JP5282914B2 - Radio identification device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a radio device identification device that can accurately identify a radio device. <P>SOLUTION: A radio device identification device receives an electric wave emitted by a radio device and determines, as a feature vector, a signal obtained by differentiating an envelope detection signal at the time of rising of the signal by a differentiation processing part, and identifies an originating device based on the degree of similarity between the feature vector and a teacher vector that is a feature vector of a known radio device that has been obtained in advance. <P>COPYRIGHT: (C)2012,JPO&amp;INPIT

Description

  The present invention relates to a radio identification device that receives radio waves emitted from a radio and identifies a source radio.

  2. Description of the Related Art Conventionally, when identifying a wireless device that performs illegal wireless communication, a wireless device identification device using an envelope detection signal at the time of signal rise or a frequency variation as a feature vector is used. However, in a wireless device identification apparatus that uses an envelope detection signal at the time of signal rise as a feature vector, the feature vector fluctuates due to fading. For this reason, the reproducibility of feature vectors is poor and accurate identification is difficult. In addition, in the wireless device identification apparatus using the frequency variation as a feature vector, there is a problem that it is difficult to identify a wireless device with a small frequency variation or a wireless device with a similar frequency variation.

  For example, Patent Document 1 (Japanese Patent Laid-Open No. 2002-026826) discloses an apparatus for identifying a radio device characterized by a voltage time lag at the time of rising. In paragraphs 0025 to 0027 of Patent Document 1, a voltage time lag is measured using a threshold value normalized with a saturation voltage as a reference. However, in an actual wireless communication environment (fading condition), the signal waveform is as shown in FIG. Therefore, as shown in FIG. 8 of Patent Document 1, it is difficult to accurately obtain the saturation voltage, and it is difficult to obtain the voltage time lag with good reproducibility. As described above, since the problem of Patent Document 1 does not consider the envelope detection signal due to fading, it is difficult to obtain an accurate saturation voltage and it is difficult to obtain a voltage time lag with high reproducibility. There is a problem that it is difficult to identify the machine.

  Patent Document 2 (Japanese Patent Application Laid-Open No. 2006-211250) discloses a radio identification device using frequency fluctuation. However, there is a problem that it is difficult to identify radio devices with small frequency fluctuations or radio devices with similar frequency fluctuations, and new feature information that does not depend on frequency information is required.

JP 2002-026826 A JP 2006-211250 A

  As described above, in the case of a radio identification device using an envelope detection signal at the time of signal rise or a frequency variation as a feature vector, there is a problem that accurate radio identification is difficult.

  The present invention has been made to solve such a problem, and provides a wireless device identification apparatus capable of accurately identifying a wireless device.

  According to one aspect of the present invention, a differentiation processing unit that differentiates a normalized envelope detection signal at the time of rising of a received signal and converts the signal into a feature vector, the feature vector, and a previously acquired known vector A wireless device identification apparatus comprising: a wireless device identification processing unit that identifies a wireless device as a transmission source based on a similarity to a teacher vector of the wireless device.

  According to another aspect of the present invention, the differential processing unit performs differential processing on the normalized envelope detection signal at the time of rising of the received signal and converts it into a feature vector, and the radio identification processing unit includes the feature vector. And a step of identifying a transmitter radio from a similarity with a teacher vector of a known radio acquired in advance.

  The radio identification apparatus of the present invention generates a feature vector obtained by differentiating a normalized envelope detection signal obtained by normalizing the amplitude value of an envelope detection signal with an effective voltage value, and stores the generated feature vector. The wireless device is identified by comparing the similarity with the teacher vector. By performing the differential processing, the envelope detection signal due to fading can be suppressed, the rising envelope detection signal characteristic can be extracted, and the radio can be accurately identified.

It is a block diagram which shows the structure of the radio | wireless machine identification apparatus of this invention. It is the wave form diagram (A) and (B) which overwritten the envelope detection signal on the ideal conditions in the radio | wireless machines A and B. FIG. FIG. 3 is waveform diagrams (A) and (B) showing an envelope detection signal by simulation when noise and fading are added to the waveform of FIG. 2. FIGS. 4A and 4B are pattern diagrams showing feature vectors obtained by differentiating the envelope detection signal in the fading environment of FIG. It is a flowchart figure which shows the operation | movement which identifies the radio | wireless machine in this invention.

(First embodiment)
Embodiments of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram showing a configuration of a radio identification device of the present invention. 2A and 2B are waveform diagrams (A) and (B) in which envelope detection signals under ideal conditions in the radio devices A and B are overwritten. FIG. 3 is a waveform diagram (A) and (B) showing an envelope detection signal by simulation when noise and fading are added to the waveform of FIG. FIGS. 4A and 4B are pattern diagrams (A) and (B) showing feature vectors obtained by differentiating the envelope detection signal in the fading environment of FIG. FIG. 5 is a flowchart showing an operation for identifying a radio in the present invention.

  A radio identification device 1 of the present invention shown in FIG. 1 is connected to a receiver 3, and includes an A / D converter 4, a rising detector 5, an effective voltage measuring device 6, an envelope detector 7, a differential process. Unit 8, feature vector database (DB) 9, and wireless device identification processing unit 10. A signal from the wireless device received by the antenna 2 is converted to an intermediate frequency by the receiver 3. The A / D converter 4 converts this intermediate frequency signal into a digital complex envelope signal. When the rising detector 5 detects the rising of the signal from the digital complex envelope signal, it notifies the effective voltage measuring device 6 and the envelope detector 7 of a rising detection trigger.

  When the effective voltage measuring device 6 receives the detection trigger, the effective voltage measuring device 6 measures the average power of the digital complex envelope signal in a certain time range and notifies the envelope detector 7 of the square root as an effective voltage value. When receiving the detection trigger, the envelope detector 7 converts a digital complex envelope signal in a certain time range into an envelope detection signal, and normalizes the voltage amplitude value of the envelope detection signal divided by the effective voltage value. An envelope detection signal is generated. The purpose of normalization is to remove the influence of the transmission distance on the envelope value. The differentiation processing unit 8 differentiates the normalized normalized envelope detection signal and converts it into a feature vector.

  The feature vector DB 9 receives the feature vector and registers the feature vector that the user desires to register as a teacher vector representing known radio information. The feature vector DB 9 also performs editing processing such as updating / deleting / referencing of teacher vectors based on user instructions. The wireless device identification processing unit 10 compares and collates the similarity between the unknown feature vector received from the differentiation processing unit 8 and the teacher vector received from the feature vector DB 9, and determines the teacher vector having the highest similarity as the wireless device of the received signal. Identify as The similarity is obtained by a method such as signal correlation or Euclidean distance.

  The generation of feature vectors will be described in detail. If the in-phase component of the digital complex envelope signal is I (n) and the quadrature component is Q (n), it is converted into an envelope detection signal ρ (n) normalized by Equation 1.

  Next, the feature vector f (n) is obtained by differentiation of the envelope detection signal (in the present invention, a difference is used for a digital signal), and is expressed by Equation 2. T is a sample period (reciprocal of sampling frequency).

  FIGS. 2A and 2B are waveforms obtained by overwriting 30 envelope detection signals at the time of start-up of the 430 MHz band radio device A and the 430 MHz band radio device B acquired under ideal conditions. Acquisition under ideal conditions is a measurement result that eliminates the effects of noise and fading by measuring the radio and receiver connected by a coaxial cable. In this example, the envelope detection signal pattern is drawn with good reproducibility for both the radio device A and the radio device B, and the envelope detection signal pattern is clearly different between the radio device A and the radio device B. From this, it is shown that the wireless device A and the wireless device B can be easily distinguished and identified by using the envelope detection signal at the time of startup under ideal conditions. 2, (A) and (B) in FIGS. 2, 3 and 4 shown below correspond to the wireless device A and the wireless device B, respectively.

  However, the influence of noise and fading is unavoidable in an actual wireless communication environment. 3A and 3B show the envelope detection signal patterns when noise and fading are added to the envelope detection signals of FIGS. 2A and 2B by simulation. Compared to FIG. 2, in FIG. 3, the reproducibility of the envelope detection signal pattern is clearly deteriorated in both the wireless device A and the wireless device B, and the difference in the envelope detection signal pattern between the wireless device A and the wireless device B Is getting smaller. This indicates that in an actual wireless communication environment, it is difficult to identify and identify the wireless device A and the wireless device B using the envelope detection signal at the time of rising. In other words, if it is an ideal condition, the wireless device can be identified even with the prior art. However, in an actual wireless communication environment, it can be understood that it is difficult to identify a wireless device as described above as the subject of the present invention.

  3 is a rice fading environment with an S / N ratio of 30 dB, a Doppler frequency of 40 Hz, and a Rice coefficient of 6.5 dB. The Doppler frequency 40 Hz is a value when a 430 MHz band radio device is moving at about 100 km / h. Measurement results of the differential (ie, feature vector) pattern of the envelope detection signal obtained by performing differentiation processing, which is an important component of the present invention, with respect to the envelope detection signal in the fading environment of FIGS. 3 (A) and 3 (B). Is shown in FIGS. 4 (A) and 4 (B). In FIG. 4, the fluctuation caused by fading that appeared in FIG. Is different. Therefore, in the present invention, even in an actual wireless communication environment, the wireless device A and the wireless device B can be easily identified by using the differentiation of the envelope detection signal at the time of rising.

  The differentiation process has a frequency characteristic of j2πf with respect to the frequency f, and plays the role of a high-frequency pass filter. Generally, the fading frequency is sufficiently smaller than the frequency of the envelope detection signal at the time of rising. In other words, the envelope detection signal due to fading changes more slowly than the rising envelope detection signal. For this reason, the fluctuation due to the fading of the low frequency component is suppressed by the differentiation process, and the envelope detection signal characteristic at the rising of the high frequency component remains.

  Next, the result of quantitatively evaluating the identification process using this feature vector is shown. Table 1 shows the result of measuring the signal correlation of the wireless device A and the wireless device B with respect to the teaching vector of 21 test wireless devices (430 MHz band) including the wireless device A and the wireless device B described above. The teacher vector is obtained by averaging 20 wave feature vectors, and the evaluation data uses 30 wave feature vectors obtained separately from the teacher vector. Both the teacher vector and the feature vector used for the evaluation data are obtained under a rice fading condition with an S / N ratio of 30 dB, a Doppler frequency of 40 Hz, and a Rice coefficient of 6.5 dB by simulating an actual wireless communication environment.

  In Table 1, bold italic letters indicate that the signal correlation is closest to 1 and indicates that the teacher vector and the evaluation data are the same wireless device. As is clear from the results in Table 1, both the wireless device A and the wireless device B showed the highest degree of similarity at the correct answer, and the respective signal correlations of 0.8332 and 0.7719 were obtained. This evaluation result uses the differential of the envelope detection signal as a feature vector, so that the wireless device A and the wireless device B can be easily identified and identified by calculating the similarity in the actual wireless communication environment. Is quantitatively shown. When signal correlation measurements were performed on all test radios (21 units) under the same conditions as in Table 1, a signal correlation value close to 1 was obtained, which was the highest similarity when 20 radio units were correct. It was. In this way, each wireless device can be identified with high probability.

  Next, the operation of the radio identification apparatus of the present invention shown in FIG. 1 will be described in detail with reference to the flowchart of FIG.

  As an operation of the wireless device identification apparatus, the A / D converter 4 converts the intermediate frequency signal input from the receiver 3 into a digital complex envelope signal and outputs it (step S1). The rise detector 5 repeatedly detects the rise of the signal with respect to the digital complex envelope signal received from the A / D converter 4, and if detected, sets a detection trigger to the effective voltage measuring device 6 and the envelope detector 7. Notice. When receiving the detection trigger, the effective voltage measuring device 6 measures the signal effective voltage within a certain time range from the rising edge (step S2). When receiving the detection trigger, the envelope detector 7 converts the digital complex envelope signal into an envelope detection signal (step S3). Further, the amplitude value of the envelope detection signal is normalized with the effective voltage measured by the effective voltage measuring device 6 to generate a normalized envelope detection signal (step S4).

  The differentiation processing unit 8 differentiates the normalized envelope detection signal to generate a feature vector (step S5). The wireless device identification processing unit 10 compares the similarity between the teacher vector stored in the feature vector DB 9 and the feature vector generated in step S5 (step S6), and determines the teacher vector having the highest similarity to the wireless device. The identification result is output (step S7). In parallel with steps S6 to S7, when the user instructs to register the feature vector acquired as a new teacher vector, the feature vector is registered in the feature vector DB 9 as a teacher vector (step S8). In step S8, as described above, a plurality of feature vectors acquired from the same radio device may be averaged to be a teacher vector. When averaged, the S / N of the teacher vector is improved, so that the identification rate can be improved.

  The radio | wireless machine identification apparatus of this invention receives the electromagnetic wave emitted from a radio | wireless machine, and makes the feature vector the signal which differentiated the envelope detection signal at the time of a signal rise (the moment of OFF to ON) by a differentiation process part. It is characterized in that the transmitter radio is identified from the similarity between this feature vector and a teacher vector, which is a feature vector of a known radio acquired in advance. By differentiating the envelope detection signal, the influence of fading can be removed, and a feature vector with good reproducibility can be obtained even with a mobile radio, so that the performance of identifying the radio is improved. In addition, by using the differential of the envelope detection signal as a feature vector, it is possible to provide an effective new radio identification technique for a radio with little frequency fluctuation or a radio with similar frequency fluctuation.

  The first effect of the present invention is that a practical wireless device identification apparatus using a feature vector obtained by differentiating an envelope detection signal at the time of rising can be provided. The reason is that the envelope detection signal due to fading can be suppressed by differential processing, and the rising envelope detection signal characteristic can be extracted with good reproducibility. The second effect is that, in the conventional radio equipment identification device using frequency fluctuation as a feature, it is possible to provide a new feature effective for identification with respect to radio equipment with low frequency fluctuation that is difficult to identify or radio equipment with similar frequency fluctuation. It can be provided. The reason is that the frequency and the envelope are orthogonal and do not depend on each other, and therefore the envelope detection signal is not necessarily similar even if the frequency variation pattern is similar. The third effect is that it is economical because it can be configured by using a conventional radio identification device. The reason is that the configuration of the DB and the identification processing unit can be used as they are, and the apparatus can be realized only by adding a differential processing unit and an effective voltage measuring device.

  As mentioned above, although this invention was demonstrated in detail as preferable embodiment, this invention is not limited to these embodiment, In the range which does not deviate from the summary of this invention, an appropriate change is possible. .

  As described above, the wireless device identification apparatus of the present invention identifies the signal obtained by differentiating the envelope detection signal by the differentiation processing unit, and identifies the transmitter wireless device based on the similarity between the feature vector and the teacher vector. To do. Since the device can be realized simply by adding a differential processing unit and an effective voltage measuring instrument to the current wireless device identification device, it is useful and highly available in the industry, and monitors radio waves emitted from illegal wireless stations. Therefore, it can be applied to an application such as a radio wave monitoring device.

DESCRIPTION OF SYMBOLS 1 Radio equipment identification apparatus 2 Antenna 3 Receiver 4 A / D converter 5 Rising detector 6 Effective voltage measuring instrument 7 Envelope detector 8 Differentiation processing part 9 Feature vector database (DB)
10 Radio identification processing unit

Claims (6)

Differentiating the normalized envelope detection signal at the rising edge of the received signal to convert it into a feature vector, and the similarity between the feature vector and a previously acquired radio equipment teacher vector And a wireless device identification processing unit for individually identifying a wireless device as a transmission source. Comprising an A / D converter for converting the signal received by the receiver into digital complex envelope signal, and the effective voltage measuring device for measuring the effective voltage value of the digital complex envelope signal, and envelope detector, the said envelope A line detector converts the digital complex envelope signal into an envelope detection signal, and generates the normalized envelope detection signal by dividing the amplitude value of the envelope detection signal by the effective voltage value. The radio | wireless machine identification apparatus of Claim 1.   3. The teacher vector according to claim 1, wherein an average value of a plurality of feature vectors acquired in advance for a specific wireless device is registered as a feature vector of the specific wireless device. Radio identification device. Detecting a rising edge of the signal from the digital complex envelope signals, said by the rise detector you Outputs a rising edge detection trigger and the effective voltage measuring device and the envelope detector comprises a receive said detection trigger An effective voltage measuring device measures an effective voltage value of the digital complex envelope signal, the envelope detector converts the digital complex envelope signal into an envelope detected signal, and an amplitude value of the envelope detected signal is converted into the effective voltage. 3. The radio identification apparatus according to claim 2, wherein the normalized envelope detection signal is generated by dividing by a value. The differential processing unit performs differential processing on the normalized envelope detection signal at the time of rising of the received signal and converts it into a feature vector, and the radio equipment identification processing unit acquires the feature vector and a previously acquired known vector. And a step of individually identifying a transmitter radio from a similarity to a teacher vector of the radio. Before the step of converting into the feature vector, step converting the received signal A / D converter receives the digital complex envelope signal, the effective voltage measuring device for measuring the effective voltage value of the digital complex envelope signal When the steps of envelope detector converts the digital complex envelope signal to envelope detection signal, to generate the normalized envelope detection signal amplitude value by dividing by the effective voltage value of the envelope detection signal The radio identification method according to claim 5, further comprising:

Images