JPH023470B2 - - Google Patents

Description

[Detailed Description of the Invention] <<Industrial Application Field>> The present invention relates to a receiver used in Loran C, which is a type of hyperbolic radio navigation, and in particular has a function of detecting a specific cycle of a carrier wave in a received pulse. Regarding Loran C receiver.

<<Background of the Invention>> In Loran C, two or more slave stations form one chain for one main station, and among these Loran C transmitting stations, the main station is as shown by M in FIG.
Generates 9 Loran pulses, and the slave stations also have S ₁ ,
Send 8 Loran pulses as shown by S ₂ . Each transmitting station generates the Loran pulse at a pulse repetition period determined for each chain. In addition, each slave station has a different delay time (coating,
generates a slave pulse with delay).

Therefore, in the Loran C receiver, the distance difference between the master station and the slave station from the two fixed points is determined by the reception delay time of the slave pulse with respect to the master station pulse, and the distance difference is calculated on the Loran hyperbola drawn between the two fixed points. This method measures the location of the receiver.

In the Loran C receiver described above, in order to determine the reception delay time of the slave pulse with respect to the master pulse, a specific cycle of the carrier wave (generally the third cycle carrier wave) in each received pulse is detected and automatically tracked. It has functions.

The carrier wave Ca of the Loran pulse (the time axis is shown enlarged in Fig. 1 b and c) has a frequency of
100KHz, so one cycle is
It will be 10μsec.

Conventionally, Loran C receivers equipped with the above functions include:
For example, JP-A-55-2938, JP-A-55-6261,
There are some such as those shown in Japanese Patent Publication No. 56-2312.

In these conventional Loran C receivers, the carrier wave
The procedure for detecting the third cycle of Ca is that when a Loran pulse LP is received, the pulse repetition period of this Loran pulse LP (99.7 msec in the Japanese sea area)
A sampling pulse is generated in synchronization with the Loran pulse LP, and if data above a predetermined threshold level is obtained from the sampling data in a state where this sampling pulse is synchronized with the Loran pulse LP, this is determined as the peak P ₁ of the carrier wave Ca. , P ₂ , P ₃ , ..., and after detecting the peak of the carrier wave obtained in this way, the first three peak data of this peak data are detected.
The configuration is such that the third cycle of the carrier wave Ca is detected by making the third cycle data the peak of the third cycle.

However, in the conventional Loran C receiver, the third cycle of the carrier wave is performed by detecting the peak of the carrier wave Ca.
If the S/N between the Loran C signal and external noise decreases, it will be difficult to reliably detect the peaks of the first and second cycles, and there is a risk that the third cycle will be incorrectly detected. It was hot.

In particular, when trying to measure the position of a moving vehicle by installing the above-mentioned Loran C receiver in a car, the attenuation of the Loran C signal is large in urban areas or mountainous areas.
There are many regions where the S/N of the received signal is smaller than 0 dB, and the above problem cannot be ignored.

<<Object of the Invention>> An object of the present invention is to provide a Loran C receiver that is capable of detecting a specific cycle of a carrier wave accurately and in a short time even under conditions where the S/N of the Loran C signal is poor.

<<Configuration of the Invention>> The configuration of the present invention will be briefly explained based on the block diagram shown in FIG. 2 which clearly shows the invention.

The Loran C receiver of the present invention receives Loran C
It is equipped with specific cycle detection means for detecting a specific cycle of a carrier wave of a signal.

The specific cycle detection means 100 includes an envelope detection means 101 that detects the envelope of the received Loran C signal carrier wave every predetermined time width, and a reference envelope that stores in advance a standard of the envelope within the predetermined time width. A storage means 102, a calculation means 103 for calculating the variance or standard deviation of the difference value between the detected envelope and the stored reference envelope, and a cycle based on the variance or standard deviation calculated by the calculation means 103. A cycle determining means 104 is provided for determining the cycle.

<<Description of Embodiment>> Hereinafter, an embodiment of the Loran C receiver according to the present invention will be described in detail with reference to FIG. 3 and subsequent drawings.

FIG. 3 is a block diagram showing the configuration of one embodiment of the Loran C receiver according to the present invention. The configuration shown in the figure shows only the main components according to the present invention, and other components, such as the initial detection circuit of the Loran C signal and the circuit that tracks the zero-crossing point of the carrier wave of the Loran C pulse, etc. As this is well known, illustration and description thereof will be omitted.

In the figure, a high frequency amplifier 2 receives a received signal including a Loran C signal received by a receiving antenna 1.
The _amplified received signal is supplied to the limiter circuit 3. This limiter circuit 3
shapes the amplified received signal into a square wave,
It converts into digital signals of "1" and "0".

The shift register 4 is a k-bit (for example, 160 bits) shift register, and sequentially receives the digitized received signal data D ₂ supplied from the limiter circuit 3 in accordance with the sampling pulse PS supplied from the counter 8. It is a matter of sampling. The latch circuit 5 latches each 4-bit output of the shift register 4, and when the shift register 4 finishes sampling the Loran pulse, the divided pulse of the counter 8
_D4 latches the 160-bit data. Microcomputer (hereinafter referred to as MPU)
Reference numeral 6 supplies address signals to the latch circuit 5 and RAM 7 through the address bus AB, and exchanges various data through the data bus DB. Then, when the divided pulse _D4 from the counter 8 is input to the interrupt terminal,
The data latched in the latch circuit 5 is read and transferred and stored in the RAM 7.

The clock generating circuit 11 generates a clock signal in synchronization with the carrier wave Ca of the Loran C signal, and the clock signal _D5 output from the clock generating circuit 11 is supplied to the preset counter 10 and the counter 8. ing.

The preset counter 10 is connected to the MPU 6.
This preset counter 10 counts a predetermined number based on preset data from
The count output _D6 of is a D-type flip-flop 9.
is supplied to. This D-type flip-flop 9
The output _D7 is "1" during the period when the count output _D6 of the preset counter 10 is supplied, and when the frequency division pulse _D4 from the counter 8 is input to the clear terminal, the output _D7 becomes "1". Returns to “0”.

During the period when the output _D7 supplied from the D-type flip-flop 9 is "1", the counter 8 divides the frequency of the clock signal _D5 inputted to the clock terminal by 1/M, and uses this divided frequency. , supplies a sampling pulse to the clock terminal of the shift register 4, and in parallel, divides the frequency of the clock signal _D5 by 1/N to obtain the divided pulse.
Connect _D4 to the latch terminal of the latch circuit 5 and the MPU
6 and the clear terminal of the D-type flip-flop 9.

Note that the values of N and M have a relationship of N>M, and further have a relationship of k=N/M. In the following explanation, specifically, N=400, M=2.5, k
= 160 for explanation. That is, the shift register 4 is a 160-bit shift register,
From counter 8, the sampling pulse PS is
One hundred and sixty pulses are output every 2.5 μsec, and the frequency-divided pulse _D4, which is also output from the counter 8, is output every 400 μsec.

In the Loran C receiver configured as described above, the SN ratio of the received signal is low, and the received signal D ₁
However, as shown in Figure 5, Loran pulse LP and noise
Assume that the signal contains Nz. and,
In the Loran C receiver, it is assumed that initial synchronization is established at point Ta of the Loran pulse LP and tracking is performed by the Loran pulse tracking circuit.

The above received signal _D1 is converted by the limiter circuit 3 into a digital signal of "1" and "0" levels as shown by _D2 in FIG.

FIG. 4 is a flowchart showing the contents of the sampling data addition and storage processing routine executed in the MPU 6. As described above, this processing is performed when the Loran pulse LP is detected and the tracking of the zero cross point of the carrier wave Ca (time Ta ) is executed after

When the process shown in FIG. 4 starts, the addition number counter A is reset and preset data is supplied to the preset counter 10 at a predetermined timing. That is, the above Loran Pulse
A count output D ₆ is supplied to the D-type flip-flop 9 for 400 μsec from a time point Ts which is 360 μsec before the initial synchronization point Ta of LP.

Along with this, the counter 8 outputs 160 sampling pulses PS at a pulse interval of 2.5 μsec for 400 μsec from the above-mentioned time point Ts, as shown in FIG. _5D3 . Therefore, each pulse of this sampling pulse group is sent to the shift register 4, and the limiter circuit 3
Sampling of the output signal D ₂ from is performed sequentially.

Sampling is performed on the shift register 4 as described above, and at the time point Te when 160 pieces of data have been sampled, that is, 400 μsec after the time point Ts, the frequency division pulse D ₄ is sent from the counter 8.
is output. Accordingly, the data (160 bits) sampled in the shift register 4 is latched into the latch circuit 5, the output _D7 of the D flip-flop 9 becomes "0", and the counting operation of the counter 8 is stopped. Further, the frequency-divided pulse _D4 of the counter 8 is supplied to the interrupt terminal of the MPU 6, thereby causing step (F-2) in FIG.
The execution effect becomes YES, and then the step (F-
The process 3) is executed, and the 160-bit sampling data latched in the latch circuit 5 are read and stored at predetermined addresses in the RAM 7, respectively.

The RAM 7 has enough addresses to store two cycles of sampling data for each Loran pulse from one master station and two slave stations.

That is, in this embodiment, sampling is performed 160 times for one Loran pulse, so
For 2 cycles, it is double that, the number of transmitting pulses for each station is 8 pulses (excluding the 9th pulse of the main station pulse), and the number of receiving stations is 3 in total for the main station and slave stations.
Since it is divided into stations, the total number is 7680 (=160×2×8
×3) memory area is required. And 1
If one memory area consists of 8 bits, a memory capacity of 7680 bytes is required.

Then, step (F-4) in Fig. 4 is executed, and the number of transmission pulses from each station is 8 for 2 cycles.
Since the number of pulse and receiving stations is three in total for the master station and slave station, it is determined whether or not the interrupt signal _D4 has been input 48 times. Therefore, the above steps (F-2) to (F-4) are performed 48 times,
For each received pulse, 160-bit sampling data is stored in the RAM 7 over two cycles.

After the 7680 sampling data are collected, the addition count counter A is incremented in step (F-5), and the memory address is reset to the initial value (step (F-6)).

Furthermore, by executing step (F-7), it is determined whether the addition number counter has reached a predetermined value L, and until it reaches the predetermined value L,
The above steps (F-2) to (F-7) are repeated. This allows the above RAM
In the memory area No. 7, L sampling data is added and stored for each Loran pulse from the main station and the slave station. Each of the above memory areas is reset to "0" in advance, so that when the stored sampling data is "1" (when the received signal data _D2 is at the "1" level at the time of sampling), 1 in memory area
is added. On the other hand, when the sampling data is "0" (when the received signal data _D2 is at the "0" level at the time of sampling), addition to the memory area is not performed.

In this way, by performing data processing in the memory area L times, the contents of each memory area become as shown by MD in FIG. 5 (note that the diagram shown in FIG. 5 is This shows the contents of the memory area for one Loran pulse among the above memory areas).

As shown in the figure, the contents of the memory area are as follows:
At the stage of initial detection of the Loran pulse, the carrier wave
Since tracking to the zero cross point of Ca (time Ta) is being performed, the above 160 sampling pulses
PS is synchronized with the carrier wave Ca. Therefore,
If the n-th sampling pulse Psn of the sampling pulses Ps is synchronized with the peak of the carrier wave Ca, the contents of the memory area storing the data sampled by this sampling pulse Psn will be L or 0.

In addition, the contents of the memory area that stores the data sampled by the sampling pulse Psn+1 synchronized with the zero-crossing point of the carrier wave Ca and the data obtained by the sampling pulse that samples the noise part other than the Loran pulse LP are , If the above noise Nz is randomly generated noise and its distribution is expressed as a Gaussian distribution, the probability that the data sampled by the above sampling pulse PS will be "1" or "0" is considered to be 1/2. Therefore, for the result of the above L operations, +1 addition is performed approximately L/2 times, and as a result, the content of the memory area becomes approximately L/2.

Furthermore, among the contents of the memory area shown in Fig. 5, there are differences in the values of the added data in the sampling portion of the Loran Pulse LP, but this is because the SN ratio of the received Loran Pulse LP is different. This is because the SN ratio of the rising portion of the Loran pulse LP is worse than that of the other portions, so the probability of sampling data error is high, and the content of the memory becomes smaller than L (or larger than 0). In this case, as long as the Loran pulse LP exists, the probability that the sampled data will be "1" or "0" will not be 1/2, and there will be a bias.
If the number of additions L of the above sampling data is large, the final memory content will always be larger (or smaller) than L/2, and the Loran pulse LP and noise
It is clear that it can be distinguished from Nz.

To distinguish between Loran pulse LP and noise Nz above,
The contents of the memory area are compared with a predetermined reference level (L/2±∆l), and if the memory contents are within the range of L/2±∆l, this is considered to be "zero crossing or noise of the Loran pulse". In other cases, “Loran pulse peak” is detected.
It can be determined that

A process for detecting a specific cycle of the Loran pulse is executed based on the sampling sum data for each peak of the Loran pulse obtained through the above processing.

First, the principle will be explained.

The envelope V _(t) of the pulse waveform from the rising portion of the Loran pulse LP, that is, from the first cycle to the sixth cycle of the carrier wave, is defined by the following equation.

V _(t) = At ² e ^-2t/65 ... (1) where A is a constant and t is time in the range of 0≦t≦65μsec.

Figure 6 shows the relative amplitude at each point in time at 0≦t≦65μsec in the rising portion of the Loran pulse LP, assuming that the amplitude at t=65μsec is 0dB.
It is expressed in dB. The points a to l in the figure represent the relative amplitudes of the positive and negative peaks of the carrier wave of the Loran pulse from the first cycle to the sixth cycle.

Therefore, the amplitude data of each peak point (hereinafter referred to as envelope data) obtained through actual measurements and each point a, b, c... obtained from equation (1)...
...l amplitude data (hereinafter referred to as reference envelope data) and the standard deviation of the difference value, it is possible to determine which portion of the reference envelope data the envelope data is.

Now, through actual measurement, envelope data at four peak points from the point initially determined to be a peak are E ₍₁₎ , E ₍₂₎ , E ₍₃₎ , E ₍₄₎ , E ₍₅₎ , E _{( 6)} . In addition to this, a, b, c, d, obtained from equation (1),
Reference envelope data R ₍₁₎ at points e and f,
The difference value D _(o) with R ₍₂₎ , R ₍₃₎ , R ₍₄₎ , R ₍₅₎ , R ₍₆₎ is
It is determined by D _(o) = R _(o) −E _(o) . (However, n is 1~
6 integer) Next, the standard deviation σ is the difference value D _(o) of each data
Since it is the square root of the square mean of the difference between and its average value, it is expressed by the following equation.

Here, = 1/6 ₆ 〓 ⁿ⁼¹ D _(o) The above value of σ shows different values depending on which peak point of the carrier wave of the Loran pulse E ₍₁₎ is.
In other words, if the data of E ₍₁₎ is the positive peak point of the first cycle of the carrier wave of the Loran pulse (point a is shown in Figure 6), the envelope data and the reference envelope data have the same shape. Since they match, the difference value D _(o) becomes a constant value of 1. Therefore, D _(o) = and there is no variation, so
The standard deviation σ is 0.

Also, due to radio interference, etc., the first cycle of the carrier wave of the Loran pulse was not detected, and E ₍₁₎ was the positive peak of the second cycle (point c is shown in Figure 6). Then, the envelope data and the reference envelope data do not match in shape, and a predetermined value of standard deviation σ is obtained.

Therefore, E ₍₁₎ is the peak point in each cycle of the carrier wave (points a, b, C...l in Fig. 6).
By calculating in advance the value of _the standard deviation σ when the It is possible to determine what cycle the peak is.

FIG. 7a shows the difference between the envelope data and the reference envelope data when the peak point determined at the beginning of the measured envelope data is each peak point a to j of the Loran pulse carrier wave shown in FIG. It is a figure showing the value of the standard deviation of the value.

However, the measured envelope data usually has some errors due to the influence of noise in the received signal. Therefore, it is necessary to consider the error when determining which cycle of the carrier wave E ₍₁₎ is.

Since the noise contained in the Loran signal is considered to be random Gaussian noise, the value of standard deviation σ is a Gaussian noise with a maximum width of σ _o centered on a to j, as shown in Figure 7b. given as a distribution. Note that, since the standard deviation σ cannot take a negative value, the distribution at point a has a shape in which the negative portion is folded around σ=0.

In this way, the value of the standard deviation σ is given stochastically within the range of the Gaussian distribution, so there is a part where the distributions of the standard deviation σ at each peak point overlap, and the value of the standard deviation σ is the part where the distributions overlap. If E ₍₁₎
It is not possible to determine exactly which cycle of the carrier wave is located. For example, if the value of standard deviation σ is
If it is in the range represented by W ₁ , it is not possible to determine whether E ₍₁₎ is point a or b because it can be either point a or point b. Also, in W ₂ ,
In E ₍₁₎ , the distributions of c, d, and e overlap, making it impossible to determine the cycle.

However, since the width of the distribution of the standard deviation σ is caused by Gaussian noise contained in the Loran signal, the width of the distribution is determined by the number of additions. If the width of the distribution when adding l times is σ _o , then the distribution when adding 2l and 4l times is σ _o /√
2, it has a value of σ _o /2. (Fig. 7, c,
Shown in d. ) In this way, if the value of the standard deviation σ found is in a part where the distributions overlap, it is possible to determine which peak point E ₍₁₎ corresponds to by increasing the number of additions. .

By the way, in order to determine which peak point E ₍₁₎ is at as described above, if the obtained value of standard deviation σ is in the overlapping part of the distributions, further addition is necessary. If this is not done, it will not be possible to accurately determine the cycle.

Therefore, in order to determine the cycle more efficiently, determine whether the data of E ₍₁₎ is a positive peak or a negative peak, and if it is positive, select the peak point on the positive side of the Loran pulse carrier wave. In other words,
Determine which peak point of a, c, e, g, i is E ( ₁ ), and if E ₍₁₎ is negative, b, d,
One possible method is to determine which peak point among f, h, and j.

A specific method is to input the data at each point a, b, c, d, e, f starting from the positive peak of the first cycle as the first standard envelope data, and then As reference envelope data, data at each point b, c, d, e, f, and g starting from the negative peak of the first cycle is entered. Then, if the value of E ₍₁₎ is positive, the standard deviation σ of the difference value between the envelope data E ₍₁₎ to _{E (6)} and the first reference envelope data is determined. Then, as shown in Fig. 8 a to c, E ₍₁₎ points a, c,
The distribution of standard deviation values for each peak of e, g, and i is calculated in advance, and from the results, E ₍₁₎
It is possible to determine which point among a, c, e, g, and i corresponds to.

Also, when E ₍₁₎ is negative, similarly find the standard deviation σ of the difference value between the envelope data E ₍₁₎ to E ₍₆₎ and the second reference envelope data, and calculate the standard deviation σ of the difference value between the envelope data E (1) to E (6) and the second reference envelope data, and The distribution of standard deviations when E ₍₁₎ is at each peak of points b, d _, f, h, and j is calculated in advance as shown in ,f,h,j
It is possible to determine which point of the object corresponds to this point.

As shown in Figure 7b and Figures 8a and d, if cycle discrimination is performed separately depending on whether the data of E ₍₁₎ is a positive peak or a negative peak, In addition, since there is little overlap between the standard deviation distributions of each peak, cycles can be determined with a small number of additions.

Based on the above principle, the process shown in FIG. 9 is actually performed. This is a flowchart of the cycle discrimination processing routine.

This processing is carried out after the sampling data addition and storage processing routine shown in FIG.
When compared with , the point that first disappears within the range of L/2±Δl is determined to be the peak. Also, if the data is larger than L/2+Δl, the polarity is positive, L/
If it is smaller than 2-Δl, the polarity is determined to be negative.

Next, in step (F-11), the absolute value of the difference from L/2 is calculated for the data of every other sampling point starting from the point that was first judged to be a peak. Starting from the data of the points, register them as envelope data of E ₍₁₎ , E ₍₂₎ , E ₍₃₎ , E ₍₄₎ , E ₍₅₎ , and E ₍₆₎ in order.

At step (F-12), step (F-12)
Determine whether the polarity obtained in step 10) is positive or not. If the polarity is positive, in step (F-13), find the standard deviation σ of the difference value between the envelope data obtained in step (F-11) and the first reference envelope data obtained in advance. . Then, in step (F-14), it is determined whether the obtained value of standard deviation σ is in the overlapping portion of the distributions shown by U ₁ to _{U 4} in FIG. 8a. Therefore,
If the value of standard deviation σ is not in the overlapping portions U ₁ to _{U 4} of the distributions, it is determined in step (F-17) what cycle E ₍₁₎ is.

In step (F-14), if the value of the standard deviation σ is in the overlapping portion of the distributions, cycle determination is not performed and the process is terminated.

If it is determined in step (F-12) that the polarity is not positive, then in step (F-15) the standard deviation σ of the difference value between the envelope data and the second reference envelope data is calculated. . and,
It is determined whether the value of the standard deviation σ obtained in step (F-16) is in the overlapping portion of the distributions indicated by _U5 to _U8 in FIG. 8d. Then, the cycle is determined in the same manner as described above, or the process is ended without determination.

The principles and processing for each part have been described above, but here we will summarize the actual flow of the entire processing. FIG. 10 is a flowchart showing the overall flow. This process is executed after the Loran pulse is detected and the zero-crossing point of the carrier wave is tracked.

First, in step (F-20), the sampling data addition and storage processing routine shown in FIG. 4 is executed, and the data at the sampling point for each station is added and stored L times. Next, if the cycle of the main station is not determined in step (F-21), the cycle determination processing routine shown in FIG. 9 is executed for the main station in step (F-22). An attempt is made to determine the cycle. Whether cycle discrimination is performed or not,
The cycle discrimination processing routine for the main station ends. Next, if the cycle of slave station 1 is not determined in step (F-23), proceed to step (F-24).
Then, cycle determination is similarly attempted for slave station 1 as well. Next, if cycle determination has not been made for slave station 2 as well, cycle determination is attempted. Then, in step (F-27),
It is determined whether cycle discrimination has been performed for all stations, including the master station, slave station 1, and slave station 2.
If there is a station that has not been identified, step (F
-20), add L times to the data that was added L times previously, for a total of 2L times,
For stations whose cycles have not yet been determined,
Cycle determination is performed.

In this way, cycles are determined for all stations, the master station, slave station 1, and slave station 2.

In addition, in the above example, the standard deviation of the difference value between the envelope data and the reference envelope data is used, but the same thing can be said even if the double value of the variance (σ ² ) is used. Needless to say.

In addition, in the example, the sampling time of the Loran signal is set to 400 μsec for each Loran pulse, but
This sampling time may be set to an optimum time by taking into consideration various conditions such as the maximum delay time of the spatial wave with respect to the surface wave.

<<Effects of the Invention>> As explained above, in the Loran C receiver according to the present invention, the S/N ratio of the received signal decreases,
Even if the first cycle, second cycle, etc. of the Loran pulse carrier wave cannot be detected, the cycle can be reliably determined using the variance or standard deviation, and the third cycle or other cycles can be detected accurately and quickly. It is possible to detect specific cycles.

[Brief explanation of drawings]

Figure 1 shows the waveform of the Loran C signal, Figure 2 shows the waveform of the Loran C signal.
FIG. 3 is a block diagram showing the configuration of an embodiment of the Loran C receiver according to the invention, and FIG. 4 is a block diagram showing the configuration of an embodiment of the Loran C receiver according to the invention. A flowchart showing the sampling data addition storage processing routine, FIG. 5 is a timing chart showing the main outputs and memory contents of the device for Loran pulses, FIG. 6 is a diagram showing the envelope of Loran pulses, and FIG. Figure 8 is a diagram showing the standard deviation value and its probability distribution based on the principle of cycle discrimination. Figure 8 is a diagram showing the probability distribution of the standard deviation and changes in the probability distribution due to changes in the number of additions. Figure 9 is the cycle discrimination process. FIG. 10 is a flowchart showing the entire cycle determination process executed by the microcomputer of the embodiment. 100... Specific cycle detection means, 101...
Envelope detection means, 102... Reference envelope storage means, 103... Calculation means, 104...
Cycle determination means.

Claims (1)

[Claims] 1. In a Loran C receiver equipped with a specific cycle detection means for detecting a specific cycle of a carrier wave of the received Loran C signal, the specific cycle detection means detects an envelope of the received Loran C signal carrier wave. Envelope detection means for detecting every predetermined time width; reference envelope storage means for pre-memorizing a standard of the envelope within the predetermined time width; and a difference value between the detected envelope and the stored reference envelope. 1. A Loran C receiver comprising: calculation means for calculating the variance or standard deviation of the calculation means; and cycle discrimination means for determining a cycle based on the value of the variance or standard deviation calculated by the calculation means.