JP4411267B2 - Call recording processor - Google Patents

Description

  The present invention relates to a call recording processing apparatus that is connected to a telephone and executes a recording process of the telephone call.

  2. Description of the Related Art Conventionally, a recording apparatus that monitors a voice signal during a call from a telephone line and records the voice signal as an analog signal or converted into a digital signal is known (for example, the following non-patent document). 1-3). In such a recording apparatus, it is necessary to detect the start and end of a call in order to record a series of audio signals for each call and not record the audio signal when the call is not in progress.

For example, in Non-Patent Document 1 below, the start of a call is determined by detecting a dial tone or conversation from a voice signal, and the end of the call is determined by detecting a no-recording state after hanging up. It is described. Non-Patent Document 2 described below describes that recording is started when the handset is raised by detecting the side sound of the telephone, and recording is stopped when the handset is lowered. On the other hand, Non-Patent Document 3 described below discloses a recording device in which a hook sensor is built in a telephone and the start of recording is controlled according to the detection of the hook sensor.
"Windows compatible USB call recording system My Logger", [online], System Kei Co., Ltd. [Search October 28, 2005], Internet <URL: http://www.systemk.co.jp/products /sapfone/index_2.html> “Easy and convenient continuous recording system“ Senran Plus ””, [online], Technology Link, Inc., [October 28, 2005 search], Internet <URL: http://park7.wakwak.com /~tlc/kinou.htm> "Network-compatible long-time recording system", [online], Sophia Systems, Inc., [October 28, 2005 search], Internet <URL: http://www.sophia-systems.co.jp/nvr/>

  As described above, when the voltage level of a predetermined tone such as a dial tone or conversation is detected from an audio signal, if the audio is interrupted or the audio is too low during the call, the call is started and ended accurately. May not be detected. In addition, when detecting the side sound of the phone, if a phone having a headset is used, the internal circuit may be constantly energized even during on-hook, so the ambient noise collected by the microphone is detected as a side sound, There was a tendency to be falsely detected as being on a regular call. On the other hand, when a sensor for detecting the state of the hook switch is provided on the telephone, there is a problem that the structure of the telephone becomes complicated and it becomes difficult to attach the sensor depending on the type of the telephone.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a call recording processing apparatus that enables accurate control of recording processing in accordance with a hook state in a telephone and reduces restrictions on a target telephone.

  In order to solve the above problems, a call recording processing apparatus of the present invention is connected to a telephone, and in a call recording processing apparatus for performing a recording process of a call on the telephone, is connected to a transmission side communication line of the telephone, A transmission signal processing unit that superimposes a probe signal modulated by a predetermined modulation method on a transmission voice signal in a telephone and a reception side communication line of the telephone, monitors the reception voice signal in the telephone, A received signal processing unit that detects an included probe signal and determines the on-hook / off-hook state of the telephone based on the detected probe signal, and a recording process when the telephone is determined to be off-hook by the received signal processor unit And a received signal processor that demodulates the received audio signal, and a demodulated signal output by the demodulator. A correlator that calculates a correlation value with the replica signal corresponding to the search signal generated by the transmission signal processing unit, and the telephone set is off-hook when the correlation value output from the correlator is equal to or greater than a predetermined threshold value. And a determination processing unit for determining.

  In such a call recording processing apparatus, a probe signal is superimposed on a transmission voice signal of a telephone so that the probe signal wraps around the received voice signal as a side sound, and the received voice signal is demodulated and then demodulated. A correlation value between the signal and the replica signal of the search signal is calculated, and when the correlation value is equal to or greater than a predetermined threshold, it is determined that the device is in the off-hook state. Based on the determination result, since the recording process is executed in the off-hook state, it is possible to accurately detect the hook state regardless of the level of the voice during the call, and between the on-hook state and the off-hook state. It is possible to reliably determine a change in the level of the side sound. In addition, by connecting to the transmission side communication line and the reception side communication line of the telephone, restrictions on the target telephone can be reduced.

  The received signal processing unit generates an analog-digital converter that converts the received audio signal into a received digital signal and a continuo signal having a previously derived code pattern, modulates the continuo signal, and superimposes it on the received digital signal And a correlator further calculating a correlation value between the demodulated signal and the code pattern, and the determination processing unit determines that the correlation value output from the correlator is a predetermined second threshold value. It is preferable to determine that the telephone is off-hook when:

  With such a configuration, the received audio signal is converted into a digital signal, the modulated signal is superimposed on the digital signal, and then the demodulated signal obtained by demodulating the digital signal is added to the continuous signal. A correlation value with the included code pattern is calculated, and when the correlation value is equal to or smaller than the second threshold value, it is determined that the device is in the off-hook state. Thus, by using the continuo signal together, it is possible to prevent a decrease in detection accuracy due to the received voice during a call in the detection of the hook state using the search signal.

  Further, when the telephone is in the off-hook state, the received signal processing unit sets the second threshold value to a second value between the correlator output level at the on-hook time and the current correlator output level. It is also preferable to further include a threshold control unit.

  In this way, for example, even if the correlator output is large in a call with a relatively low voice or noise level, the threshold for detecting a continuo signal is set according to the voice or noise level generated during off-hook. It can be set to an appropriate value.

  Further, the search signal is preferably a signal modulated by spread spectrum. By using such a search signal modulated by spread spectrum, it is possible to prevent the reception voice during the call from being annoying and to prevent the detection of the hook state from being disturbed by the call voice.

  According to the call recording processing apparatus of the present invention, it is possible to accurately control the recording process in accordance with the hook state of the telephone, and to reduce the restrictions on the target telephone.

  Hereinafter, preferred embodiments of a call recording processing apparatus according to the present invention will be described in detail with reference to the drawings. In the description of the drawings, the same or corresponding parts are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a diagram showing a configuration of a call recording system 1 including a call recording processing apparatus 10 according to a preferred embodiment of the present invention. As shown in the figure, a call recording system 1 includes a telephone 20 that performs communication related to a call to be recorded, a call recording processing device 10 connected to the telephone 20, a call recording processing device 10, and a LAN (Local Area Network). And a server device 30 connected via a network NW such as a WAN (Wide Area Network).

  The telephone 20 includes a hybrid circuit 21 connected to an external two-wire telephone line W1, a transmitter 22 connected to the hybrid circuit 21 via a transmission side communication line W2 inside the telephone 20, and a hybrid circuit. 21 is provided with a handset 23 connected to the telephone 20 via a receiving communication line W3. The hybrid circuit 21 transmits a voice signal received from the outside on the two-wire telephone line W1 as a received voice signal to the reception side communication line W3 side, and from the transmitter 22 via the transmission side communication line W2. The transmission voice signal received in this way is sent out on the telephone line W1. That is, the hybrid circuit 21 separates or synthesizes transmission voice and reception voice, which are analog signals, between a four-wire telephone line inside the telephone 20 and a two-wire telephone line outside the telephone 20. It is a 4-wire conversion circuit. Here, a part of the transmission voice signal transmitted from the transmitter 22 of the telephone 20 wraps around the reception side communication line W3 as “side sound” via the hybrid circuit 21. “Side tone” refers to a sound that is generated when a part of the sound that enters from the transmitter or around the telephone is output from the receiver through the internal circuit of the telephone.

  The call recording processing device 10 is connected to the transmission side communication line W2 of the telephone set 20, generates a search signal by a predetermined modulation method, and superimposes the search signal on the transmission voice signal in the telephone set 20; A reception signal processing unit 12 connected to the reception side communication line W3 of the telephone set 20, monitors a received voice signal, detects a probe signal, and determines the hook state of the telephone set 20 based on the detection result; and a reception signal processing unit A data transmission processing unit (recording processing unit) 13 that performs recording processing of the received audio signal based on the determination of 12, and a transmission level determination unit 14 that determines the signal level of the search signal or the like to an appropriate level. .

  The server device 30 is a server system that receives voice data for each call transmitted from the data transmission processing unit 13 of the call recording processing device 10, and creates a voice file, constructs a database, stores a voice file, and the like. .

  Hereinafter, each component of the call recording processing apparatus 10 will be described with reference to FIGS.

  As shown in FIG. 2, the transmission signal processing unit 11 includes a code generator 111, a signal double flow unit 112, a PSK modulator 113, and a D / A converter 114.

を用いて、Ｍ系列拡散符号を生成する。符号発生器１１１は、図３に示すように、ｎ段（ここでは、ｎ＝１０）のシフトレジスタ１１１ａと、式（１）の生成多項式に対応してシフトレジスタ１１１ａの複数段の状態の論理結合をシフトレジスタ１１１ａの入力にフィードバックする論理回路１１１ｂとから構成されている。 The code generator 111 has a function of deriving a digital code sequence for generating a search signal in synchronization with an externally input timing clock CL. In the call recording processing apparatus 10, a direct sequence modulation (DS) method, which is one type of spread spectrum modulation, is used as a method of modulating a search signal. The probe signal modulated by such direct spread modulation is excellent in signal secrecy and noise resistance because the signal energy is spread over a wide bandwidth, and has the following advantages. That is, even if it is superimposed on the received voice signal, it does not give an irritating impression to the user, detection of the search signal is hardly disturbed by the voice during a call, and the detection sensitivity is high even if the side sound level is low. Specifically, the code generator 111 is, for example, the following equation (1) as a generator polynomial for generating a digital code sequence;

Is used to generate an M-sequence spreading code. As shown in FIG. 3, the code generator 111 includes an n-stage (here, n = 10) shift register 111a and a logic of a plurality of stages of the shift register 111a corresponding to the generator polynomial of Expression (1). The logic circuit 111b feeds back the coupling to the input of the shift register 111a.

The M-sequence spreading code generated in this way is a code sequence that is repeated with a period of 2 ⁿ −1, and the difference between the numbers of 1 and 0 in each period is 1 or less (balanced characteristic). And a sequence shifted by an arbitrary number, the difference between the number of columns that match each other and the number of columns that do not match each other is 1 or less (correlation characteristics), and the run (number of consecutive identical digits) is 1 in each cycle. The number of cases is 1/2 of the total number of runs, the number of cases of 2 is 1/4, the number of cases of 3 is 1/8,... (Run characteristic), and two Ms that are arbitrarily shifted When the exclusive OR of the sequence spreading codes is taken, an M sequence spreading code subjected to a phase shift different from any of them is obtained (generic characteristics).

  Returning to FIG. 2, the signal double stream unit 112 replaces the digital value “0” of the M sequence spread code generated by the code generator 111 with “−1”, thereby double stream the M sequence spread code and PSK. Output to the modulator 113.

PSK modulator 113 generates a probing signal a carrier signal which is output from the carrier signal source S _c is modulated by PSK (Phase Shift Keying) modulation with the M-sequence spreading codes. Specifically, the PSK modulator 113 performs multiplication processing of the carrier signal and the spread code sequence.

  The D / A converter 114 converts the search signal output from the PSK modulator 113 from a digital signal to an analog signal, and sends the converted analog signal to the transmission side communication line W2 of the telephone set 20 so as to be superimposed on the transmission voice signal. Let

  As shown in FIG. 4, the reception signal processing unit 12 includes a continuo signal generation unit 115 that generates a continuo signal that is used together with a probe signal to detect the hook state of the telephone set 20, and an A / D converter (analog). (Digital converter) 116, adder 117, PSK demodulator 118, filter unit 119, limiter unit 120, and replica signal (replicated signal) corresponding to the continuo signal generated by continuo signal generator 115 Includes a replica signal generation unit 121, a correlator 122, an ATC unit (threshold control unit) 123, and a determination processing unit 124.

を用いて、符号発生器１１１が生成する符号系列と干渉しないＭ系列拡散符号を生成する。ここで、通奏信号は、受信音声信号がアナログ−デジタル変換された受信デジタル信号に重畳され、電話機２０のフック状態を検出するために探り信号と併用される信号である。信号複流化部１２７及びＰＳＫ変調器１２８は、それぞれ、信号複流化部１１２及びＰＳＫ変調器１１３と同様の処理を行う。アンプ部１２６は、符号発生器１２５によって出力された拡散符号系列に対して可変乗算処理を実行して信号複流化部１２７に出力する。より詳細には、アンプ部１２６は、Ｑ１５フォーマットの固定小数点演算による乗算処理を施す。 The continuo signal generation unit 115 includes a code generator 125, an amplifier unit 126, a signal double flow unit 127, and a PSK modulator 128. The code generator 125 has the same configuration and function as the code generator 111, and generates a digital code sequence for generating a continuo signal in synchronization with the timing clock CL. Specifically, the code generator 125 generates, for example, the following equation (2) as a generator polynomial for generating a digital code sequence in direct spreading modulation:

Is used to generate an M-sequence spreading code that does not interfere with the code sequence generated by the code generator 111. Here, the continuo signal is a signal that is superimposed on the received digital signal obtained by analog-digital conversion of the received audio signal and is used together with the search signal to detect the hook state of the telephone set 20. The signal double flow unit 127 and the PSK modulator 128 perform the same processing as the signal double flow unit 112 and the PSK modulator 113, respectively. The amplifier unit 126 performs variable multiplication processing on the spread code sequence output from the code generator 125 and outputs the result to the signal double stream unit 127. More specifically, the amplifier unit 126 performs multiplication processing by fixed-point arithmetic in the Q15 format.

  The A / D converter 116 obtains a received digital signal by performing analog-digital conversion on the received audio signal monitored from the receiving side communication line W3 of the telephone set 20, and outputs the received digital signal to the adder 117. This received digital signal is added to the continuo signal by the adder 117 and input to the PSK demodulator 118.

The PSK demodulator 118 performs demodulation processing on the received digital signal including the continuo signal, extracts a code sequence modulated by PSK modulation as a demodulated signal, and outputs it to the filter unit 119. That, PSK demodulator 118 obtains a carrier signal from a carrier signal source S _c inputted to the PSK modulator 113,128, by performing multiplication processing between the received digital signal, a demodulated signal by performing synchronous detection .

  The filter unit 119 performs a smoothing process using a moving average method to remove a high-frequency signal component from the demodulated signal extracted by the PSK demodulator 118 and outputs the result to the limiter unit 120. In this smoothing process, a Nyquist filter may be used, but here the moving average method is used in that the process can be simplified. For example, when the sampling frequency of the modulation signal is 8 kHz and the bit rate of the code is 1 kHz, the filter unit 119 executes the averaging process 8 times.

  The limiter unit 120 performs a positive / negative determination process on the data included in the demodulated signal and outputs the result to the correlator 122. In this positive / negative determination process, when each code data in the demodulated signal is larger than 0, it is determined as “1”, and when each code data in the demodulated signal is smaller than 0, it is determined as “−1”. An amplitude noise component included in the signal is removed to generate a demodulated code sequence.

  The replica signal generation unit 121 includes a code generator 129, a signal double flow unit 130, and a delay processing unit 131. The functions of the code generator 129 and the signal doubling unit 130 are the same as those of the code generators 111 and 125 and the signal doubling units 112 and 127, respectively, and the same code sequence as the spreading code sequence corresponding to the probe signal And a replica code pattern that is the same code sequence as the spreading code sequence corresponding to the continuo signal is generated. The two replica code patterns generated in the code generator 129 and the signal doubling unit 130 are synchronized with the spreading code sequence included in the demodulated code sequence, so that the delay processing unit 131 transmits the transmission signal processing unit 11 and the received signal processing unit. 12, a delay corresponding to the delay time caused by the conversion process, sampling process, etc. is given to the correlator 122.

により相関値φ（ｍ）を求める。ここで、Ｘ_２（ｎ＋ｍ）_ｍｏｄＮは、Ｎ点のデータに対するサイクリックシフト演算を示す。相関器１２２は、探り信号に対応して算出した相関値φ_１（ｍ）の１周期分の積算値と、通奏信号に対応して算出した相関値φ_２（ｍ）の１周期分の積算値とを更に算出し、それらの積算値をＡＴＣ部１２３を経由して判定処理部１２４に出力する。 The correlator 122 calculates the correlation value of the demodulated code sequence, the replica code pattern corresponding to the search signal, and the replica code pattern corresponding to the continuo signal for one period. Specifically, the correlator 122 uses the following equation (3) when the demodulated code sequence is X ₁ (n) and the replica code pattern is X ₂ (n);

To obtain a correlation value φ (m). Here, X ₂ (n + m) _mod N indicates a cyclic shift operation on data at N points. The correlator 122 is an integrated value for one cycle of the correlation value φ ₁ (m) calculated corresponding to the probe signal and one cycle of the correlation value φ ₂ (m) calculated corresponding to the continuo signal. The integrated value is further calculated, and the integrated value is output to the determination processing unit 124 via the ATC unit 123.

The determination processing unit 124 determines the on-hook / off-hook state of the telephone set 20 based on the respective integrated values of the correlation values φ ₁ (m) and φ ₂ (m) output from the correlator 122, and determination result data Is output to the data transmission processing unit 13. That is, when the integrated value of the correlation value φ ₁ (m) is equal to or greater than the predetermined threshold φ _th1, it is determined that the probe signal has sneak into the received voice signal, and the hook state of the telephone 20 is off-hook. On the other hand, if it is less than the threshold φ _th1 , the hook state of the telephone 20 is determined to be on-hook.

Here, when only the probe signal is detected, there is a possibility that the probe signal is disturbed by the voice during a call and the detection becomes impossible. Therefore, the determination processing unit 124 goes off-hook by the integrated value of the correlation value φ ₁ (m). After the determination, the continuo signal is preferentially detected. The continuo signal is a signal that is virtually injected at a constant level in the circuit on the receiver side. By constantly monitoring this signal, a superimposition of a probe signal or a reception voice signal from the transmitter 22 side when off-hook is performed. It is possible to detect a decrease in detection sensitivity due to. The hook state is determined based on the change in detection sensitivity. Specifically, when the integrated value of the correlation value φ ₂ (m) is equal to or less than a predetermined threshold φ _th2 , the determination processing unit 124 determines that the received audio signal includes an audio signal or a search signal. Therefore, it is determined that the hook state of the telephone 20 is off-hook. On the other hand, if the threshold value _φth2 is exceeded, the hook state of the telephone 20 is determined to be on-hook.

The ATC unit 123 detects threshold values φ _th1 and φ _th2 for detecting the hook state of the telephone set 20 based on the integrated values of the correlation values φ ₁ (m) and φ ₂ (m) output from the correlator 122. Is automatically set and output to the determination processing unit 124. For example, while the determination processing unit 124 determines that it is on-hook, the threshold φ _th2 is set to a set value of a predetermined ratio (for example, ½) of the maximum output level of the integrated value of the correlation value φ ₂ (m). and, while it is determined that the off-hook, the threshold phi _th2, an intermediate value between the output level of the integrated value of the maximum output level and current correlation values phi ₂ of the integrated value of the correlation value φ _{2 (m) (m)} Set to. In a general telephone, the side tone level is considerably suppressed even in an off-hook state due to the influence of a noise canceller or an echo canceller, and the difference in the output level of the correlator 122 tends to be small. Since the threshold value control described above increases the threshold value in such a case, the threshold value for detecting a continuo signal is appropriately set according to the level of voice and noise generated during off-hook and on-hook. Can be set to any value.

  Returning to FIG. 1, based on the determination result output from the received signal processing unit 12, the data transmission processing unit 13 performs a call recording process in the telephone 20 in accordance with the timing when the telephone 20 changes from the on-hook state to the off-hook state. And the recording process is stopped at the timing when the telephone 20 changes from the off-hook state to the on-hook state. This recording process is performed by creating voice data based on the voice signal of the monitored call. The data transmission processing unit 13 sequentially transmits the created audio data to the server device 30.

  The transmission level determination unit 14 performs a calibration process for setting the levels of the search signal and the continuo signal to appropriate levels. FIG. 5 is a flowchart showing the calibration process for the search signal by the transmission level determination unit 14, and FIG. 6 is a flowchart for the calibration process for the continuo signal by the transmission level determination unit 14. Hereinafter, the procedure of the calibration process will be described.

  As shown in FIG. 5, first, the transmission level determination unit 14 monitors the reception side communication line W3, and receives a DT (dial tone) coming from a telephone line W1 side exchange (not shown) when the telephone is in an off-hook state. ) The signal level is measured and set as a reference level (step S01). Next, the transmission signal processing unit 11 is controlled to generate a search signal at a preset maximum level (step S02). At this time, the received voice signal on which the search signal is superimposed is monitored, and the DT signal is removed from the monitored received voice signal by processing using a 400 Hz comb filter (step S03). Further, the level of the received audio signal from which the DT signal has been removed is measured (step S04). Finally, the level of the search signal is set so that the level of the received audio signal is not disturbing to the user and can be detected with reference to the reference level, and the search signal is generated at the set level. The transmission signal processing unit 11 is controlled as described above (step S05).

  Next, referring to FIG. 6, transmission level determination unit 14 superimposes the search signal on the transmission audio signal at the level set by the process shown in FIG. 5 when telephone 20 is in the on-hook state (step S11). Next, the amplifier unit 126 of the received signal processing unit 12 is controlled to generate a continuo signal at a preset level (step S12). Thereafter, the transmission level determination unit 14 starts monitoring the output level corresponding to the continuo signal output from the correlator 122 of the reception signal processing unit 12 (step S13). Then, the output level from the correlator 122 is compared with a predetermined maximum receivable value (step S14). If the output level exceeds the maximum receivable value (step S14; YES), the reception signal processing unit 12 is processed. Is controlled to reduce the level of the continuo signal (step S15). On the other hand, when the output level does not exceed the maximum receivable value (step S14; NO), the reception signal processing unit 12 is controlled to increase the level of the continuo signal (step S16). As a result, when the output level of the correlator 122 reaches a level close to the maximum receivable value (step S17; YES), the level of the continuo signal is fixed (step S18). On the other hand, when the output level of the correlator 122 does not reach a level close to the maximum receivable value (step S17; NO), the process returns to step S14 to adjust the level of the continuo signal (steps S14 to S17). repeat. With the calibration process as described above, the continuo signal can be set to a level that can maintain sufficient detection sensitivity when on-hook.

  Hereinafter, the operation and effect of the call recording processing apparatus 10 described above will be described.

In the call recording processing apparatus 10, the probe signal is superimposed on the transmission voice signal of the telephone 20, so that the probe signal wraps around the received voice signal as a side sound, and the demodulated signal is demodulated after the received voice signal is demodulated. And a replica code pattern corresponding to the probe signal are calculated, and when the correlation value is equal to or greater than a predetermined threshold φ _th1 , it is determined to be in the off-hook state. Based on the determination result, since the recording process is executed in the off-hook state, it is possible to accurately detect the hook state regardless of the level of the voice during the call, and between the on-hook state and the off-hook state. It is possible to reliably determine a change in the level of the side sound. In particular, since the search signal is continuously supplied even during the off-hook state, the off-hook state can be continuously detected even if the voice in the call is interrupted or the voice level is low. Furthermore, even when the internal circuit is always energized in a telephone or the like with a headset, the difference in the side sound level that returns to the receiving communication line can be sufficiently determined. In addition, by connecting the telephone 20 to the transmission side communication line W2 and the reception side communication line W3, restrictions on the target telephone can be reduced.

Further, after converting the received audio signal into a received digital signal, after superimposing a continuo signal modulated using a spreading code sequence on the received digital signal, a demodulated signal obtained by demodulating the received digital signal; A correlation value between the spread code sequence and the same replica code pattern is calculated, and when the integrated value of the correlation values is equal to or less than a predetermined threshold φ _th2, it is determined as an off-hook state. By using the continuo signal inserted in the circuit on the receiving side in this way, it is possible to prevent a decrease in detection accuracy due to the received voice during a call in the detection of the hook state using the search signal.

  Further, by using a signal modulated by spread spectrum as the search signal, it is possible to prevent the reception voice during the call from being annoying and to prevent the call voice from interfering with the detection of the hook state.

  Next, the experimental result regarding the detection of the hook state for the call recording processing apparatus 10 according to the embodiment of the present invention will be described.

  Here, as the call recording processing apparatus 10 to be tested, the frequency (chip frequency) per code bit is 1000 Hz, the spreading code cycle is 1.023 sec, the code sequence is 1023 bit M-sequence code, the carrier frequency is 2000 Hz, and code synchronization. The method is fixed synchronization, the maximum occupied frequency band is 977Hz to 3023Hz, the sampling frequency is 8000Hz, the audio data format is μ-Law or A-Law, the frequency characteristic of audio is 600 to 3,400Hz (± 4dB), S / N ratio Was about 42 dB.

FIG. 7 shows a configuration of a call recording system 201 used for detecting a hook state. As shown in the drawing, an NCU (Network Control Unit) 202 having a built-in MODEM is connected to a control PC 203 for controlling a call operation by the NCU 202. In addition, the telephone 20 is connected to the NCU 202 via a telephone line, and a partner telephone 205 is connected via the telephone line and the pseudo exchange 204. The NCU 202 sets the telephone 20 to the off-hook state by a built-in relay circuit, and then sets an arbitrary call time and repeatedly connects the telephone 20 and the partner telephone 205. The NCU 202 sets the telephone 20 in an on-hook state immediately after the end of each call connection. At this time, an answering machine function is set in the other party's telephone 205, and a recorded voice corresponding to the call time is played. Table 1 shows the measurement conditions in this test.

In an experiment using such a call recording system 201, by comparing the call time between the control log of the NCU 202 and the voice file stored in the server device 30 based on the call start time, as shown in Table 2 Experimental results were obtained. As a result, it has been found that hook detection can be performed almost satisfactorily regardless of the type of telephone.

It is a figure which shows the structure of the call recording system containing the call recording processing apparatus which is preferable one Embodiment of this invention. FIG. 2 is a block diagram of a transmission signal processing unit in FIG. 1. It is a figure which shows the structure of the code generator of FIG. It is a block diagram of the received signal processing part of FIG. It is a flowchart which shows the calibration process by the transmission level determination part of FIG. It is a flowchart which shows the calibration process by the transmission level determination part of FIG. It is a figure which shows the structure of the call recording system containing the call recording processing apparatus which is a to-be-tested apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Call recording processing apparatus, 11 ... Transmission signal processing part, 12 ... Reception signal processing part, 13 ... Data transmission processing part (recording processing part), 20 ... Telephone, 115 ... Consecutive signal generation part, 116 ... A / D Converter (analog / digital converter), 118... PSK demodulator, 122... Correlator, 123... ATC unit (threshold control unit), 124 ... determination processing unit, W2... Transmission side communication line, W3.

Claims (4)

In a call recording processing apparatus that is connected to a telephone and performs a recording process of a call on the telephone,
A transmission signal processing unit which is connected to the transmission side communication line of the telephone and superimposes a probe signal modulated by a predetermined modulation method on a transmission voice signal in the telephone;
Connected to the receiving communication line of the telephone, monitors a received voice signal at the telephone, detects the probe signal included in the received voice signal, and detects the probe signal based on the detected probe signal. A received signal processing unit for determining an on-hook / off-hook state;
A recording processing unit that executes the recording process when the received signal processing unit determines that the telephone is in an off-hook state,
The received signal processor is
A demodulator that demodulates the received audio signal;
A correlator for calculating a correlation value between the demodulated signal output by the demodulator and the replica signal corresponding to the probe signal generated by the transmission signal processing unit;
A determination processing unit that determines that the telephone is in an off-hook state when a correlation value output from the correlator is equal to or greater than a predetermined threshold;
A call recording processing apparatus. The received signal processor is
An analog-to-digital converter that converts the received audio signal into a received digital signal;
A continuo signal generating unit that generates a continuo signal having a previously derived code pattern, modulates the continuo signal and superimposes it on the received digital signal;
The correlator further calculates a correlation value between the demodulated signal and the code pattern;
The determination processing unit determines that the telephone is in an off-hook state when a correlation value output from the correlator is equal to or smaller than a predetermined second threshold value.
The call recording processing apparatus according to claim 1. The received signal processor is
A threshold control unit configured to set the second threshold value to a second value between an output level of the correlator at the time of on-hook and a current output level of the correlator when the telephone is in an off-hook state; In addition,
The call recording processing apparatus according to claim 2.   The call recording processing apparatus according to claim 1, wherein the search signal is a signal modulated by spread spectrum.

Images