JP4968737B2 - Sampling frequency search method and program for undersampling for desired channel - Google Patents

Description

  The present invention relates to a sampling frequency search method and program for undersampling for a desired channel. In particular, the present invention relates to a technique for searching for a sampling frequency Fs for collectively sampling each desired channel of a plurality of system bands included in an RF (Radio Frequency) signal by undersampling.

  Undersampling refers to a method of intentionally generating an aliasing image by sampling a received signal at a frequency lower than the Nyquist frequency and converting a high-frequency carrier wave to a low frequency (see Non-Patent Document 1, for example). Since the sampling frequency is low, the processing capability of the receiver can be relatively low, but the design of the filter for noise removal becomes complicated. Conversely, oversampling refers to a method of sampling a received signal at a frequency higher than the Nyquist frequency. Since the sampling frequency is high, the design of a filter for noise removal can be simplified, but a high processing capability is required for the receiver.

  FIG. 1 is an image diagram of frequency components when undersampling is used.

  According to FIG. 1, the horizontal axis represents frequency. A white rectangle represents a system band, and a hatched mountain shape represents a desired channel. By sampling using the undersampling frequency Fs for the system band and the desired channel, aliasing occurs in a low frequency band below the frequency Fs / 2.

The code is defined as follows.
Fc: Center frequency of system band Fch: Center frequency of desired channel Fif: Center frequency of aliasing system band after sampling at sampling frequency Fs Fifch: Center frequency of aliasing desired channel after sampling at sampling frequency Fs BW: System bandwidth, bandwidth of aliasing system bandwidth BWch: desired channel width BH: frequency difference between highest boundary frequency of desired channel and frequency Fs / 2 BH = Fs / 2− (Fifch + BWch / 2)
BL: frequency difference between lowest boundary frequency of desired channel and frequency 0 BL = Fifch−BWch / 2

  FIG. 2 is a diagram illustrating the correspondence of aliasing images in undersampling.

Define the following functions:
fix (a): A function that rounds off the decimal part of a
rem (a, b): Function that obtains the remainder when a is divided by b

The following Fifch is derived for the sampling frequency Fs at the time of undersampling.
Nch = fix (Fch / (Fs / 2))
If Nch is an even number, Fifch = rem (Fch, Fs)
= Fch-Nch x (Fs / 2)
If Nch is an odd number, Fifch = Fs−rem (Fch, Fs)
= (Nch + 1) x Fs / 2-Fch

  The folding of the frequency signal occurs every half of the sampling frequency (Fs / 2). Accordingly, the number of times Nch at which the center frequency of the desired channel is folded is calculated by Fch / (Fs / 2) by the above-described equation. The calculated value is truncated by fix (). The position of the image in the aliasing system band varies depending on the even / odd value Nch obtained by the above equation.

  FIG. 2A shows the aliasing system band that is folded back to Fs / 2 or less when the number of times of folding Nch is an even number.

  FIG. 2B shows an aliasing system band that is folded back to Fs / 2 or less when the number of times of folding Nch is an odd number.

  As a technique using undersampling, there is a sampling frequency search method and program technique capable of searching for the lowest possible sampling frequency when sampling a plurality of wireless system bands for each system band (for example, Patent Document 1). reference). According to this technique, the lowest possible sampling frequency Fs can be searched, so that a high-speed A / D converter is not required, the amount of data for digital signal processing can be reduced, and power consumption is further reduced. Bring.

  There is also a technique for calculating a sampling frequency range in which a plurality of RF signals received simultaneously are not overlapped with each other after sampling by using undersampling (see, for example, Non-Patent Document 2). In this technique, cases are classified according to the positional relationship of folding caused by sampling of each RF signal, and an available sampling frequency is calculated from the number of undersampling and the frequency of the RF signal.

  Here, a general functional configuration of a receiver that samples a plurality of system bands collectively will be briefly described. The receiver includes an antenna, a low noise amplifier LNA (Low Noise Amplifier), a band pass filter BPF (Band Path Filter), and an A / D converter ADC (Analog / Digital Converter). The signal received by the antenna is amplified by a low noise amplifier, and only a predetermined band is extracted by a band pass filter. The A / D converter samples only the band by undersampling. Bandpass filters are provided in parallel for each system band.

Dennis M. Akos, Michael Stockmaster, JamesB. Y. Tsui and Joe Caschera "Direct Bandpass Sampling of Multiple Distinct RF Signals", IEEE Transactions on communications, vol.47, No.7, July 1999, pp.983-988 Ching-Hsiang Tseng and Sun-Chung Chou, “DirectDownconversion of Multiband RF Signals Using Bandpass Sampling,” IEEE Trans.Commun., Vol. 5, no. 1, Jan. 2006. JP 2006-180373 A

  According to the prior art, an algorithm that repeats the determination while increasing the sampling frequency by ΔFs in order from the lowest frequency until the sampling frequency that satisfies the condition is found is adopted (for example, a feature by the same applicant). See application 2006-161708). As ΔFs decreases, the number of repetitions increases, and as a result, the amount of calculation also increases. On the other hand, as ΔFs is increased, the possibility of missing a sampling frequency that can be undersampled increases.

  Therefore, the present invention searches for an available sampling frequency without increasing the amount of calculation when each desired channel in two system bands is sampled collectively using one A / D converter. It is an object of the present invention to provide a sampling frequency search method and program that can be used.

According to the present invention, there is provided a sampling frequency search method of a receiver for searching for a sampling frequency Fs for sampling each desired channel in a plurality of system bands included in an RF (Radio Frequency) signal by undersampling. There,
Sampling frequency range that can be used or not used for all aliasing system bands so that a aliased image of the aliasing system band based on frequency 0 and Fs / 2 does not overlap the desired channel of the aliasing system band on the frequency axis. A first step of setting Fs to Fs + ΔFs;
A second step of repeating the first step again by substituting Fs + ΔFs for Fs when the unusable sampling frequency range Fs to Fs + ΔFs is set;
Within the available sampling frequency range Fs to Fs + ΔFs,
A third step of calculating available or unusable sampling frequency ranges Fs to Fs + ΔFs so that all other aliasing system bands do not overlap on the frequency axis for every desired channel;
When the unusable sampling frequency range Fs to Fs + ΔFs is set, the fourth step is performed by substituting Fs + ΔFs into Fs and repeating the first step again.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
After performing the third step and the fourth step,
Within the available sampling frequency range Fs to Fs + ΔFs,
It is also preferable to perform the first step and the second step.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
The number of times the center frequency Fch of the desired channel is turned back by the sampling frequency Fs is Nch (= fix (Fch / (Fs / 2)), fix (): function for truncating the decimal point of the argument)
The first step is for each desired channel:
When Nch is an even number or an odd number, the aliasing system band based on the frequency 0 or Fs / 2 is used with the characteristic that the desired channel moves in the low frequency direction or the high frequency direction as the sampling frequency Fs increases. However, it is also preferable to calculate an available sampling frequency range when the desired channel does not overlap on the frequency axis, and to calculate an unusable sampling frequency range when it overlaps on the frequency axis.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
The desired channel bandwidth is BWch (Band Width channel), the difference between the highest boundary frequency and frequency Fs / 2 is BHch (Band High channel), and the difference between the lowest boundary frequency and frequency 0 is BLch (Band Low channel). channel),
The first step is for each desired channel:
When the number of turn-backs Nch is an even number,
Sampling available as ΔFsOK = (BL−BWL / 2) / (Nch / 2) when the aliasing system band aliasing image based on frequency 0 or Fs / 2 does not overlap the desired channel on the frequency axis Calculate the frequency range Fs ~ Fs + ΔFsOK,
When the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis, ΔFsL = (BL + BWch + BWH / 2) / (Nch / 2) is calculated, and the unusable sampling frequency range Fs to Fs + ΔFsL is calculated. And
When the aliasing system band aliasing image based on the frequency Fs / 2 overlaps with the desired channel on the frequency axis, ΔFsH = (BWH / 2−BH) / (Nch / 2 + 1/2), and an unusable sampling frequency The range Fs to Fs + ΔFsH is calculated,
When the number of times Nch is an odd number,
Sampling available as ΔFsOK = (BH−BWL / 2) / (Nch / 2) when the aliasing system band aliasing image based on frequency 0 or Fs / 2 does not overlap the desired channel on the frequency axis Calculate the frequency range Fs ~ Fs + ΔFsOK,
When the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis, ΔFsL = (BWH / 2−BL) / (Nch / 2 + 1/2), and an unusable sampling frequency range Fs ~ Fs + ΔFsL is calculated,
When the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis, ΔFsH = (BH + BWch + BWH / 2) / (Nch / 2), and an unusable sampling frequency range Fs to Fs + ΔFsH It is also preferable to calculate.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
The third step is
If you want to sample each desired channel in two system bands together,
The frequency shift amount of the aliasing system band per unit increment of the sampling frequency Fs is v (the high frequency direction is positive), the frequency shift amount v1 of the first desired channel, and the frequency shift of the second desired channel. The relative movement difference from the amount v2 is Δv (= v1−v2),
The interval from the second aliasing system band to the first desired channel is ΔFa, the interval from the second desired channel to the first aliasing system band is ΔFb,
When one or both of ΔFa or ΔFb is 0 or more,
When the movement amount relative difference Δv is 0 or more, the larger of BW2−ΔFa + BWch1 or BW1−ΔFb + BWch2 is set to ΔF, and ΔFs = ΔF / Δv, and the unusable sampling frequency range Fs to Fs + ΔFs is calculated.
When the movement amount relative difference Δv is smaller than 0, the larger of ΔFa and ΔFb is set as ΔF, and ΔFs = ΔF / (− Δv) is used to calculate the unusable sampling frequency range Fs to Fs + ΔFs,
When both ΔFa and ΔFb are smaller than 0, the larger of ΔFa and ΔFb is set as ΔF,
When the movement amount relative difference Δv is equal to or greater than 0, the available sampling frequency range Fs to Fs + ΔFsOK is calculated with ΔFsOK being the minimum value among (−ΔF) / Δv, ΔFsOK1 and ΔFsOK2.
When the movement amount relative difference Δv is smaller than 0, it is also preferable to calculate the available sampling frequency range Fs to Fs + ΔFsOK with ΔFsOK being the minimum value among ΔFsOK1 and ΔFsOK2.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
It is also preferable that any one of the first desired channel and the first aliasing system band or the second desired channel and the second aliasing system band is a band detected by oversampling.

According to another embodiment of the sampling frequency search method of the receiver of the present invention,
The third step is
When sampling each desired channel of three or more system bands at once,
It is also preferable to calculate an available sampling frequency range Fs to Fs + ΔFsOK for each combination of desired channels in two adjacent system bands, and to derive an overlapping sampling frequency range Fs to Fs + ΔFsOK in all these ranges.

According to the present invention, a frequency search program for executing a computer mounted on a receiver for searching for a sampling frequency Fs for sampling each desired channel of a plurality of systems included in an RF signal by batch undersampling. Because
Sampling frequency range that can be used or not used for all aliasing system bands so that a aliased image of the aliasing system band based on frequency 0 and Fs / 2 does not overlap the desired channel of the aliasing system band on the frequency axis. A first step of setting Fs to Fs + ΔFs;
A second step of repeating the first step again by substituting Fs + ΔFs for Fs when the unusable sampling frequency range Fs to Fs + ΔFs is set;
Within the available sampling frequency range Fs to Fs + ΔFs,
A third step of calculating available or unusable sampling frequency ranges Fs to Fs + ΔFs so that all other aliasing system bands do not overlap on the frequency axis for every desired channel;
When the unusable sampling frequency range Fs to Fs + ΔFs is set, Fs + ΔFs is substituted for Fs, and the computer is executed as the fourth step for repeating the first step again.

According to another embodiment of the sampling frequency search program for the receiver of the present invention,
After performing the third step and the fourth step,
Within the available sampling frequency range Fs to Fs + ΔFs,
It is also preferable to perform the first step and the second step.

According to the present invention, a receiver comprising:
A computer that executes the aforementioned program;
Sampling clock generating means for generating a sampling clock according to the sampling frequency Fs output from the computer;
An analog / digital conversion means for converting a received analog signal into a digital signal according to a sampling clock is provided.

  According to the sampling frequency search method and program of the present invention, even if each desired channel in two system bands is sampled collectively using one A / D converter, the amount of calculation is increased. Without being able to search for available sampling frequencies. This does not require a high-speed A / D converter, can reduce the amount of calculation of digital signal processing, and further reduces power consumption. As a result, the receiver can be downsized.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

  FIG. 3 is a functional configuration diagram of the receiver.

  FIG. 3 depicts a receiver that supports batch sampling of two desired channels. A signal received by the antenna 101 is amplified by a low noise amplifier 102, and two system bands are extracted by a band pass filter 103 for each system band. The signals of the two system bands are input to the A / D converter 104, and batch sampling is performed.

  The sampling frequency Fs of the A / D converter 104 is supplied from the sampling clock generator 109. The sampling frequency Fs is searched for by the CPU 110. The two desired channels detected by the sampling frequency Fs may be an aliasing system band based on undersampling, an aliasing system band based on undersampling, and a system band based on oversampling. Good.

  The two digital signals output from the A / D converter 104 are input to the signal separation unit 105 and separated. Each of the two digital signals is input to the quadrature demodulator 106. The quadrature demodulator 106 multiplies the input digital signal with the sine wave signal output from the numerically controlled oscillator NCO by two multipliers. Here, the oscillation frequency of the numerically controlled oscillator NCO is the same as the center frequency Fifch of the aliasing desired channel. The Q channel is multiplied by a sine wave whose phase is delayed by π / 2 radians by a phase shifter. As a result, each multiplier outputs a baseband signal of the I channel and Q channel of the channel.

  The baseband signals (I channel and Q channel) output from the quadrature demodulator 106 are each removed by the low-pass filter 107 from high frequency components generated by the quadrature demodulation. The signal output from each low-pass filter 107 is demodulated by the demodulator 108.

  Note that the processing of the flowchart for searching for the sampling frequency Fs described below is all executed by the CPU 110.

  FIG. 4 is a flowchart of the sampling frequency search method in the present invention.

  The processing of the present invention includes the following four steps. According to FIG. 4, the initial value of the sampling frequency Fs is provisionally determined as Fs = BW1 + BW2 + BWch1 + BWch2 which is the sum of two system bandwidths and two channel bandwidths. For example, as another embodiment, for example, the initial value Fs = a fixed value of 40 MHz, and Fs can be gradually decreased. The provisional Fs is set to a minimum value.

(First step)
(S401) A channel band determination process is executed for the first desired channel. The channel band determination process can be used so that the aliasing image of the first aliasing system band based on the frequency 0 or Fs / 2 does not overlap on the frequency axis with the first desired channel of the first aliasing system band or An unusable sampling frequency range Fs to Fs + ΔFs is set. For example, for the first desired channel, the available sampling frequency range is represented by ΔFsOK1, and the unavailable sampling frequency range is represented by ΔFsH1 and ΔFsL1.
(S402) For the second desired channel 2, a channel band determination process is executed as in S401. For example, for the second desired channel, the available sampling frequency range is represented by ΔFsOK2, and the unavailable sampling frequency range is represented by ΔFsH2 and ΔFsL2.

(Second step)
(S403) It is determined whether any one of ΔFsH1, ΔFsL1, ΔFsH2, or ΔFsL2 is set, that is, whether or not an unusable sampling frequency range Fs to Fs + ΔFs is set.
(S404) If any one of ΔFsH1, ΔFsL1, ΔFsH2, and ΔFsL2 is set, the maximum value is set to ΔFs. Then, Fs + ΔFs is substituted for Fs, and the process returns to S401 again.

(Third step)
(S405) When ΔFsH1, ΔFsL1, ΔFsH2, and ΔFsL2 are not set, that is, when the available sampling frequency ranges Fs to Fs + ΔFs are set, a sampling frequency range calculation process is executed. In this processing, the second aliasing system band does not overlap the first desired channel on the frequency axis within the available sampling frequency range Fs to Fs + ΔFs, and the first aliasing system band is the first. The usable or unusable sampling frequency range Fs to Fs + ΔFs is calculated so as not to overlap the two desired channels on the frequency axis.

(Fourth step)
(S406) When ΔFsOK is not set, that is, when the unusable sampling frequency range Fs to Fs + ΔFs is set, Fs + ΔFs is substituted for Fs, and the process returns to S401 again.

(S407) If ΔFsOK is set, Fs + ΔFs is determined as the sampling frequency. However, when another sampling frequency is searched for next time, Fs + ΔFs is substituted for Fs, and the process returns to S401 again.

  Note that the embodiment of replacing each step in the sequence of FIG. 4 described above is within the technical idea of the present invention. Here, after executing the third step and the fourth step, even if the first step and the second step are executed within the available sampling frequency range Fs to Fs + ΔFs, it is the same as the present invention. There is an effect. As another embodiment, after a single channel determination process in the first step, when an unusable sampling frequency range Fs to Fs + ΔFs is already set, Fs + ΔFs is substituted for Fs at that time. The first step can be repeated again. Specifically, after executing S401, when the unusable sampling frequency range Fs to Fs + ΔFs is set (S403), Fs + ΔFs is substituted for Fs, and S401 is repeated again.

  FIG. 5 is a flowchart of the channel band determination process.

  FIG. 6 is an image diagram showing movement of a desired channel according to ΔFs when Nch is an even number as compared to FIG.

  FIG. 7 is an image diagram showing movement of a desired channel according to ΔFs when Nch is an odd number, compared to FIG.

  Below, it demonstrates along the flowchart of FIG. 5, referring FIG.6 and FIG.7. 6 and 7, the movement vector is “positive / +” in the high frequency direction and “negative / −” in the low frequency direction.

(S501) First, the number of folds Nch in the desired channel is calculated as follows. Further, the center frequency Fifch of the desired channel is calculated as follows with respect to the sampling frequency Fs.
Nch = fix {Fch / (Fs / 2)}
When Nch is an even number,
Fifch = rem (Fch, Fs) = Fch−Nch × (Fs / 2)
When Nch is an odd number,
Fifch = Fs−rem (Fch, Fs) = (Nch + 1) × (Fs / 2) −Fch

(S502) Next, BH and BL are calculated for the desired channel as follows.
BH: frequency difference between the highest boundary frequency of the desired channel and the frequency Fs / 2 BH = Fs / 2− (Fifch + BWch / 2)
BL: frequency difference between lowest boundary frequency of desired channel and frequency 0 BL = Fifch−BWch / 2

(S503) It is determined whether the number of folds Nch is an even number or an odd number.

When Nch is an even number, if Fs is increased by ΔFs, it moves to the left as shown in FIG. 6, that is, in the direction of “negative / −”. At this time, the movement amount of the center frequency Fifch of the desired channel moves in the “negative” direction by the following ΔFifch.
ΔFifch = − (Nch / 2) × ΔFs

When Nch is an odd number, if Fs is increased by ΔFs, it moves rightward as shown in FIG. At this time, the movement amount of the center frequency Fifch of the desired channel moves by the following ΔFifch in the “positive” direction.
ΔFif = ((Nch1 + 1) / 2) × ΔFs

(S504) If Nch is an even number, it is determined whether or not BH ≧ BWH / 2 (the highest boundary frequency of the desired channel is equal to or lower than the lowest boundary frequency of the aliasing system band aliasing image based on Fs / 2). To do. When BH ≧ BWH / 2, the aliasing system band aliasing image based on the frequency Fs / 2 does not overlap the desired channel on the frequency axis.

(S505) Next, if BH ≧ BWH / 2, is BL ≧ BWL / 2 (the minimum boundary frequency of the desired channel is equal to or higher than the maximum boundary frequency of the aliasing system band aliasing image based on frequency 0)? Determine whether or not.

(S506) When BL ≧ BWL / 2 (YES), the aliasing system band aliasing image based on frequency 0 does not overlap the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsOK based on the following ΔFsOK can be used.
ΔFsOK = (BL−BWL / 2) / (Nch / 2)
By increasing Fs, the desired channel moves to the left as shown in FIG. The desired channel can be moved to a boundary position that does not overlap the aliasing system band aliasing image based on the frequency 0 on the frequency axis. Based on Fs + ΔFs at that position, an available sampling frequency range Fs to Fs + ΔFsOK is calculated.

(S507) When BL ≧ BWL / 2 is not satisfied (NO, BL <BWL / 2), the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsL based on the following ΔFsL cannot be used.
ΔFsL = (BL + BWch + BWH / 2) / (Nch / 2)
By increasing Fs, the desired channel moves to the left as shown in FIG. While the aliasing system band aliasing image based on the frequency 0 overlaps the desired channel on the frequency axis, the sampling frequency range Fs to Fs + ΔFsL between them is determined to be unusable.

(S508) When BH ≧ BWH / 2 is not satisfied (BH <BWH / 2), the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsH based on the following ΔFsH cannot be used.
ΔFsH = (BWH / 2−BH) / (Nch / 2 + 1/2)
By increasing Fs, the desired channel moves to the left as shown in FIG. While the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis, it is determined that the sampling frequency range Fs to Fs + ΔFsH therebetween cannot be used.

(S514) If Nch is an odd number, it is determined whether BL ≧ BWH / 2 (the lowest boundary frequency of the desired channel is equal to or higher than the highest boundary frequency of the aliasing system band aliasing image based on frequency 0). . When BL ≧ BWH / 2, the aliasing system band aliasing image based on the frequency 0 does not overlap the desired channel on the frequency axis.

(S515) Next, when BL ≧ BWH / 2, BH ≧ BWL / 2 (the highest boundary frequency of the desired channel is equal to or lower than the lowest boundary frequency of the aliasing system band aliasing image based on the frequency Fs / 2). It is determined whether or not there is.

(S516) When BH ≧ BWL / 2 (YES), the aliasing system band aliasing image based on the frequency Fs / 2 does not overlap the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsOK based on the following ΔFsOK can be used.
ΔFsOK = (BH−BWL / 2) / (Nch / 2)
By increasing Fs, the desired channel moves to the right as shown in FIG. The desired channel can be moved to a boundary position that does not overlap the aliasing system band aliasing image based on the frequency Fs / 2 on the frequency axis. Based on Fs + ΔFs at that position, an available sampling frequency range Fs to Fs + ΔFsOK is calculated.

(S517) When BH ≧ BWL / 2 is not satisfied (NO, BH <BWL / 2), the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsH based on the following ΔFsH cannot be used.
ΔFsH = (BH + BWch + BWH / 2) / (Nch / 2)
By increasing Fs, the desired channel moves to the right as shown in FIG. While the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis, it is determined that the sampling frequency range Fs to Fs + ΔFsH therebetween cannot be used.

(S518) When BL ≧ BWH / 2 is not satisfied (BL <BWH / 2), the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis. The desired channel is at the position of the image in FIG. At this time, it is determined that the sampling frequency range Fs to Fs + ΔFsL based on the following ΔFsL cannot be used.
ΔFsL = (BWH / 2−BL) / (Nch / 2 + 1/2)
By increasing Fs, the desired channel moves to the right as shown in FIG. While the aliasing system band aliasing image based on the frequency 0 overlaps the desired channel on the frequency axis, the sampling frequency range Fs to Fs + ΔFsL between them is determined to be unusable.

  FIG. 8 is a flowchart of the sampling frequency range calculation process.

  FIG. 9 is an image diagram showing movement of two desired channels in S801 of FIG.

  FIG. 10 is an image diagram showing a positional relationship of desired channels when ΔFa and ΔFb are positive or negative.

  Below, it demonstrates along the flowchart of FIG. 8, referring FIG.9 and FIG.10.

(S800) First, the relationship of Fifch1 <Fifch2 is established. That is, if Fifch1> Fifch2, the first aliasing system band and the first desired channel are switched with the second aliasing system band and the second desired channel. That is, Fifch1 and Fifch2 are interchanged, and of course, BWch1 and BWch2 are also interchanged.

(S801) Let v be the frequency shift amount of the desired channel per unit increment of the sampling frequency Fs. The frequency shift amount is expressed as follows.
When Nch is an even number: v = −Nch / 2
When odd: v = Nch / 2 + 1/2

  Then, a movement amount relative difference Δv (= v1−v2) between the frequency movement amount v1 of the first desired channel and the frequency movement amount v2 of the second desired channel is calculated. In addition, an interval from the second aliasing system band to the first desired channel is ΔFa, and an interval from the second desired channel to the first aliasing system band is ΔFb.

  Referring to FIG. 9, ΔFa and ΔFb are calculated as follows.

According to FIG. 9A, when Nch1 = even and Nch2 = even, by increasing the sampling frequency Fs, the first desired channel moves to the left (low frequency direction), and the second desired The channel also moves in the left direction (low frequency direction).
ΔFa = (Fifch1 + BWch1 / 2)-(Fifch2-BWch2 // 2-BWL2)
ΔFb = (Fifch1 + BWch1 / 2 / BWH1) − (Fifch2−BWch2 / 2)

According to FIG. 9B, when Nch1 = odd number and Nch2 = even number, the first desired channel moves to the right (high frequency direction) by increasing the sampling frequency Fs, and the second desired channel The channel moves in the left direction (low frequency direction).
ΔFa = (Fifch1 + BWch1 / 2)-(Fifch2-BWch2 // 2-BWL2)
ΔFb = (Fifch1 + BWch1 / 2 / BWL1) − (Fifch2−BWch2 / 2)

According to FIG. 9C, when Nch1 = even and Nch2 = odd, the first desired channel moves to the left (low frequency direction) by increasing the sampling frequency Fs, and the second desired The channel moves to the right (high frequency direction).
ΔFa = (Fifch1 + BWch1 / 2)-(Fifch2-BWch2 // 2-BWH2)
ΔFb = (Fifch1 + BWch1 / 2 / BWH1) − (Fifch2−BWch2 / 2)

According to FIG. 9D, when Nch1 = odd and Nch2 = odd, the first desired channel moves to the right (high frequency direction) by increasing the sampling frequency Fs, and the second desired The channel also moves in the right direction (high frequency direction).
ΔFa = (Fifch1 + BWch1 / 2)-(Fifch2-BWch2 // 2-BWH2)
ΔFb = (Fifch1 + BWch1 / 2 / BWL1) − (Fifch2−BWch2 / 2)

(S802) It is determined whether one or both of ΔFa and ΔFb is 0 or more.

(S803) When one or both of ΔFa and ΔFb are equal to or greater than 0, the two desired channels are at the position of the image in FIG. That is, the second aliasing system band overlaps the first desired channel on the frequency axis, or the first aliasing system band overlaps the second desired channel on the frequency axis. At this time, it is next determined whether or not the movement amount relative difference Δv is 0 or more.

(S804) When one or both of ΔFa and ΔFb is 0 or more and the movement amount relative difference Δv is 0 or more, the larger of BW2−ΔFa + BWch1 or BW1−ΔFb + BWch2 is set to ΔF. Then, an unusable sampling frequency range Fs to Fs + ΔFs is calculated as ΔFs = ΔF / Δv. By increasing Fs, the positions of the two desired channels are moved so that one aliasing system band does not overlap the other desired channel on the frequency axis.

(S805) If one or both of ΔFa and ΔFb is equal to or greater than 0 and the movement amount relative difference Δv is smaller than 0, the larger ΔFa or ΔFb is defined as ΔF. Then, the unusable sampling frequency range Fs to Fs + ΔFs is calculated as ΔFs = ΔF / (− Δv).

(S812) If both ΔFa and ΔFb are smaller than 0, the two desired channels are at the position of the image in FIG. 10B. That is, the second aliasing system band does not overlap with the first desired channel on the frequency axis, and the first aliasing system band does not overlap with the second desired channel on the frequency axis. At this time, the larger of ΔFa or ΔFb is defined as ΔF. That is, the smaller absolute value is ΔF.

(S813) Next, it is determined whether the movement amount relative difference Δv is 0 or more.

(S814) When both ΔFa and ΔFb are smaller than 0 and the movement amount relative difference Δv is 0 or more, the minimum value among (−ΔF) / Δv, ΔFsOK1, and ΔFsOK2 can be used as ΔFsOK. The sampling frequency range Fs to Fs + ΔFsOK is calculated. Even if Fs is increased, a range in which one aliasing system band does not overlap the other desired channel on the frequency axis is derived.

(S815) When both ΔFa and ΔFb are smaller than 0 and the relative displacement Δv is smaller than 0, the available sampling frequency range Fs to Fs + ΔFsOK is set with ΔFsOK being the minimum value among ΔFsOK1 and ΔFsOK2. calculate.

  According to the above-described embodiment of the present invention, the calculation of the sampling frequency when sampling each desired channel of the two system bands collectively has been described. According to the present invention, it is also possible to calculate a sampling frequency when sampling each desired channel of three or more system bands at once using these embodiments.

  FIG. 11 is an explanatory diagram of a calculation order when sampling each desired channel of three or more system bands at once.

  According to FIG. 11, three system bands are represented. Here, the available sampling frequency range Fs to Fs + ΔFsOK is calculated for each combination of desired channels in two adjacent system bands. According to FIG. 11, the sampling frequency range that can be used between the system band 1 and the system band 2 is calculated, and then the sampling frequency range that can be used between the system band 2 and the system band 3 is calculated. . Then, the overlapping range of these two available sampling frequency ranges can be calculated as the available sampling frequency range Fs to Fs + ΔFsOK.

  As described above in detail, according to the sampling frequency search method and program of the present invention, each desired channel in two system bands is sampled collectively using one A / D converter. Even in such a case, an available sampling frequency can be searched without increasing the amount of calculation. This does not require a high-speed A / D converter, can reduce the amount of calculation of digital signal processing, and further reduces power consumption. As a result, the receiver can be downsized. Such a sampling frequency search method and program have particular application to an oscillation frequency search circuit that searches for a sampling frequency.

  In the various embodiments of the present invention described above, various changes, modifications, and omissions in the scope of the technical idea and the viewpoint of the present invention can be easily made by those skilled in the art. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

It is an image figure of the frequency component in the case of using undersampling. It is a corresponding explanatory view of an aliasing image in undersampling. It is a functional block diagram of a receiver. It is a flowchart of the sampling frequency search method in this invention. It is a flowchart of a channel band determination process. FIG. 6 is an image diagram showing movement of the aliasing system band according to ΔFs when Nch is an even number. FIG. 6 is an image diagram showing movement of the aliasing system band according to ΔFs when Nch is an odd number. It is a flowchart of a sampling frequency range calculation process. FIG. 9 is an image diagram showing movement of two desired channels in S801 of FIG. It is an image figure showing the positional relationship of a desired channel about the case where (DELTA) Fa and (DELTA) Fb are positive or negative. It is explanatory drawing of the calculation order in the case of sampling each desired channel of three or more system bands collectively.

Explanation of symbols

101 antenna 102 low noise amplifier, LNA
103 Band pass filter, band pass filter, BPF
104 A / D converter, ADC
105 Signal Separator 106 Orthogonal Demodulator 107 Low-Pass Filter, Low-Pass Filter 108 Demodulator 109 Sampling Clock Generator 110 CPU, Computer

Claims (10)

A receiver sampling frequency search method for searching for a sampling frequency Fs for sampling each desired channel in a plurality of system bands included in an RF (Radio Frequency) signal by undersampling,
Sampling frequency range that can be used or not used for all aliasing system bands so that a aliased image of the aliasing system band based on frequency 0 and Fs / 2 does not overlap the desired channel of the aliasing system band on the frequency axis. A first step of setting Fs to Fs + ΔFs;
A second step of repeating the first step again by substituting Fs + ΔFs for Fs when the unusable sampling frequency range Fs to Fs + ΔFs is set;
Within the available sampling frequency range Fs to Fs + ΔFs,
A third step of calculating available or unusable sampling frequency ranges Fs to Fs + ΔFs so that all other aliasing system bands do not overlap on the frequency axis for every desired channel;
A sampling frequency search method for a receiver, comprising: a fourth step of repeating the first step by substituting Fs + ΔFs into Fs when the unusable sampling frequency range Fs to Fs + ΔFs is set . After performing the third step and the fourth step,
Within the available sampling frequency range Fs to Fs + ΔFs,
The receiver sampling frequency search method according to claim 1, wherein the first step and the second step are executed. The number of times the center frequency Fch of the desired channel is turned back by the sampling frequency Fs is Nch (= fix (Fch / (Fs / 2)), fix (): a function for truncating the decimal part of the argument)
The first step is for each desired channel:
When the Nch is even or odd, the aliasing system band based on the frequency 0 or Fs / 2 is used by using the characteristic that the desired channel moves in the low frequency direction or the high frequency direction as the sampling frequency Fs increases. When the aliasing image does not overlap the desired channel on the frequency axis, an available sampling frequency range is calculated. Conversely, when the aliasing image overlaps on the frequency axis, an unusable sampling frequency range is calculated. The sampling frequency search method of the receiver of Claim 1 or 2. The desired channel bandwidth is BWch (Band Width channel), the difference between the highest boundary frequency and frequency Fs / 2 is BHch (Band High channel), and the difference between the lowest boundary frequency and frequency 0 is BLch (Band Low channel). channel),
The first step is for each desired channel:
When the number of turn-backs Nch is an even number,
Sampling available as ΔFsOK = (BL−BWL / 2) / (Nch / 2) when the aliasing system band aliasing image based on frequency 0 or Fs / 2 does not overlap the desired channel on the frequency axis Calculate the frequency range Fs ~ Fs + ΔFsOK,
When the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis, ΔFsL = (BL + BWch + BWH / 2) / (Nch / 2) is calculated, and the unusable sampling frequency range Fs to Fs + ΔFsL is calculated. And
When the aliasing system band aliasing image based on the frequency Fs / 2 overlaps with the desired channel on the frequency axis, ΔFsH = (BWH / 2−BH) / (Nch / 2 + 1/2), and an unusable sampling frequency The range Fs to Fs + ΔFsH is calculated,
When the number of times Nch is an odd number,
Sampling available as ΔFsOK = (BH−BWL / 2) / (Nch / 2) when the aliasing system band aliasing image based on frequency 0 or Fs / 2 does not overlap the desired channel on the frequency axis Calculate the frequency range Fs ~ Fs + ΔFsOK,
When the aliasing system band aliasing image based on frequency 0 overlaps the desired channel on the frequency axis, ΔFsL = (BWH / 2−BL) / (Nch / 2 + 1/2), and an unusable sampling frequency range Fs ~ Fs + ΔFsL is calculated,
When the aliasing system band aliasing image based on the frequency Fs / 2 overlaps the desired channel on the frequency axis, ΔFsH = (BH + BWch + BWH / 2) / (Nch / 2), and an unusable sampling frequency range Fs to Fs + ΔFsH The receiver sampling frequency search method according to claim 3, wherein: The third step is
If you want to sample each desired channel in two system bands together,
The frequency shift amount of the aliasing system band per unit increment of the sampling frequency Fs is v (the high frequency direction is positive), the frequency shift amount v1 of the first desired channel, and the frequency shift of the second desired channel. The relative movement difference from the amount v2 is Δv (= v1−v2),
The interval from the second aliasing system band to the first desired channel is ΔFa, the interval from the second desired channel to the first aliasing system band is ΔFb,
When one or both of ΔFa or ΔFb is 0 or more,
When the movement amount relative difference Δv is 0 or more, the larger of BW2−ΔFa + BWch1 or BW1−ΔFb + BWch2 is set to ΔF, and ΔFs = ΔF / Δv, and the unusable sampling frequency range Fs to Fs + ΔFs is calculated.
When the movement amount relative difference Δv is smaller than 0, the larger of ΔFa and ΔFb is set as ΔF, and ΔFs = ΔF / (− Δv) is used to calculate the unusable sampling frequency range Fs to Fs + ΔFs,
When both ΔFa and ΔFb are smaller than 0, the larger of ΔFa and ΔFb is set as ΔF,
When the movement amount relative difference Δv is equal to or greater than 0, the available sampling frequency range Fs to Fs + ΔFsOK is calculated with ΔFsOK being the minimum value among (−ΔF) / Δv, ΔFsOK1 and ΔFsOK2.
5. The receiver according to claim 4, wherein when the movement amount relative difference Δv is smaller than 0, an available sampling frequency range Fs to Fs + ΔFsOK is calculated with ΔFsOK being the minimum value among ΔFsOK1 and ΔFsOK2. Sampling frequency search method.   The first desired channel and the first aliasing system band or the second desired channel and the second aliasing system band are bands detected by oversampling. Sampling frequency search method for receivers. The third step is
When sampling each desired channel of three or more system bands at once,
The available sampling frequency range Fs to Fs + ΔFsOK is calculated for each combination of desired channels in two adjacent system bands, and the sampling frequency range Fs to Fs + ΔFsOK that overlaps in all these ranges is derived. The reception sampling frequency search method according to claim 5 or 6. A frequency search program for causing a computer mounted on a receiver to search for a sampling frequency Fs for sampling each desired channel of a plurality of systems included in an RF signal by undersampling collectively,
Sampling frequency range that can be used or not used for all aliasing system bands so that a aliased image of the aliasing system band based on frequency 0 and Fs / 2 does not overlap the desired channel of the aliasing system band on the frequency axis. A first step of setting Fs to Fs + ΔFs;
A second step of repeating the first step again by substituting Fs + ΔFs for Fs when the unusable sampling frequency range Fs to Fs + ΔFs is set;
Within the available sampling frequency range Fs to Fs + ΔFs,
A third step of calculating available or unusable sampling frequency ranges Fs to Fs + ΔFs so that all other aliasing system bands do not overlap on the frequency axis for every desired channel;
Sampling for a receiver characterized in that, when an unusable sampling frequency range Fs to Fs + ΔFs is set, Fs + ΔFs is substituted for Fs and the computer is executed again as a fourth step to repeat the first step. Frequency search program. After performing the third step and the fourth step,
Within the available sampling frequency range Fs to Fs + ΔFs,
The receiver sampling frequency search program according to claim 8, wherein the first step and the second step are executed. A computer that executes the program according to any one of claims 8 and 9,
Sampling clock generating means for generating a sampling clock according to the sampling frequency Fs output from the computer;
A receiver comprising analog / digital conversion means for converting a received analog signal into a digital signal in accordance with the sampling clock.

Images