notation will be adopted: Interference source S k Regarding: -k is the index of the source, -θ k is the source position, -φ k is the operating frequency of the source. Observation value Z n Regarding: -n is the index of the observation, -T n is the position of the observation device, -W n is the received interference power, -F n is the observed frequency.

[0046] For the estimation of the received power, the observed value Z n At the time when k The received power (e.g., in dB) at the observing device can be modeled as:

number

[0047] where a and b are the two coefficients of the path loss model,

number

number

[0048] where ρ0, d c are the two coefficients of the shadowing model.

[0049] Interference source S k The operating frequency of can be modeled by two parameters. φ k =(f k ,B k )

[0050] where f k represents the starting frequency, and B k represents the bandwidth.

[0051] To be able to analyze interference, the observations must be classified into distinct sources. A database of each source can then be built and updated according to the observations it belongs to. This database then serves as prior knowledge to the next classification of observations.

[0052] From now on, V n is the observed value Z n represents a latent variable indicating the source to which the observed value Z is assigned. n At the start of the analysis, the following statements are assumed: ·(.) -n is the observed value Z n represents the existing set of previous variables with There are K interference sources: k=1..K. Sauce S k Regarding

number

number

number

[0053] The membership calculation in step S1 can be based on any of the following: a) Euclidean distance: average position

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

number

[0054] The probability of a frequency observation is

number

number

number

number

[0055] Next, for the implementation of step S2, the observation classification can then be based on either: a) Euclidean distance Observations are classified by hard decision according to their distance to the source, in this sense, an observation is associated with the source with the smallest distance, for example:

number

number

number

[0056] The database update in step S3 can be based on any of the following: a) Euclidean distance Observed value Z n Assuming that k is associated with source k, the average position and frequency can be updated as follows:

number

number

number

[0057] Figures 4a and 4b show an example of an interference environment: the interference activity in the frequency domain is shown in Figure 4a, and the respective geometric locations of the interferers are shown in Figure 4b.

[0058] In this interference environment, the mobile device observes the interference environment via a frequency hopping (FH) pattern, as generally presented in Figure 2b, for example. Details of such observations can be found in document European Patent Application No. 3716506.

[0059] The initial corresponding observations obtained for the mobile device are shown in Figure 5a (frequency domain) and Figure 5b (spatial domain).

[0060] In this example, Bayesian inference is used for membership calculations and hard decisions are used for classification to analyze the environment. The estimation is performed ab initio without an initial database.

[0061] The results of the observation classification are shown in Figures 6a and 6c in the frequency domain and the received power domain, respectively. In both figures, each label I1, I2, I3, I4, and I5 represents an interference source. As can be seen, the method embodiment allows estimating five interference sources and attributing observations to each of them. In Figure 6b, the estimates of operating frequency and activation time are plotted for each interference source. In Figure 6d, the contours of the source location probabilities resulting from Bayesian updating are displayed along with the true locations of the interference sources.

[0062] Thus, the present disclosure enables the use of frequency hopping systems, for example in the 2.4 GHz ISM band, to characterize interference along a given track, such as a railroad track, thereby enabling the reduction of the impact of such interference on track equipment and / or communication devices mounted on trains.

[0063] With reference to FIG. 7, a device DEV for implementing the above method may for example comprise: an input interface IN (e.g. a radio frequency antenna, etc.) for receiving radio frequency signals and measuring the radio frequency power in each frequency channel according to a frequency hopping pattern, for example as presented in FIG. 2b; a memory MEM storing at least the instructions of a computer program implementing the method presented above and possibly also storing data of a database DB used in the method; a processor PROC for implementing the method and for accessing said memory MEM and said database DB in order to finally determine the current interference sources (location and radio frequency active bands), an output interface OUT for delivering data of such decisions, which can be fed into the database DB in order to update the contents of this database DB; and optionally, in one embodiment where the device DEV is mobile, a GPS chip to determine the exact location of the device at each of its observations.

Claims

1. A method for characterizing radio frequency interference produced by multiple sources, comprising: The observation device sequentially scanning said plurality of radio frequency bands at different spatial locations and at different times to perform interference measurements (I2) in said plurality of radio frequency bands; Current observation frequency, the current location of the observation device; and the received interference power in the observed frequency band, providing observations at successive time points of Calculating (S1) for each interference source and each observation a likelihood probability of attributing an observation to an interference source; assigning each of said observations to one interference source using likelihood probability calculations (S2); for each interference source assigned one of the observations, estimating the location and frequency occupancy of the interference source; [0033] A method, wherein the likelihood probability calculation (S1) is performed based on both frequency observations and interference power observations.

2. 2. The method of claim 1, wherein the likelihood probability calculation is performed by using a database that stores, for each of the interference sources, at least data on the location of the source and data on the occupancy of at least two radio frequency bands by the source.

3. 3. The method according to claim 2, wherein the database is updated after each assignment (S2) of an observation to an interference source, taking into account the next iteration of the method.

4. For each interference source, the database includes: The historical average position, and a probability map of past discretized positions, 4. The method of claim 2 or 3, further comprising providing at least one of:

5. For each interference source, the database includes: Historical average radio frequency band occupancy, and Probability maps of past radio frequency range occupancy, 4. The method of claim 2 or 3, further comprising providing at least one of:

6. 4. The method according to claim 2 or 3, wherein the database stores, for each interference source, occurrence data of previous observations already assigned to the interference source.

7. For a current observation, the likelihood probability of one of the interference sources being associated with the current observation is calculated (S1) based on a comparison between data in the database and data of the provided observation, the comparison comprising: a similarity between the level of radio frequency power measured in the current observation and the closest level of radio frequency power corresponding to this source given the past location of the source in the database and the current location of the observation device; a similarity between the radio frequency band occupied by the current observation and the radio frequency band occupied by the source according to the database; 4. The method of claim 2 or 3, comprising determining:

8. The assignment (S2) of at least one observation to an interference source is sorting the likelihood probabilities calculated for each interference source to attribute the at least one observation to the interference source; selecting the highest calculated likelihood probability and assigning the at least one observation to the interference source having the highest calculated likelihood probability; The method according to any one of claims 1 to 3, comprising:

9. The likelihood probability is a distance between the observation device and a source in the radio frequency power domain; the distance between the frequency band occupied by said source according to the database and the frequency band in which interference is measured according to said observations; The method of claim 8, wherein the distance is calculated based on a Euclidean distance calculation of

10. The assignment (S2) of at least one observation to an interference source is sampling each source allocation based on an associated calculated likelihood probability, where the initial random draw is followed by calculation iterations until successive samples converge to the associated calculated likelihood probability; The method according to any one of claims 1 to 3, comprising:

11. Each likelihood probability is compared to a threshold value, if the calculated likelihood probability is above the threshold, selecting the highest calculated likelihood probability and assigning the at least one observation to the interference source having the highest calculated likelihood probability; The method of claim 10 , wherein if the calculated likelihood probability is below the threshold, the source assignments are sampled until convergence.

12. A method according to any one of claims 1 to 3, wherein the observation device moves in successive known positions while the interference source is assumed to be in a fixed position.

13. 4. The method of claim 1, wherein the scanning of the multiple radio frequency bands is performed by at least three observation devices having known spatial positions, the spatial positions being unaligned, and source location determination is performed by triangulation across the three observation devices.

14. The method according to any one of claims 1 to 3, wherein the likelihood probability calculation (S1) involves simultaneous Bayesian inference of frequency and interference power observations.

15. An observation device for carrying out the method according to any one of claims 1 to 3.

16. A system comprising at least three observation devices that perform the method according to any one of claims 1 to 3.

17. A computer program comprising instructions for causing a computer to carry out the method according to any one of claims 1 to 3.

Images