JP5958453B2 - Radio station identification device, radio station identification method, and radio station identification program - Google Patents

Description

  The present invention relates to a radio station identification device, a radio station identification method, and a radio station identification program, and more particularly to a radio station identification device, a radio station identification method, and a radio station identification that enable radio stations to be identified at high speed and with high accuracy. Regarding the program.

  In a radio station identification device for identifying a radio station, every time a radio signal from a radio station is normally received and a feature waveform vector of the radio signal is detected, all teacher data to be identified that are stored in advance in a database Similarity calculation is performed with (waveform vector that can specify a radio station), and teacher data with high similarity is output as an identification result. Therefore, it is necessary to repeat the similarity calculation as many times as the number of target teacher data. In order to output the identification result in a short time, it is important to narrow down the number of target teacher data to be subjected to similarity calculation in advance. Become.

  However, depending on the parameters used for the index and the fineness to narrow down, radio stations may be missed, and the identification accuracy may deteriorate, so the number of teacher data can be narrowed down finely without degrading the identification accuracy. The selection of possible parameters has been an issue.

  For this reason, for example, in Japanese Patent No. 4760967 of "Patent Document 1," "Radio Station Identification Device, Radio Station Identification Method, Radio Station Identification Program", a feature vector waveform in an unknown radio station and a known registered in a database When calculating the similarity with the feature vector waveform (teacher data) in the radio station, the identification target teacher data is narrowed down from the database using the reception frequency and the measurement time as an index, and the similarity is calculated for the identification target teacher data. By doing so, a technology that makes it possible to improve the identification accuracy and speed up the identification process is proposed.

Japanese Patent No. 4760967 (page 9-11)

  However, when narrowing down according to the measurement time as in Patent Document 1, it is possible to limit the identification target to teacher data within the past several years in order to eliminate the influence of aging degradation of radio stations. There is a problem that it is difficult to narrow down the details in detail.

  Also, with regard to narrowing down by reception frequency, multiple frequency channels can be selected even for the same radio station, so if you finely set the reception frequency to narrow down, you will miss radio stations that appeared in different channels. There is a problem that there is a limit to narrowing down using only the reception frequency.

(Object of the present invention)
The present invention has been made in order to solve the above-described problems, and enables radio station identification apparatus and radio station that can fine-tune teacher data and identify radio stations at high speed and with high accuracy. An object of the present invention is to provide an identification method and a radio station identification program.

  In order to solve the above-described problems, a radio station identification device, a radio station identification method, and a radio station identification program according to the present invention mainly adopt the following characteristic configuration.

  (1) A radio station identification device according to the present invention is a database in which a feature vector waveform indicated by a frequency variation waveform at the time of rising or falling of a signal obtained by converting a radio signal from each radio station is registered in advance as teacher data. A radio station identification device that performs processing for identifying a radio station and includes a reception unit for identifying a radio station that is transmitting a received radio signal when receiving the radio signal from the radio station to be identified And the refinement parameter generated from the feature vector waveform in the wireless signal is used as an index, and the teacher data to be collated is refined and extracted from the teacher data registered in the database unit The feature vector waveform of the radio signal received from the radio station to be identified is sequentially compared with the extracted teacher data. To, based on the extracted result of calculating the waveform similarity between the teacher data, and identifying the radio station to be identified.

  (2) The radio station identification method according to the present invention is a database in which a feature vector waveform indicated by a frequency fluctuation waveform at the time of rising or falling of a signal obtained by converting a radio signal from each radio station is registered in advance as teacher data. A radio station identification method for performing a process of identifying a radio station using a unit, wherein when a radio station transmitting a received radio signal is identified, the reception frequency of the radio signal from the radio station to be identified And the refinement parameter generated from the feature vector waveform in the wireless signal is used as an index, and the teacher data to be collated is refined and extracted from the teacher data registered in the database unit The feature vector waveform of the radio signal received from the radio station to be identified is sequentially compared with the extracted teacher data. To, based on the extracted result of calculating the waveform similarity between the teacher data, and identifying the radio station to be identified.

  (3) A radio station identification program according to the present invention is characterized in that the radio station identification method described in (2) is implemented as a program executable by a computer.

  According to the radio station identification device, radio station identification method, and radio station identification program of the present invention, the following effects can be obtained.

  The first effect is to use the index of the reception frequency and the narrowing-down parameter to identify the degree of similarity with the radio signal (feature vector waveform) output from the radio station from the teacher data in the database unit. That is, the target teacher data can be narrowed down. Thus, even when a large amount of teacher data is registered in the database unit, the wireless station identification process can be performed in a short time and with high accuracy.

  The second effect is to calculate the similarity to the radio signal (feature vector waveform) output by the radio station by narrowing down the teacher data to be identified by using the index of the reception frequency and the narrowing parameter. The processing time is shortened, and furthermore, the load on the calculation process can be reduced. Thus, an inexpensive computer can be used for wireless station identification processing, and an inexpensive wireless station identification device can be realized.

It is a block block diagram which shows an example of an internal structure of the radio station identification device by this invention. It is a wave form diagram which shows an example of the feature vector waveform extracted in the feature vector waveform extraction part. It is explanatory drawing which shows an example of the data structure of the teacher data stored in a database part. It is explanatory drawing which shows the example of registration of the teacher data of the database part in a certain receiving frequency and a narrowing down parameter.

  Preferred embodiments of a radio station identification device, radio station identification method, and radio station identification program according to the present invention will be described below with reference to the accompanying drawings. In the following description, the radio station identification apparatus and radio station identification method according to the present invention will be described. However, the radio station identification method may be implemented as a radio station identification program that can be executed by a computer. Alternatively, it goes without saying that the wireless station identification program may be recorded on a computer-readable recording medium. In addition, it is needless to say that the drawing reference numerals attached to the following drawings are added for convenience to the respective elements as an example for facilitating understanding, and are not intended to limit the present invention to the illustrated embodiments. Yes.

(Features of the present invention)
Prior to the description of the embodiments of the present invention, an outline of the features of the present invention will be described first. The present invention relates to a frequency fluctuation waveform (hereinafter, referred to as “OFF → ON switching instant”) or a signal falling edge (ON → OFF switching instant) of a signal obtained by converting a radio signal received from a radio station. In a wireless station identification apparatus that includes a database unit registered as a feature (referred to as a feature vector waveform) and performs a process of identifying a wireless station, by narrowing down teacher data according to the feature vector waveform, Even when data is registered, the main feature is that the identification result of the wireless station can be output accurately in a short time.

  More specifically, the present invention relates to a database unit that includes a reception frequency of a radio signal received from a radio station, and a narrowing parameter such as a waveform average energy and a waveform average envelope calculated from a feature vector waveform related to the radio signal. It is possible to narrow down the identification target teacher data from the database part in more detail, and to calculate the similarity only for the identification target teacher data. Even when the teacher data is registered, the identification result of the wireless station can be output quickly and with high accuracy.

(Configuration example of embodiment)
Next, a configuration example of the radio station identification device according to the present invention will be described in detail with reference to FIG. Here, FIG. 1 is a block diagram showing an example of the internal configuration of the radio station identification apparatus according to the present invention. As an example of the present invention, a radio station identification device 1 shown in FIG. 1 includes an antenna 2, a receiver 3, an A / D converter 4, a feature vector waveform extraction unit 5, a refinement parameter calculation unit 6, a feature data generation unit 7, a radio The station identification processing unit 8, the database unit 9, and the teacher data generation unit 10 are included at least.

  The antenna 2 is a part that acquires a signal of a specific frequency set in advance transmitted from a radio station as a received signal. The receiver 3 detects the reception frequency of the reception signal acquired by the antenna 2, converts the reception frequency of the reception signal into an intermediate frequency signal, and converts the converted intermediate frequency signal into the A / D converter 4 and characteristics. This is a part to be transmitted to the data generation unit 7. The A / D converter 4 converts the intermediate frequency signal received from the receiver 3 into a digital complex envelope signal that is a digital signal, and transmits the converted digital complex envelope signal to the feature vector waveform extraction unit 5 It is.

  The feature vector waveform extraction unit 5 analyzes the digital complex envelope signal received from the A / D converter 4 and at the time of rising of the digital complex envelope signal (that is, the timing of switching from the OFF state to the ON state) or the falling A part that extracts a fluctuation waveform of the frequency at the time (that is, the timing at which the ON state is switched to the OFF state) as a feature vector waveform, and transmits the extracted feature vector waveform to the refinement parameter calculation unit 6 and the feature data generation unit 7 It is. Note that the feature vector waveform is a waveform in which the center frequency at which the frequency variation converges is offset-corrected to 0 Hz.

The refinement parameter calculation unit 6 is a part that analyzes the feature vector waveform received from the feature vector waveform extraction unit 5 to calculate a refinement parameter, and transmits the computed refinement parameter to the feature data generation unit 7. FIG. 2 is a waveform diagram showing an example of a feature vector waveform extracted by the feature vector waveform extraction unit 5, where the horizontal axis represents time and the vertical axis represents frequency. As shown in FIG. 2, when the feature vector waveform 11 is divided into m sections (m: positive integer) on the time axis, and the frequency value in the i-th section (1 ≦ i ≦ m) is _{f_i} , As the frequency of the feature vector waveform 11 in each section, the frequencies of _{f_1} , _{f_2} , _{f_3} , ..., _{f_m} are acquired.

  Here, for example, a waveform average energy or a waveform average envelope can be used as the narrowing parameter. The waveform average energy is given by Equation 1 below. That is, the waveform average energy is calculated as a square value of the frequency in each section obtained by dividing the feature vector waveform 11 into m sections (m: positive integer) on the time axis, and the square value is averaged over m sections. Given by value.

On the other hand, the waveform average envelope is given by the following equation 2. In Expression 2, the envelope value of the square value of the frequency in the i-th section is indicated as _{A_i} . As shown in Equation 2, the waveform average envelope is the average of the envelope of the square of the frequency in each section obtained by dividing the feature vector waveform 11 into m sections (m: positive integer) on the time axis. Is given as a value.

  The feature data generation unit 7 uses the frequency indicated by the intermediate frequency signal received from the receiver 3 as the reception frequency of the feature vector waveform received from the feature vector waveform extraction unit 5, the reception frequency of the feature vector waveform, and the narrowing parameters This is a part that generates feature data obtained by combining the refinement parameter received from the calculation unit 6 with the feature vector waveform and transmits the feature data to the radio station identification processing unit 8 and the teacher data generation unit 10. Here, the feature data means signal data including a reception frequency, a narrowing parameter, and a feature vector waveform as described above. For this reason, when receiving the feature data, the radio station identification processing unit 8 can limit the identification target (range) of the teacher data stored in the database unit 9 using the reception frequency and the narrowing parameters as keys. . Thus, the wireless station identification process can be speeded up, and the wireless station identification accuracy can be improved.

  Upon receiving the feature data from the feature data generation unit 7, the radio station identification processing unit 8 transmits the reception frequency included in the feature data and the narrowing parameter, for example, the waveform average energy, to the database unit 9 as keys, and the database unit 9, the teacher data for similarity calculation corresponding to the key is received as data to be collated with the feature vector waveform included in the feature data. Then, the radio station is identified from the waveform similarity obtained as a result of matching between the feature vector waveform included in the feature data received from the feature data generation unit 7 and the teacher vector waveform of the teacher data received from the database unit 9, The result is output.

The database unit 9 registers the teacher data received from the teacher data generation unit 10. The teacher data is stored as a database of teacher vector waveforms received from the teacher data generation unit 10 using the reception frequency and the narrowing parameters as indexes. FIG. 3 is an explanatory diagram showing an example of the data structure of the teacher data stored in the database unit 9. That is, as shown in FIG. 3, the teacher data stored in the database unit 9 has a tree structure including a reception frequency 21, a narrowing parameter 22, and a teacher vector waveform 23, and an index f of a certain reception frequency 21. _The teacher data included in _{_k} is classified according to the size of the narrowing-down parameter 22 (for example, three of E_1, E_2, and E_3), and the corresponding teacher vector waveform 23 is stored as teacher data under that.

  When the database unit 9 receives the reception frequency and the refinement parameter of the feature data from the radio station identification processing unit 8, the database unit 9 searches the tree structure index using the received reception frequency and the refinement parameter as a key, and is stored as a database. Only the corresponding teacher data (teacher vector waveform) is extracted as the identification target from the teacher data, and is transmitted to the radio station identification processing unit 8. As described above, in addition to the method of determining the index classification for each refinement parameter in advance, it is included in the range of ± X (X: arbitrary real number) of the refinement parameter of the feature vector waveform in an unknown radio station. It is also possible to search the teacher data (teacher vector waveform) to be searched and extract the search result as an identification target. Here, X can be given as a constant or a variable such as a function of a narrowing parameter.

  When the teacher data generation unit 10 receives the feature data from the feature data generation unit 7, the teacher data generation unit 10 receives a registration instruction from the outside (for example, a measurer who is an operator of the radio station identification device 1). Teacher data is generated by processing a plurality of feature data accumulated in advance for each. With respect to generation of teacher data, a teacher vector waveform is generated as teacher data by performing a process of averaging feature vector waveforms of a plurality of accumulated feature data, and the generated teacher data is stored in the database unit 9. Sent. As described above, the S / N ratio in the teacher data can be improved by performing a process of averaging feature vector waveforms of a plurality of feature data accumulated with respect to a radio signal from a certain radio station.

(Description of operation of embodiment)
Next, an example of the operation of the wireless station identification device 1 shown in FIG. 1 will be described in detail. First, when the receiver 3 receives a radio signal from a radio station via the antenna 2, the reception frequency of the received radio signal is detected, and the A / D converter 4 is combined with an intermediate frequency signal obtained by converting the radio signal. And transmitted to the feature data generation unit 7. The A / D converter 4 converts the intermediate frequency signal received from the receiver 3 into a digital signal and transmits the digital signal to the feature vector waveform extraction unit 5.

  Next, the feature vector waveform extraction unit 5 analyzes the digital signal received from the A / D converter 4 and at the time of rising (OFF → ON switching) or falling (ON → OFF switching) of the digital signal. The instantaneous frequency fluctuation waveform is extracted as the feature vector waveform of the wireless station and transmitted to the refinement parameter calculation unit 6 and the feature data generation unit 7. The refinement parameter calculation unit 6 calculates a refinement parameter, for example, waveform average energy, of the received feature vector waveform, and transmits it to the feature data generation unit 7 together with the feature vector waveform.

  The feature data generation unit 7 generates feature data by combining the reception frequency received from the receiver 3 and the refinement parameter received from the refinement parameter calculation unit 6 with the feature vector waveform received from the feature vector waveform extraction unit 5. And transmitted to the radio station identification processing unit 8 and the teacher data generation unit 10.

Next, the radio station identification processing unit 8 performs an index search of the database unit 9 using the reception frequency included in the feature data and a narrowing parameter such as waveform average energy as keys, and the corresponding teacher data is used as identification target teacher data. Extract. FIG. 4 shows a registration example of teacher data in the database unit 9 at a certain reception frequency and narrowing parameters. In the example shown in FIG. 4, the index _{f _k} certain received frequency 21, E_1 sequentially small size of the narrowing parameter 22, E_2, teacher vector display 23 when indexed for E_3 as teacher data Registered. Here, as illustrated in FIG. 4, it can be seen that the size of the narrow-down parameter 22 corresponds to the degree of frequency variation of the teacher vector waveform 23.

  The radio station identification processing unit 8 is narrowed down and extracted by indexing the feature vector waveform included in the feature data received from the feature data generation unit 7 and the reception frequency and the narrowing parameters from the database unit 9 as keys. Sequentially comparing the teacher vector waveforms with each other, calculating the waveform similarity between the two, and identifying the radio station transmitting the radio signal of the reception frequency included in the feature data based on the waveform similarity, The identified result is output as a wireless station identification result.

  In addition, as a result of evaluating the effect of the index of the reception frequency and the narrowing parameter by simulation for a plurality of radio stations, the ratio of the number of teacher data extracted as a comparison target with respect to all the teacher data registered in the database unit 9 It was found that can be narrowed down to several tens of percent. In addition, as a result of evaluating the rate at which teacher data of the same radio station could not be extracted, that is, the miss rate of radio stations, it was found that the miss rate could be suppressed to about several percent.

  Therefore, by using not only the reception frequency but also the narrowing parameters including the waveform average energy and the waveform average envelope related to the feature vector waveform for the index, the number of identification targets of the teacher data can be greatly reduced. In addition, it has been found that since the frequency of missed radio stations to be identified can be suppressed to a very low level, radio station identification can be realized quickly and with high identification accuracy. That is, the narrowing-down parameter is an effective index for narrowing down the identification target of the teacher data. As the number of types of teacher vector waveforms increases, the narrowing-down effect by the index using the narrowing-down parameter increases.

  If the teacher data narrowed down and extracted from the database unit 9 is set on the memory instead of the hard disk, the I / O operation with the hard disk can be reduced, and the wireless station It is also possible to further speed up the identification.

(Explanation of effect of embodiment)
As described in detail above, the following effects are obtained in the present embodiment.

  The first effect is to collate the similarity with the radio signal (feature vector waveform) output from the radio station from the teacher data in the database unit 9 by using the index of the reception frequency and the narrowing-down parameter. It is possible to narrow down the teacher data to be collated. Thus, even when a large amount of teacher data is registered in the database unit 9, the wireless station identification process can be performed in a short time and with high accuracy.

  The second effect is to collate the similarity with the radio signal (feature vector waveform) output by the radio station by narrowing down the teacher data to be collated by using the index of the reception frequency and the narrowing parameter. The processing time is shortened, and furthermore, the load on the collation process can be reduced. Thus, an inexpensive computer can be used for wireless station identification processing, and an inexpensive wireless station identification device can be realized.

  The present invention can be suitably applied to uses such as a radio wave monitoring device for automatically monitoring radio waves emitted from illegal radio stations and the like.

  The configuration of the preferred embodiment of the present invention has been described above. However, it should be noted that such embodiments are merely examples of the present invention and do not limit the present invention in any way. Those skilled in the art will readily understand that various modifications and changes can be made according to a specific application without departing from the gist of the present invention.

DESCRIPTION OF SYMBOLS 1 Radio station identification device 2 Antenna 3 Receiver 4 A / D converter 5 Feature vector waveform extraction part 6 Narrowing parameter calculation part 7 Feature data generation part 8 Radio station identification processing part 9 Database part 10 Teacher data generation part 11 Feature vector waveform 21 Reception frequency 22 Refinement parameter 23 Teacher vector waveform

Claims (8)

A process of identifying a radio station is performed by including a database unit in which a feature vector waveform indicated by a frequency variation waveform at the time of rising or falling of a signal obtained by converting a radio signal from each radio station is registered as teacher data. A wireless station identification device,
When identifying a radio station that is transmitting a received radio signal, an index is given to the reception frequency of the radio signal from the radio station to be identified and the narrowing parameter generated from the feature vector waveform in the radio signal. By using, the teacher data to be collated is extracted from the teacher data registered in the database unit,
Based on the result of calculating the waveform similarity with the extracted teacher data by sequentially collating the feature vector waveform of the radio signal received from the identification target radio station with the extracted teacher data. Identify the station,
When indexing with the refinement parameter and extracting the teacher data to be collated from the database part, the index division of the database part is within a range of ± X (X: any real number) of the refinement parameter. Set the included teacher data as the search target ,
The size of the narrowing-down parameter corresponds to the degree of frequency fluctuation of the teacher vector waveform corresponding to the teacher data .   As the narrowing parameter, the square of the frequency in each section obtained by dividing the feature vector waveform of the radio signal received from the radio station to be identified into m sections (m: positive integer) on the time axis is averaged over m sections. 2. The radio station identification apparatus according to claim 1, wherein a waveform average energy given by the obtained value is used.   As the narrowing-down parameter, the envelope value of the square value of the frequency in each section obtained by dividing the feature vector waveform of the radio signal received from the radio station to be identified into m sections (m: positive integer) on the time axis. 2. The radio station identification apparatus according to claim 1, wherein a waveform average envelope given by a value averaged over m sections is used.   The teacher data registered in the database section is a vector waveform generated by averaging a plurality of the feature vector waveforms obtained from a radio signal received from the radio station. 4. The radio station identification device according to any one of 3. Performs processing for identifying a radio station using a database unit in which a feature vector waveform indicated by a frequency variation waveform at the time of rising or falling of a signal obtained by converting a radio signal from each radio station is registered as teacher data A radio station identification method for generating a radio station transmitting a received radio signal from a reception frequency of the radio signal from a radio station to be identified and the feature vector waveform in the radio signal By using the refined parameter as an index, a radio signal received from an identification target radio station by extracting and extracting teacher data to be collated from the teacher data registered in the database unit The feature vector waveform is sequentially compared with the extracted teacher data, and the waveform with the extracted teacher data is Based on the result of calculating the similarity, when the identification target radio station is identified and indexed by the refinement parameter and the teacher data to be verified is extracted from the database unit, the index classification of the database unit Is set to search for teacher data included in a range of ± X (X: any real number) of the narrowing parameters ,
The size of the narrowing-down parameter corresponds to the degree of frequency variation of the teacher vector waveform corresponding to the teacher data .   As the narrowing parameter, the square of the frequency in each section obtained by dividing the feature vector waveform of the radio signal received from the radio station to be identified into m sections (m: positive integer) on the time axis is averaged over m sections. 6. The radio station identification method according to claim 5, wherein the waveform average energy given by the calculated value is used.   As the narrowing-down parameter, the envelope value of the square value of the frequency in each section obtained by dividing the feature vector waveform of the radio signal received from the radio station to be identified into m sections (m: positive integer) on the time axis. 6. The radio station identification method according to claim 5, wherein a waveform average envelope given by a value averaged over m sections is used.   8. A radio station identification program, wherein the radio station identification method according to claim 5 is implemented as a program executable by a computer.

Images