JP5005446B2 - Independent high-sensitivity satellite signal receiver - Google Patents

Description

  The present invention is installed in a mobile body such as a cellular phone or a car, and even in a very weak environment such as a satellite signal from a navigation satellite such as GPS indoors or under an elevated position, the edge position (navigation data) And a self-supporting high-sensitivity satellite signal receiver that can determine the tracking frequency by itself.

  A global system such as the Global Positioning System (GPS) in the United States, which uses information about the distance to the navigation satellite (artificial satellite) that orbits the earth and the orbit of the navigation satellite to determine the position and speed of objects on the earth. In the satellite signal receiver of the Navigation Satellite System (GNSS), there are various situations for the satellite signals received. Hereinafter, the GPS will be described as an example.

  When the S / N ratio of the I and Q correlation signals for 1 ms is good (situation 1), a satellite receiver with normal sensitivity can be used. In a normal-sensitivity autonomous satellite signal receiver, first, PN code detection, frequency acquisition, and carrier phase acquisition are performed, and then the edge position can be detected by inverting the I correlation signal every 1 ms (Patent Document 1). ).

  The S / N ratio of the I and Q correlation signals for 1 ms is low and buried in noise, but the S / N ratio of the coherent result for 20 ms with the frequency estimated correctly is in a good situation (situation 2) Then, it is possible to estimate the edge position and inversion of the navigation data every 20 ms (Patent Document 2).

  In this Patent Document 2, (1) the edge position between each 20 ms is selected from the 20 candidates with the addition timing shifted by 1 ms (Sum1 is selected in the example of FIG. 2C of Patent Document 2). (2) The navigation data transition between each 20 ms block exceeded 20 ms by selecting the one that maximizes the signal power overall, taking into account the inversion of the navigation data of the continuous 20 ms blocks. Signal acquisition is performed by coherent addition.

  However, in the method of Patent Document 2, since the determination is made every 20 ms as described above, the method is applied when the S / N ratio of the coherent result for 20 ms in a state where the frequency is correctly estimated is also bad. I can't.

  Further, since the method of Patent Document 2 aims at signal acquisition, it does not matter if the edge positions estimated independently for each 20 ms are slightly different. That is, in the signal acquisition, when 20 ms is added, if the true edge position is obtained, the 1 ms addition value * 20 is obtained, but even if the 1 ms edge position is incorrect, a sufficiently large signal power of 1 ms addition value * 18 can be obtained. Therefore, there is no problem in the signal acquisition stage as long as the edge position roughly matches the true edge position. However, in the present invention, since it is necessary to accurately obtain the edge position, the method of Patent Document 2 cannot be used as an edge detection method.

  The S / N ratio of the I and Q correlation signals for 1 ms is low and buried in noise, and the S / N ratio of the coherent result for 20 ms in a state where the frequency is correctly estimated is also bad (situation 3). Has improved the S / N ratio by extending the addition time by using coherent addition of I and Q correlation signals for 1 ms and noncoherent addition for noncoherent addition of the addition result of the coherent addition. A satellite signal tracking device has been proposed by the present inventor (Japanese Patent Application No. 2006-293440; hereinafter referred to as the prior invention).

  Coherent addition is an addition method in which an I (carrier positive phase) correlation signal and a Q (carrier 90 ° phase shift) correlation signal are added as they are, and a method capable of greatly improving the S / N ratio. In GPS, the inversion of the I and Q correlation signals occurs only every 20 ms, so coherent addition is performed up to 20 ms. Since frequency analysis by FFT is possible within a range where the inversion of the I and Q correlation signals does not occur, frequency analysis by FFT may be performed within the coherent addition interval.

On the other hand, non-coherent addition is a method of adding I, Q correlation signals or signal power P (= I ² + Q ² ) obtained by coherent addition of these signals. Although the S / N ratio improvement of this method is inferior to coherent addition, it is not affected by the inversion of the I and Q correlation signals, so the S / N ratio can be improved even when the inversion pattern of the navigation data from the satellite is unknown. There is an advantage that can be done.

  However, the invention of the prior application aims to perform stable frequency tracking even in a very weak environment such as indoors or under an elevated position, and cannot be applied as it is to accurately perform edge detection.

Further, a network type satellite signal receiver that uses assist information from a network is known as a method that enables edge detection even in the situation 3 (Patent Document 3). In this network type high-sensitivity satellite signal receiver, the headquarter server receives satellite signals, calculates GPS navigation data, and sends them to the terminal side. In this case, by using the inversion pattern of the transmitted navigation data and changing the sign of the navigation data received on the terminal side, the sign of the I signal and Q signal is not reversed by the navigation data. As a result, it is possible to perform coherent addition for a navigation data repetition period (20 ms for GPS) or more, and high sensitivity of edge position detection is realized. However, since the present invention is a self-supporting high-sensitivity satellite signal receiver that cannot obtain any navigation data information from the network, the method of Patent Document 3 cannot be used.
Japanese Patent Laid-Open No. 10-246768 US2007 / 0008217A1 JP 2000-193735 A

  The present invention has been made in view of the above points, and in an environment where the satellite signal from a navigation satellite such as GPS is extremely weak such as indoors or under an elevated position, an edge position (of navigation data) is required. It is an object of the present invention to provide a self-supporting high-sensitivity satellite signal receiver that can determine the switching timing) and the tracking frequency.

The satellite signal receiver according to claim 1 includes a carrier local oscillator that receives a frequency control signal and generates a carrier frequency signal, converts a satellite signal including navigation data, and receives a received signal related to a specific satellite signal and PN A 1 ms correlation channel circuit that takes a code correlation with a code and a carrier correlation with the carrier frequency signal, and simultaneously outputs an I correlation signal and a Q correlation signal for 1 ms every 1 ms;
A plurality of J frequency converters for calculating and outputting a frequency-converted I-correlation signal and a frequency-converted Q-correlation signal that are respectively frequency-converted with respect to different frequency candidates for the I-correlation signal and the Q-correlation signal;
The frequency conversion I correlation signal and the frequency conversion Q correlation signal from the plurality of J frequency converters are added to each of the addition sections corresponding to the plurality of K edge position candidate positions with respect to the edge position indicating the switching timing of the navigation data. A plurality of J * K coherent adders, each of which outputs an I coherent addition value and a Q coherent addition value,
The signal power for each coherent adder is obtained by receiving the I coherent addition value and the Q coherent addition value from the plurality of J * K coherent adders, and the non-coherent addition during that period for at least one predetermined period. A plurality of J * K non-coherent adders,
An edge detection circuit controller for simultaneously determining an edge position and a tracking frequency indicating the switching timing of navigation data from the addition results of the plurality of J * K non-coherent adders;
It is provided with.

  The satellite signal receiver according to claim 2 is the satellite signal receiver according to claim 1, wherein an edge position detection is performed between the 1 ms correlation channel circuit 1 and the plurality of J frequency converters 51 to 5J. And an addition start time selector 3 for setting an addition interval for use, and the setting of the addition interval is performed based on an addition start time instruction indicating the addition interval from the edge detection circuit controller 10. .

  According to a third aspect of the present invention, in the satellite signal receiver according to the first or second aspect, when the edge position E0 is determined by the edge detection circuit controller 10, one frequency candidate k (k is , 1 to K), the edge position determination is performed when the ratio of the maximum non-coherent result to the second largest coherent result is equal to or greater than a predetermined threshold.

  The present invention also provides a first coherent adder 811 to 8JK having M times of coherent times and a second coherent addition of L times (L> M) as a plurality of J * K non-coherent adders. It is preferable to provide the devices 911 to 9JK. Thereby, the detection time of the edge position is shortened according to the degree (good or bad) of the S / N ratio of the I and Q correlation signals for 1 ms.

  Further, when the S / N ratio targeted by the present invention is poor, the edge detection error rate is high, and therefore the probability that the edge positions are continuous is low. When the difference in the number of appearances between the candidate with the largest number of occurrences and the candidate with the second largest number of appearances exceeds a predetermined threshold, the edge detection is completed (edge detection end).

  According to the satellite signal receiver of the present invention, in the environment where the satellite signal from the navigation satellite such as GPS is very weak such as indoors or under the overhead, no assist information from a network often used for high sensitivity is used. The edge position (navigation data switching timing) and tracking frequency necessary for positioning position calculation can be determined independently.

  Hereinafter, with reference to the drawings, an embodiment of the self-supporting high-sensitivity satellite signal receiver according to the present invention will be described using GPS as an example.

  In the present invention, the edge position (navigation data switching timing) and the tracking frequency necessary for positioning position calculation are determined. In the signal acquisition process as the preceding process, the PN code detection, PN code phase acquisition, and Frequency pulling to within several tens of Hz is performed. This signal acquisition processing may be performed by providing a dedicated signal acquisition circuit, or may be performed using the edge detection addition circuits 41 to 4J of FIG. 1 of the present invention. Detailed explanation is omitted because it is not directly related to the problem.

  In FIG. 1 showing an overall configuration diagram of a satellite signal receiver according to an embodiment of the present invention, a satellite signal arriving at an antenna 101 from a satellite is converted to an intermediate frequency by a down converter 102 and is converted by an A / D conversion circuit 103. It is converted into a digital signal for digital processing and input as a received signal to a 1 ms correlation channel circuit 1 including a carrier NCO2.

  The 1 ms correlation channel circuit 1 includes a carrier local oscillator 2 that receives a carrier frequency control signal obtained in the previous acquisition process and generates a carrier frequency signal, converts a satellite signal, and receives a received signal related to a specific satellite signal that is input. The code correlation with the PN code and the carrier correlation with the carrier frequency signal are taken, and the I correlation signal I0 and the Q correlation signal Q0 for 1 ms are simultaneously output every 1 ms. The 1 ms correlation channel circuit 1 may be the same as that used in a conventional GPS receiver.

  The addition start time selector 3 supplies the I correlation signal I0 and the Q correlation signal Q0 for 1 ms from the 1 ms correlation channel circuit 1 to the edge detection addition circuits 41 to 4J in the next stage as they are, or starts addition. Only at the start time, data I0 and Q0 for a certain time are discarded, that is, the data I0 and Q0 after that time are supplied to the edge detection addition circuits 41 to 4J without being sent to the edge detection addition circuits 41 to 4J. This is selected according to the instruction to start addition from the edge detection circuit controller 10. At the start of the edge detection, the I correlation signal I0 and the Q correlation signal Q0 for 1 ms are supplied as they are to the next edge detection addition circuits 41 to 4J. An addition start time instruction is given according to the status of subsequent processing.

  The edge detection adding circuits 41 to 4J are provided in accordance with J frequency candidates, and are supplied with the I correlation signal I0 and the Q correlation signal Q0 for 1 ms, respectively, and output the addition result. Each of the edge detection addition circuits 4j (where j = 1 to J) includes a frequency converter 5j to the frequency candidate j and K edge candidate addition circuits 6jk (where k = 1 to K). Or K may be 20, for example.

  In the frequency converter 5j, with respect to the I correlation signal I0 and the Q correlation signal Q0 for 1 ms, the frequency conversion I correlation signal I0j and the frequency conversion Q corresponding to the frequency of the frequency candidate j are obtained by the same frequency rotation method as in the prior invention. A correlation signal Q0j is obtained and output.

  In the frequency conversion in this frequency conversion device 5j, as in the prior invention, the I correlation signal I0 and the Q correlation signal Q0 are multiplied by a multiplier (cos ωNT−jsin ωNT), and the real part is used as a frequency conversion I correlation signal. The imaginary part is used as a frequency conversion Q correlation signal. The rotation angle ωNT is given by a digital word from a numerically controlled oscillator in accordance with each candidate frequency, and the values of cos ωNT and sin ωNT can use a ROM table.

  The frequency conversion process in the frequency conversion device 5j may be performed with a very slow cycle of the target I and Q correlation signals I0 and Q0 every 1 ms, and one conversion circuit sequentially time-divides each candidate frequency. Therefore, it can be executed with an extremely small circuit, and the current consumption can be extremely reduced. This is clearly superior in comparison with the case where frequency conversion is performed for a plurality of candidate frequencies using a local oscillator (carrier NCO).

  Here, the frequency control signal F0 obtained by the signal acquisition process in the previous stage is in a state having an error of ± 30 Hz, and the frequency FS to be obtained by the edge detection circuit controller 10 in the present invention is obtained by a tracking circuit (not shown). Assuming that the frequency can be within 10 Hz, the frequency candidates obtained from the frequency converters 5j (j = 1 to J) are F0 + 30 [Hz], F0 + 20 [Hz], F0 + 10 [Hz], and F0 [Hz]. , F0-10 [Hz], F0-20 [Hz], and F0-30 [Hz]. In this case, J = 7.

  The edge candidate addition circuit 6jk coherently adds each frequency transformed I correlation signal I0j and frequency transformed Q correlation signal Q0j in each addition section corresponding to the corresponding edge position candidate position (edge candidate k of the frequency candidate j). Nms coherent adder 7jk that outputs an I-coherent addition value and a Q-coherent addition value, and receives an I-coherent addition value and a Q-coherent addition value from the coherent adder 7jk, And M times of the non-coherent adder 8jk that performs non-coherent addition during a predetermined period M times, and the signal power of each coherent adder jk is obtained for a predetermined period L times (where L> L times non-coherent adder 9jk that performs non-coherent addition during period M) Kick. Providing M times (first) coherent adder 8jk and L times (second) coherent adder 9jk allows the degree of S / N ratio of the I and Q correlation signals for 1 ms (good or bad). ) In order to shorten the edge position detection time.

  The edge detection circuit controller 10 determines and outputs the edge position E0 and the frequency FS based on the addition results from all the J * K edge candidate addition circuits 6jk. This edge position E0 is used for pseudorange calculation and confirmation of the starting point of 20 ms. The frequency FS is used for carrier frequency tracking. Further, the edge detection circuit controller 10 supplies an addition start time instruction to the addition start time selector 3 when necessary prior to the determination of the edge position E0 and the frequency FS. Hereinafter, the determination method thereof will be described.

  First, the coherent and non-coherent results of the edge candidate k in the edge detection addition circuit 4j will be described. FIG. 5 is a diagram for explaining a difference in noise contribution at the addition section boundary. In the case of GPS, there are a total of 20 candidates from start positions 00 to 19 with the addition interval changed by 1 ms in FIG. 5 as edge start positions.

Here, as an example
(1) The true edge position is the start position 6
(2) The frequency of the frequency of the frequency candidate j matches the true frequency to be tracked, and the absolute values of the added values in each 1 ms section of FIG. 5 are all “1”.
(3) The added value in each 1 ms section of FIG. 5 is “−1” in the section of 00 to 05 and “+1” in the section of 06 to 25.
The coherent result in the first half of FIG. 5 will be described.

  In this case, for example, the coherent result for 20 ms of the edge candidate at the start position 00 is −6 + 14 = + 8 since 00 to 05 is “−1” and 06 to 19 is “+1”. On the other hand, the coherent result for 20 ms of the edge candidate at the start position 06, which is the true edge position, becomes +20 because 06 to 25 are all “+1”.

  When this is obtained for all start positions (edge candidate numbers), the maximum coherent result at the true edge position as shown in FIG. 2A when there is edge inversion in FIG. 2 showing the correspondence between edge inversion and coherent results. Since this is the (first peak), the edge candidate from which the maximum coherent value is obtained may be set as the true edge position.

  However, the navigation data may not be reversed. In the above example, when the section from 00 to 05 is also “+1”, the coherent result for 20 ms of the edge candidate at the start position 00 is also +20. That is, when the inversion of the navigation data is not included, the coherent result is almost the same at any edge position as shown in FIG. 2B, and the edge value cannot be determined.

  Therefore, in the present invention, unlike Patent Document 2, the second position shown in FIG. 2 among the coherent results regarding one frequency candidate k (k is any one of 1 to K) when determining the edge position E0. If the ratio of the first peak (first peak value / second peak value), which is the maximum non-coherent result to the second peak, which is a very large coherent result, is equal to or greater than a predetermined threshold, the edge position determination is performed accurately. Edge detection can be performed.

  In the above description, in order to make the description easy to understand, the description is made with a single coherent result. However, in actuality, in order to perform edge detection even when the S / N ratio is low, a plurality of coherent results are reflected. The above edge position E0 is determined based on the non-coherent result of the number of times.

  Further, the difference in the non-coherent result due to the difference in frequency from the frequency converter 5j becomes significant with the non-coherent time as shown in FIG. 3, so when the difference (or ratio) becomes equal to or greater than the threshold value, Determine the true frequency. Since the frequency control signal F0 is drawn to several tens of Hz by the signal acquisition process in the previous stage, for example, the frequency candidates may be set at intervals of 10 Hz.

  In determining the true frequency, there are a plurality of edge position candidates in each of the frequency candidates 1 to J, but the non-coherent result is simply compared among all the edge candidate addition circuits 611 to 6JK to obtain the maximum non-coherent result. The true frequency may be determined independently from the edge position determination time from the candidate for which the coherent result is obtained, and after the edge candidates are almost determined within each of the frequency candidates 1 to J, The true frequency may be determined from the non-coherent result.

  As described above, the present invention can be limited to only a limited number of types of frequency candidates and is efficient, and thus can save energy. Note that in the case of FFT as in Patent Document 2, only the coherent addition interval can be used, and even if a narrow frequency range may be examined as in the present invention, all frequencies are required until the final stage. Candidate ranges must be calculated, which is inefficient.

  Now, the edge position E0 and the tracking frequency FS can be determined by the method described above. However, if the error is 1 ms in the edge detection, the positioning position is greatly mistaken as 300 km or more, so the final edge detection error rate A is It is necessary to make the value sufficiently small. “Situation 3” targeted by the present invention; the S / N ratio of the I and Q correlation signals for 1 ms is low and buried in noise, and the S / N of the coherent result for 20 ms in a state where the frequency is correctly estimated. In a situation where the ratio is also bad, it is inevitable that a detection error occurs with a probability of being at the detected edge position. Therefore, in the edge detection, the final edge position detection error rate A needs to be set to a sufficiently small value so as to meet the user's request.

  Only the edge position detection error rate B of one non-coherent result cannot satisfy the final edge detection error rate A. When the S / N ratio is good, the edge position detection can be completed when the same edge position is continuously detected several times or more. However, when the S / N ratio targeted by the present invention is poor, the edge position detection error rate is high, and therefore the probability that the same edge position is continuously detected is low. It cannot usually be expected to be detected continuously. Therefore, in the present invention, in edge position detection in situation 3, the difference between the number of detections (number of appearances) of the edge position candidate with the largest number of detections (number of appearances) and the second most frequent edge position candidate (number of appearances) exceeds a predetermined threshold. It is preferable to complete the edge position detection (edge position detection end).

  In the present invention, as shown in FIG. 1, the signal power for each coherent adder jk is obtained, and M times non-coherent adder 8jk that performs non-coherent addition during a predetermined period M times, as well as each coherent addition. The signal power for each device jk is obtained and used in combination with the L-time non-coherent adder 9jk that performs non-coherent addition during a predetermined period of L times (where L> M). Providing the M times (first) coherent adder 8jk and the L times (second) coherent adder 9jk allows the S / N ratio of the I and Q correlation signals for 1 ms (good, or This is for shortening the detection time of the edge position in response to (bad).

  When the S / N ratio of the I and Q correlation signals for 1 ms is good, the edge error rate is sufficiently small even if the non-coherent time is short. Therefore, the edge is shorter when the non-coherent time is short as shown in FIG. Detection time is short. On the other hand, when the signal S / N ratio is poor, the edge error rate is large when the non-coherent time is short, and it is necessary to check many times. Therefore, when the non-coherent time is long as shown in FIG. Edge detection time is short. That is, the shortest edge position detection time can be realized by performing both of these methods at the same time and adopting the method that ends earlier.

  By the way, the edge position detection error rate (edge error rate) is true as shown in FIG. 6A showing the case where the true edge position in FIG. 6 for explaining the edge error rate is not near the addition section boundary. The case where the addition interval is different by ± 1 ms from the edge position is sufficiently higher than other different addition intervals. Note that since the edge error rate is the probability that the edge position is “not true, wrong”, the edge error rate at the true edge position cannot be defined. In FIG. 6, instead of the edge error rate, the detection rate of edge detection can be expressed in the same way as the edge error rate, and the edge detection rate other than the true edge position can be expressed. The occurrence rate is expressed by a higher occurrence rate than the occurrence rate of edge detection other than the true edge position.

  Further, when the true edge position is in the vicinity of the addition section boundary, the edge error rate is further increased when the addition section is different by +1 ms or −1 ms as shown in FIG.

  The reason for this will be described with reference to FIG. 5 in the case where the true edge position is 0 ms. When the addition section is shifted by +1 ms, there is a difference between the addition section only in the addition head and the last 1 ms, which are blacked out in FIG. That is, most of the sections are the same. On the other hand, as compared with the case where the addition section is −1 ms and the start position = 19, the hatched portion of 19 ms * 2 in FIG. 5 is another time addition section. In a state where the signal level is low, the signal component is less than or equal to noise by only coherent addition for 20 ms. Therefore, the difference in noise due to different addition sections is easily reflected in the addition result, and the edge error rate is further increased. End up.

  Therefore, in the present invention, the edge detection circuit controller 10 observes the edge determination candidate obtained from the non-coherent result prior to the determination of the edge position E0 and the frequency FS, and the edge determination candidate is near the addition section boundary. In such a case, it is necessary to shift the addition time given to the addition start time selector by a predetermined interval (for example, 10 ms). In order to shift this addition time by a predetermined interval, an addition start time instruction is supplied to the addition start time selector 3. As a result, it is possible to always perform edge detection in the state as shown in FIG. The initial value of the addition time is the time when the edge detection addition circuits 41 to 4J are started to be used.

  In addition, when the number of edge determination candidates is larger in the ± 5 ms section of the addition section boundary, the above-described adjustment of shifting by 10 ms may be performed separately in the ± 5 ms section of the addition section boundary. Further, the edge determination candidate information before the 10 ms adjustment may be stored as position information shifted by 10 ms, and the information may be used effectively after the 10 ms adjustment.

  Since it is sufficient that the addition start time can be distinguished in the 5 ms section, the error rate can be performed with a smaller number of times than the edge position determination. For example, if the number threshold for determining the addition start time is 2 and the number threshold for determining the edge position is 4 times, if two times before the adjustment of the addition start time are also effectively used, two times after the adjustment You will have to wait.

  In the above description, the case where the addition start time selector 3 and the edge detection addition circuits 41 to 4J are configured by hardware has been described as an example, but these may be realized by software.

  Further, the configuration has been described in which the assist information from the network is not used. However, if there is usable information, for example, information such as the tracking frequency, the approximate edge position, and the navigation data, they may be used.

1 is an overall configuration diagram of a satellite signal receiver according to an embodiment of the present invention. Diagram showing correspondence between edge inversion and coherent results Diagram showing correspondence between frequency difference and non-coherent result Diagram showing the relationship between non-coherent time and edge detection time Diagram for explaining the difference in noise contribution at the addition section boundary Diagram for explaining edge error rate

Explanation of symbols

101 antenna, 102 down converter, 103 A / D converter 1 1 ms correlation channel circuit, 2 carrier NCO, 3 addition start time selector,
41-4J edge detection adder circuit, 51-5J frequency converter,
611-6JK Nms coherent adder,
811-8JK M times non-coherent adder,
911 to 9JK L times non-coherent adder, 10 edge detection circuit controller

Claims (3)

A carrier local oscillator that receives a frequency control signal and generates a carrier frequency signal, converts a satellite signal including navigation data and receives a code correlation between a received signal and a PN code related to a specific satellite signal, and the carrier frequency signal A 1 ms correlation channel circuit that takes carrier correlation and outputs an I correlation signal and a Q correlation signal for 1 ms simultaneously every 1 ms;
A plurality of J frequency converters for calculating and outputting a frequency-converted I-correlation signal and a frequency-converted Q-correlation signal that are frequency-converted with respect to different frequency candidates for the I correlation signal and the Q correlation signal,
The frequency conversion I correlation signal and the frequency conversion Q correlation signal from the plurality of J frequency converters are added to each of the addition sections corresponding to the plurality of K edge position candidate positions with respect to the edge position indicating the switching timing of the navigation data. A plurality of J * K coherent adders, each of which outputs an I coherent addition value and a Q coherent addition value,
The signal power for each coherent adder is obtained by receiving the I coherent addition value and the Q coherent addition value from the plurality of J * K coherent adders, and the non-coherent addition during that period for at least one predetermined period. A plurality of J * K non-coherent adders,
An edge detection circuit controller for simultaneously determining an edge position and a tracking frequency indicating the switching timing of navigation data from the addition results of the plurality of J * K non-coherent adders;
A satellite signal receiver.   An addition start timing selector for setting an addition section for detecting an edge position is provided between the 1 ms correlation channel circuit and the plurality of J frequency converters, and the setting of the addition section is performed by the edge detection circuit controller. The satellite signal receiver according to claim 1, wherein the satellite signal receiver is performed based on an addition start time instruction that indicates an addition interval from the satellite. When the edge position is determined by the edge detection circuit controller, among the coherent results for one frequency candidate k (k is any one of 1 to K), the maximum non-coherent result with respect to the second largest coherent result is obtained. The satellite signal receiver according to claim 1, wherein edge position determination is performed when the ratio is equal to or greater than a predetermined threshold value.