JP4845598B2 - Signal detection method and noise removal method - Google Patents

Description

  The technical field to which the present invention belongs is that when a digital signal is modulated at a high frequency, the digital signal is transmitted as a radio wave, the radio signal is received and the digital signal is demodulated, the demodulated digital signal is compared with the digital signal before the high frequency modulation. It belongs to the technical field of signal correction methods for increasing the identity. In particular, this is a method for demodulating a digital signal used when the electric field strength of a radio signal is weak or when machines are crowded and many noises and radio waves of the same frequency are emitted.

  In recent years, wireless communication that performs data transfer by modulating a digital signal at a high frequency has become common with the spread of wireless sensor networks. As a method for removing noise in wireless communication, there is a method of removing interference and noise components by sharpening filter characteristics and narrowing a pass band when performing Morse communication. In recent years, in the intermediate frequency band and the low frequency band, a digital signal converted from an analog signal to a digital signal using an A / D converter and converted using a DSP (digital signal processor) element. A method of removing unnecessary portions and noise and further converting the signal into an analog signal using an A / D converter is realized. However, these two methods are complicated in structure, so that the size of the device is large, so that it is difficult to apply in a small device, and since many elements are used, the current consumption increases. There was a disadvantage that it could not be used for a small device driven by a power source.

Furthermore, in these two methods, the problem is that the frequency of the signal wave must be adjusted to a specific frequency or frequency range from the element characteristics, so that an electric signal generated by a crystal oscillation circuit or the like. Is required to generate a signal wave frequency that matches the characteristics of the element to be used, and to filter harmonics, that is, a superheterodyne mechanism. Therefore, both the device scale and the current consumption increase. On the other hand, as a wireless device is reduced in size and power consumption, it is better to assemble a wireless circuit using a package-in type wireless module or a wireless IC than to assemble a wireless circuit using discrete elements. It is also advantageous.
Japanese Patent No. 365783

  The purpose of the present invention is to increase the accuracy rate of the received signal when demodulating the received signal. However, the radio unit used in the present invention is a digital signal, that is, baseband input and output and subsequent. Wireless processing, that is, processing for high-frequency modulation of a signal and transmission from an antenna, or processing for demodulating a high-frequency signal received by the antenna and converting it to a baseband signal wave is all performed by a package-type wireless module. That is, if hardware or software changes are made to the wireless module itself, it may affect the performance of the wireless module, so it is necessary to process the baseband signal wave output from the wireless module. This is a challenge.

  In recent years, central processing devices that have a remarkably low operating power but a high processing speed have appeared, and it has become possible to shape a baseband signal waveform by using these central processing devices.

  The signal detection method of the present invention detects the signal position that should be originally detected from the characteristics of the received waveform when demodulating a digital signal transmitted wirelessly in a wireless device using a coupling capacitor in a detection circuit. Distinguish between signal and noise.

  The reception noise elimination method of the present invention uses a method for determining a data communication speed from the characteristics of a received waveform and the determination method in a radio apparatus using a coupling capacitor in a detection circuit.

  The number detection method of the present invention calculates the position on the time axis where the next signal rises from the position of the signal that should be originally determined and the determined communication speed, slices before and after, and measures the signal voltage. To detect the rising edge of the signal.

    The signal detection method of the present invention determines that there is no signal input when the rising edge of the signal is not detected.

  The main part of the wireless device used in the present invention is composed of a wireless unit mainly composed of the wireless module in hardware, a central processing unit that controls and controls communication, and a power source. In order to increase the accuracy rate in signal demodulation, means to determine whether the baseband signal wave output from the radio unit is a signal desired by the central processing unit, an interference signal or a noise component is used. It is done.

  Further, as a known method, the data transfer speed according to the communication distance, that is, when the communication distance is long, the communication speed is reduced or the Manchester method, that is, when the digital signal 1 (one) is sent, There is a method of sending 0 (zero) of the signal. In wireless communications, keeping the sideband width as constant as possible helps to improve reception efficiency, but if this wireless device is specified in an environment with a lot of environmental noise or spark noise, it is a signal that does not exist originally. It is difficult to improve the reception efficiency without removing.

  Wireless communication using digital signals often has different data transfer speeds, and even if the data is sent at a certain speed, the frequency of the crystal oscillator used in the central processing unit related to data transmission varies from one individual to another. Therefore, there are few cases where the frequency exactly matches the frequency specified by the numerical value. In addition, since the crystal oscillator is sensitive to temperature, in a crystal resonator that does not have a temperature guarantee function, frequency drift due to temperature affects the data transfer rate. Therefore, it is first necessary to correctly identify the speed of data communication. As a method of identifying the data transfer rate, a method of measuring using a preamble is common. In normal wireless communication, a method called transmitting a signal called a preamble and then transmitting data to be originally transmitted is adopted. In many cases, 1 and 0 are sent alternately in the preamble, and even in the communication used in the present invention, data is sent after sending 1 and 0 alternately in 32 bits.

  INDUSTRIAL APPLICABILITY The present invention is useful for demodulating an original signal by minimizing the influence of the environment when there is a lot of environmental noise and spark noise in the usage environment of the wireless device, and thus can improve communication reliability. . In addition, when the radio device has a lot of interference such as indoors and the antenna electromotive force due to the signal wave is remarkably reduced, or when the signal wave is very close to the noise level in long-distance communication, the accuracy of the received signal is increased. Useful.

  Many preambles in digital communication usually send 1 and 0 signals alternately. This has the purpose of stabilizing the detection circuit of the receiver. In particular, when a coupling capacitor is used in the detection circuit, it has the effect of stabilizing the potential of the capacitor and speeding up the response to the signal.

  The output waveform of a digital signal detection circuit using a coupling capacitor is a waveform with the characteristics of a capacitor output, and it is possible to measure the speed of the signal by utilizing this characteristic, and the preamble is sent. In the meantime, the length of the signal, that is, the data transfer rate is determined from the waveform of the signal wave, and based on this, the signal that should originally be separated from the noise can be easily read.

  In the wireless device used in the experiment of the present invention, a microprocessor equipped with a flash memory is used as a transceiver module for transmitting and receiving an OOK signal. OOK modulation is a type of amplitude modulation method that outputs a high-frequency output when the baseband input signal is 1 and does not output when the input signal is 0. The frequency of the main carrier was 315 MHz, and the signal wave was a digital signal obtained by OOK modulation.

  As an implementation method, the method of calculating the baseband data transfer rate using the output voltage of the coupling capacitor at the time of preamble transmission is changed to the first embodiment, and the signal wave is changed from 1 to 0 and further changed from 0 to 1. A peak determination method when the signal rises in this case is described in the second embodiment, and similarly, a peak determination method when the signal wave shifts from 1 to 0 and the next signal is 0 is described in the third embodiment.

  In this embodiment, a method for determining a data transfer rate from a preamble using balance data will be described. In the detection circuit using a coupling capacitor, the voltage value indicated by the capacitor is the data that the output waveform from the detection circuit is sent on the transmission side because the charge accumulated in the capacitor is gradually discharged by the signal wave. It is often different from the waveform.

  FIG. 1 shows a baseband waveform 100, a transmission waveform 110, and a reception waveform 120 when a coupling capacitor is used in the detection circuit when data changes from 0 to 1 and then to 0. When the signal changes from 0 to 1, a spire voltage is generated and a large peak appears in the received waveform 120. In the process of accumulating charges in the steeple voltage and capacitor, a valley is formed once and the maximum voltage value is seen. Here, when the transmission waveform 110 changes from 1 to 0, the voltage drops according to the discharge waveform of the capacitor.

  FIG. 2 shows a temporal relationship between the transmission waveform and the reception waveform. The received signal is sliced at certain time intervals and the voltage is measured. The noise threshold and the signal voltage exceeding the threshold are set as the signal voltage 1, and the first signal voltage when the received signal falls below the noise threshold is set as the signal voltage. N. This is shown in FIG.

  Next, between the signal voltage 1 and the signal voltage N, there are three singular points, that is, points where the slope of the tangent is zero. Of the three points, the lowest point of the signal voltage is defined as a singular point B shown in FIG. The singular point B is required to be in the middle of the other two points. The singular point on the signal voltage 1 side of the singular point B is set as P1, and the singular point on the signal voltage N side of the singular point B on Saturday is set as the singular point P2. As a specific method for searching for a singular point, a method in which a slope is measured from two adjacent signal voltages between signal voltage 1 and signal voltage N, and a point where the sign of the slope changes is taken as a singular point is generally used. Is.

  As the noise threshold value, a value obtained by inputting a numerical value or continuously receiving a noise component for a certain time in a state where there is no signal and calculating an average value of integrated voltages is generally used.

  The singular point P1 is the spire voltage when the signal changes from 0 to 1, and the singular point P2 is the point at which the signal switches from 1 to 0. When the signal wave is 1, the charge is stored in the capacitor and the voltage rises. However, when the signal wave is switched from 1 to 0, the charge stored in the capacitor is not supplied and changes to discharge, so the signal voltage drops. become. Therefore, the length of the baseband signal, time T, is obtained by multiplying the difference between the signal voltage 1 and the singular point P2 by the sampling interval. FIG. 4 shows time T.

  In this example, in order to improve time accuracy, time measurement was performed for 24 times except for the first 8 times for 32 preambles, and the result was taken as an average value.

  When the shape of an actual received signal is viewed with a spectrum analyzer or the like, the shape is basically the shape shown in FIGS. 1 to 5, but many deformed waveforms can be seen. In particular, the pinnacle voltage is often much higher than the voltage at the singular point P2, and when a value of 0 or 1 is determined using a numerical value set in a table as in the past, it can be taken as two different signals. Even if the signal is not as sharp as spike noise, it has been found that the signal may be misidentified. Therefore, it has been experimentally proved that it is effective to determine the data transfer rate by using the method shown in this embodiment.

  This embodiment is a method for determining whether the next signal is 1 or 0 after the signal wave is changed from 0 to 1 and further changed from 1 to 0 when actual data is received. Will be described. In this embodiment, a case where the next signal is 1 will be described.

  Using the singular point P2 and time T found in the previous embodiment, the position for searching for the rising edge of the signal is determined. The search start reference point in FIG. 5 is the base point for signal rise search. This point is at a position after time T from the singular point P2. Since the time T is a value corresponding to the baseband data transfer time, the signal wave is near the singular point P2, that is, the point where the signal is assumed to have changed from 1 to 0, and the point after the time T has elapsed. May change from 0 to 1.

  However, since there is a frequency error in the data transfer clock, it is necessary to search for the rising edge of the signal in a time zone that takes into account the error.

  In FIG. 6, the search start point and the search end point will be described. The search start point is at a point that is a certain time T2 before the search reference point. From this point, the signal is sliced at an appropriate fixed interval, and the voltage is measured. The slice of the signal wave and the voltage measurement are performed until the end of the search as shown in FIG. The search end point is at a position delayed by time T2 from the search start reference point. Since the time T2 depends on the stability of the crystal oscillator that determines the accuracy of the high frequency, in this embodiment, the accuracy of the crystal resonator is a range, that is, the accuracy of the crystal resonator is ± 10 PPM. On the other hand, the search start point is set at a position on the time axis 20PPM before the range. Whether the signal wave rises can be confirmed by the sliced voltage value exceeding the noise threshold, but considering the effects of spike noise etc., at least two voltage measurements exceed the noise threshold. Preferably it is. In this embodiment, the condition is that three consecutive slice measurement voltages exceed the noise threshold.

  In the implementation condition of the present invention, the Manchester method is used. In this method, the signal wave does not have two or more 0s or two or more consecutive 1s. Accordingly, when two 1s are consecutive, 0 is sure to come next, so by performing the operation shown in FIG. 7 from the time when the time T is tripled from the rise of the first signal wave as the base point, It can be determined whether the next signal is 0 or 1.

  In the present embodiment, the case where the next signal is 0 after the signal wave is changed from 0 to 1 and further changed from 1 to 0 when actual data is received will be described. In the method of searching for the rising edge of the signal, a method according to the second embodiment is used.

  A method for confirming that no signal wave peak exists will be described with reference to FIG. Similarly to FIG. 7, the place assumed to be the rising position of the signal wave is sliced along the time axis from the search start point to the search end point, and the signal voltage at each slice point is measured. However, when the signal is 0, unlike the case of 1, since there is no input as a signal wave, a noise level is output. Therefore, the slice voltage value does not continuously exceed the noise level. However, since there is a possibility of spike noise, the signal wave is again transmitted at the point where the time axis is lowered by the time T from the exploration start time when the slice voltage value is highest, at the peak confirmation point as shown in FIG. And slice current value at that position is measured. If the slice current value measured here is below the noise threshold value, it can be determined that there is no input signal. However, if the slice voltage value at the peak confirmation point exceeds the noise threshold due to spark noise or the like and does not exceed 70% of the peak threshold in FIG. 3, it is determined that there is no signal input. In other cases, it is necessary to slice the vicinity of the peak confirmation point and determine the value. If the slice value of the neighboring point does not exceed the noise threshold value, it is determined that there is no signal input, and if the noise threshold value is exceeded and 70% of the peak threshold value in FIG. .

  The present invention is effective as a method for discriminating noise from a signal in a noisy environment or when a signal is attenuated and approaches a noise level. As a place with a high noise level, a place with a lot of machinery such as a factory, a vicinity of a power line, etc. can be raised. Examples of cases where the signal is attenuated include cases where the communication distance is long and transmission power is low.

Diagram showing comparison between transmitted waveform and received waveform Diagram showing the relationship between transmit and receive waveforms Diagram showing peak threshold and noise threshold The figure which showed the position of singular point P1, P2 and singular point B Figure showing the search start reference point for peak presence / absence Diagram showing the exploration start point and exploration end point Diagram showing signal wave slicing near the exploration start reference point The figure which showed the waveform when there is no rise from 0 to 1 of the signal wave

Explanation of symbols

100 Baseband waveform 110 Transmission waveform 120 Reception waveform

Claims (3)

In a wireless device that uses a coupling capacitor in the detection circuit, when demodulating a digital signal transmitted wirelessly, it is determined from the position of the signal that should be detected from the characteristics of the received waveform and the characteristics of the received waveform. A signal detection method for detecting the rising edge of a signal by determining the position on the time axis at which the next signal rises from the data communication speed, slicing the signal before and after it, and measuring the signal voltage. 2. The signal detection method according to claim 1, wherein when no rising edge of the signal is detected, it is determined that there is no signal input. A noise removal method for distinguishing noise from a signal by using the method according to claim 1.

Images