JPH04135308A - Hilbert transforming unit due to time region processing - Google Patents

Abstract

PURPOSE:To form a Hilbert transforming unit into a large scale integrated circuit (LSI) by digitizing signals and processing them in real time by a logic circuit. CONSTITUTION:When a digitized signal m(I) is ampplied to an input terminal 1 by an A/D converter, this is guided to shift registers 21-2N composed of an N pieces of elements. Clock pulses at 0.1 intervals are applied to the shift registers for shifting the signals, and outputs m(I-0.1) to m(I-0.1N) are respectively outputted from those elements. Coefficients k1/pi-kN/Npi are respectively mulitplied to these outputs by coefficient multipliers 31-3N. In such as case, k1-kN are defined as correcting values. All the results are guided to an adder 4, added and outputted to an output terminal 5. Therefore, the digital signal of an approximate Hilbert transform output m(I) is obtained from the terminal 5. Thus, the Hilbert transformer can be formed into LSI.

Description

[Detailed description of the invention] The present invention relates to a Hilbert transformer with time domain processing.

Conventionally, Hilbert transformers have been called all-pass networks, quadrature phase shifters, etc., and treated in the frequency domain. In this case, the value of the actance element in the constructed circuit becomes relatively large, making it difficult to integrate the circuit.

The present invention is a Hilbert converter LSI (Large Scale Integrated Circuit)
The aim is to realize this by increasing the number of 5SB.
-8C (carrier suppression single sideband method) modulation circuit LSI
The aim is to make this possible.

In order to achieve the above object, the present invention employs a method in which signals are digitized and processed in real time by a logic circuit. Since a logic circuit can be easily manufactured using an LSI, a Hilbert transformer based on time domain processing can be implemented as an LSI.

The objects and features of the present invention will become clear from the description of the embodiments with reference to the following drawings.

The Hilbert transform m(1) of the causal signal m(1) is given by the following integral.

m(t)−fo(m(u)/7r(t u))du=
, f o (m (tu)/ffu) du...
・・・・・・・・・(1) In this equation, when the time t is discretized, the integral value is very greatly influenced by t when t×1, so the Nyquist interval of the signal m(1) is set as the time l. For example, if the actual sampling clock is 01 and the discrete time ■ is written, then equation (1) becomes m(■)-〇, 1o=1Σm (I-0,1n) 1
0. 1 n T = n, Σm (1-0, L n ) /n 1r...
(2) However, it is sufficient to set the number N of sample values to 100. That is, the denominator of equation (2) is T when n = 1.

= 100, it is 100'Ir, and due to the causality of the signal, for n where l<0.1n, m(Io-tn) = 0
By being.

Based on equation (2), the circuit configuration of the Hilbert transformer is as shown in FIG. In the figure, 1 is the input terminal, and A/
When the signal m(1) digitized by the D-converter is applied, it is guided to a shift register 21, 22.degree. 2N consisting of N elements. Clock pulses at intervals of 0 and 1 are applied to the shift register for signal shifting, and each of these elements outputs m(1-0
,1),m(1-0,2), --,m(1-0,
I N ), which are also supplied with coefficient multipliers 31, 32 . ..., 3N makes the coefficient □/'Ir,
k/2'I'r, . . . , kN/N'Ir are respectively multiplied. However, k, 2. , kN are correction values. All of these results are led to adder 4, added, and output to output terminal 5. In this way, a digital signal with an approximate Hilbert transform output m(1) is obtained from the terminal 5.

The circuit configuration shown in FIG. 1 can be fabricated in its entirety as an LSI.

As explained above, the Hilbert converter of the present invention has the advantage of being able to realize the characteristics of LSI, such as small size, light weight, low power consumption, long life, and low cost.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of the configuration of a Hilbert transformer using time domain processing. 1: Digitized signal input terminal 21,...
2N=Shift register element 31,..., 3N: Coefficient multiplier

Claims (1)

[Claims] A signal can be approximated by digitizing the signal at the Nyquist interval or a sampling interval narrower than that, leading it to a shift register consisting of many elements, and multiplying the outputs of these elements by appropriate coefficients and adding them. A Hilbert transformer that realizes the Hilbert transform in a practical manner.