JP2553595B2 - Batch demodulator - Google Patents

Description

TECHNICAL FIELD The present invention relates to a batch demodulator for a plurality of FDM waves.

2. Description of the Related Art In a conventional batch demodulator, a sampling frequency is selected to be more than twice the highest frequency of a signal of interest according to a sampling theorem. For example, when n FDM waves of band ₀ per wave are continuously arranged, sampling is performed at a frequency of 2 × ₀ × n (for example, “multiple PSK wave batch demodulator using differential detection”). "Composition" IEICE Technical Report, CS84-41, "Application of Digital Signal Processing" supervision: Nobuo Inoue Publisher: IEICE / Corona Inc.).

Problems to be Solved by the Invention For reasons such as simplifying the selective filter of the receiver when there is a margin in frequency, FDM waves are not arranged consecutively but are arranged every other wave like a comb tooth. It often happens. In such a case, if the sampling frequency is selected according to the sampling theorem as in the case of the conventional batch demodulator, for example, the FDM wave of band ₀ per wave is arranged every other wave like the teeth of an n-wave comb. If so, sampling must be performed at 4 × ₀ × n. This is twice as large as the case where the same n waves are continuously arranged, and there is a problem in that the amount of calculation of digital signal processing after conversion into a digital signal is at least doubled.

The present invention has been made in view of the above point, and by appropriately selecting the sampling frequency, even when the FDM waves are arranged every other wave like the teeth of a comb, the N waves of the FDM waves continue. It is an object of the present invention to provide a collective demodulator capable of performing digital signal processing with an amount of calculation equivalent to that when arranged.

Means for Solving the Problems In order to solve the above problems, the present invention provides an A / D converter that performs sampling at a frequency half the frequency shown in the sampling theorem, and continuously performs sampling after sampling. N arranged
This is to separate and demodulate FDM waves individually.

According to the sampling theorem, if sampling is performed at frequency _S , _S /
The so-called "aliasing" in which the frequency spectrum is folded back in units of 2 occurs. Therefore, the sampling frequency is usually selected to be more than twice the highest frequency of the signal of interest. If sampling is performed at a frequency of 2 × ₀ × n when the FDM waves in the band ₀ are arranged every other wave like the teeth of an n-wave comb, the above-mentioned “alias phenomenon” occurs and ₀ × n
The frequency spectrum is repeated in units of. At this time, since the original FDM waves are arranged every other wave like comb teeth, the spectrum from ₀ × n to 2 × ₀ × n enters the gap between the spectra from ₀ to ₀ × n. , All signals included in frequencies 0 to 2 × ₀ × n before sampling from frequencies O to ₀ × n are included. Therefore, if signals of frequencies 0 to ₀ × n are regarded as signals in which n waves of FDM waves of band ₀ per wave are arranged consecutively and this FDM wave is separated and demodulated, the same as if the original signal was separated and demodulated. Result. At this time, the sampling frequency is 2 × ₀ × n, and the signal processing after being converted into a digital signal is the same as that in the case where n waves of FDM waves are continuously arranged. This is equivalent to the case where n waves are continuously arranged.

Embodiment FIG. 1 is a block diagram of an embodiment of the present invention. In FIG. 1, 200 has a sampling frequency of 8 ×
A / D converter selected as ₀ , 201 is a digital signal processing unit that collectively demodulates four consecutively arranged FDM waves, 202 is an input terminal for inputting the FDM waves,
203 to 206 are output terminals. Digital signal processing unit 201
Is composed of delay elements 211-214, sub-filters 221-224, FFT230 and the like. From the input terminal 202, as shown in FIG. 2A, n-wave (n-4) FDM waves of band ₀ per wave arranged every other wave like comb teeth are input. The FDM wave input from the input terminal 202 is converted into a digital signal by the A / D converter 200. At this time, the sampling frequency is 8 ×
Since it is selected as ₀ , aliasing occurs,
The output spectrum is as shown in FIG. 2B. Original FD
M waves are arranged every other wave like the teeth of a comb
The spectrum of from ₀ × 2 to ₀ × 8 from ₀ 0 × 4
All the signals included in the frequencies 0 to ₀ × 8 before sampling are included in the range 0 to ₀ × 4.

The digital signal processing unit 201 has a conventionally known 0
The output terminal 203 is a device for collectively demodulating four FDM waves sequentially arranged from ₀ to ₀ × 4 in units of ₀ .
From ~ 206, the separated and demodulated FDM waves are output. However, since the vector of the input signal is created by a synonymous phenomenon as shown in Fig. 2B, its spectrum is an odd-numbered FDM.
The waves have already reversed, was required for batch demodulator collectively demodulates the four waves FDM waves which are successively arranged a ₀ from 0 which has been known to ₀ × 4 units It is unnecessary to invert the spectrum of the odd-numbered signal among the output terminals 203 to 206.

In this embodiment, the case where the four FDM waves are collectively demodulated has been shown for simplicity, but it goes without saying that the FDM waves can be expanded to n waves in the same manner. Further, although there are various signal processing alternatives for the digital signal processing unit 201 surrounded by the dotted line in FIG. 1, the present invention can be applied to any alternatives for this portion.

EFFECTS OF THE INVENTION As described above, according to the present invention, sampling is performed at a frequency that is half the frequency shown in the sampling theorem, and after that, n FDM waves that are continuously arranged are obtained. By placing a batch demodulator for batch demodulation and separating and demodulating n FDM waves, even when the FDM waves in band ₀ are arranged every other wave like the teeth of an n-wave comb. Digital signal processing can be performed with an amount of calculation equivalent to that in the case where n FDM waves are continuously arranged, which is extremely useful in practice.
Also, when the real input type is adopted for the batch demodulator, the spectrum of the input signal of the batch demodulator is created by aliasing, so that the spectrum is naturally inverted in the odd-numbered FDM wave and There is also an effect that the process of inverting the spectrum of the signal becomes unnecessary.

[Brief description of drawings]

FIG. 1 is a block diagram of a batch demodulator in one embodiment of the present invention, and FIG. 2 is a characteristic diagram showing the relationship between the sampling frequency and the spectrum of the present invention. 200 ... A / D converter, 201 ... digital signal processing unit, 202 ... input terminal, 203-206 ... output terminal, 211 ... 21
4 …… Delay element, 221 …… 224 …… Sub filter, 230 ……
FFT.

Claims (1)

(57) [Claims] 1. An A / D converter collectively converts F-waves of n waves each having a band of ₀ , which are regularly arranged every other wave like comb teeth, into a digital signal, and then a digital signal. A demodulator for separating and demodulating each FDM wave by processing, wherein the sampling frequency of the A / D converter is selected to be 2 × n × ₀ lower than the frequency indicated by the sampling theorem. Demodulator.