JP3153693B2 - Dual tone signal transmission device - Google Patents

Abstract

PURPOSE:To reduce a noise of demodulation at the receiving side in the communication of a vector sum drive linear prediction coding system by superposing a prescribed noise on a dual tone signal produced based on dial information. CONSTITUTION:A tone dialer 3 receives the dial information from a key pad 1-1 of a handset 1 via a dial detector means 2 and produces a prescribed dual tone signal. A VSELP shows an information quantity compression system which is premised on the human voices. Therefore a narrow band signal like the dual tone signal increases noises when it is decoded. In this respect, a noise superposing means 4 adds a white noise, for example, received from a noise source 4-1 to the dual tone signal and sends these superposed signals to a VSELP coding/decoding device 5. So that the increase of a band is prevented when the compression coding is performed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a dual tone signal transmitting apparatus for a VSELP voice coding type mobile telephone which is a Japanese standard digital mobile telephone.

[0002]

2. Description of the Related Art VSELP (Vector Sum Driven Linear Prediction Coding: Vect)
or Sum Excited Linear Prediction) A speech coding method is a type of vector quantization that combines waveform coding of a speech signal and analysis-synthesis coding, and combines low bit rate coding with excellent voice quality. I have.

FIG. 4 shows a basic configuration of the VSELP system.
The VSELP method is characterized in that a driving sound source vector is represented by a linear combination of basic vectors. That is, a codebook is provided in the form of a basic vector having only a few bits, and a driving excitation vector is created by linear combination of the basic vectors. One bit is assigned to each basic vector depending on whether the polarity of each basic vector is plus or minus, a combination of 2 bits is realized, and the polarity itself is transmitted as a code. This reduces the amount of computation and memory,
At the same time, three improvements, that is, reduction in deterioration of transmission voice quality due to code errors, are achieved.

In the coding of VSELP, first, a reflection coefficient is obtained from a speech input, and the obtained reflection coefficient is used as a linear prediction coefficient (α parameter).
After that, interpolation is performed for each subframe (5 msec) while maintaining the stability of the linear prediction coefficient.

Next, a code book of the driving sound source vector is searched. The speech input is passed through an auditory weighting filter to be a reference signal. The state of the drive excitation vector and the pitch prediction filter are searched from the drive vector codebook and the pitch adaptive codebook, respectively. After multiplying the output obtained from the codebook by each gain, the combination of them becomes the driving sound source waveform, and the sum of the square of the error between the result of passing through the weighted vocal tract filter and the reference signal is minimized Such a code is selected.

[0006] VSELP decoding generates a driving excitation vector for a selected codebook code based on the code, and drives a vocal tract filter.

[0007] In the conventional VSELP speech coding system of the Japanese standard digital car telephone described above, no special consideration is given to the transmission of the dual tone signal.

[0008]

The above-mentioned conventional VSELP
In the audio coding method, codebook data optimized to transmit a human voice clearly and with a reduced number of bits is stored. In other words, VSEL
The P-speech coding system performs information capacity compression for efficiently transmitting only the voice of a human voice from all kinds of speech signals. For this reason, when a narrow band signal such as a dual tone signal is applied to the audio input, the driving sound source waveform combining the codes obtained from the codebook is not a narrow band but a wide band waveform, and Will be a noisy audio signal as shown by the dashed line in FIG. 3, and depending on the state of the transmission line, a message may be transmitted to a pager (pager) by key operation of the telephone, or a ticket may be transmitted. There has been a problem that communication with a telephone service computer for making reservations is hindered. In view of the above problems, an object of the present invention is to provide a VSE
It is an object of the present invention to provide a dual tone signal transmitting apparatus capable of encoding a dual tone signal with low noise in speech encoding in a mobile telephone adopting the LP speech encoding method.

[0009]

In order to solve the above-mentioned problems , the present invention provides a vector sum driven linear predictive coding (VSE).
In a dual tone signal transmitting device for a digital mobile phone of the LP) type, a dial detecting means for detecting dial information from a keypad, a tone dialer for converting dial information into a dual tone signal, and noise for adding noise to the dual tone signal And a noise superimposing means.
The stage is twice the frequency of the high frequency side of the dual tone signal.
Higher frequency bands and lower frequencies of dual tone signals
Noise is superimposed on the frequency band less than half of the side frequency
It is characterized in that it is a means to be.

[0010]

[0011]

A band of frequencies higher than twice the frequency of the high band side frequency of the working dual-tone signal by superimposing the noise and bandwidth of 1/2 or less of the frequency of the low band side frequency dual tone signal, Extend the spectrum of the input signal.
When the spectrum-spread signal is subjected to VSELP encoding, a driving excitation vector of a codebook having a small error from the input signal is searched. For this reason, in VSELP decoding, a dual tone signal having a small noise component is reproduced, and erroneous determination of key information in a communication partner device is eliminated.

[0012]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an embodiment of a digital mobile telephone of the VSELP encoding system equipped with a dual tone signal transmitting apparatus according to the present invention. In FIG. 1, a handset 1 having a keypad 1-1 is connected to a mobile phone main body 10. The mobile phone body 10 has a keypad 1-
Dial detecting means 2 for detecting dial information from 1
, Tone dialer 3, noise superimposing means 4, VSEL
A P code decoding device 5 is provided.

Next, the operation of this embodiment will be described. When any key on the keypad 1-1 of the handset 1 is pressed, the dial detecting means 2 transmits a signal corresponding to the pressed key to the tone dialer 3 and superimposes noise on the noise superimposing means 4. To instruct. The tone dialer 3 generates a predetermined dual tone signal based on a signal from the dial detecting means 2. The noise superimposing means 4 includes a dual tone signal from the tone dialer 3 and a noise source 4-
The white noise signal from No. 1 is added by the adding circuit 4-2 to generate a dual tone signal on which noise is superimposed.
The dual tone signal on which the noise is superimposed is VSELP encoded by the VSELP code decoding device 5.

FIG. 2 shows measured values showing the spectrum distribution of the reproduced audio signal when VSELP encoding and decoding are performed by adding sufficiently small white noise to the dual tone signal.
It is confirmed that the effect is remarkable when compared with the spectrum distribution of the reproduced audio signal at the time of encoding and decoding according to the conventional technique of FIG.

In this embodiment, a sufficiently small white noise is superimposed on the dual tone signal. However, the frequency band higher than twice the frequency of the high frequency side of the dual tone signal and the low frequency side of the dual tone signal are superimposed. The same effect can be obtained even when noise is superimposed on a band having a frequency equal to or less than half the frequency.

[0016]

As described above, according to the present invention, V
When transmitting a dual tone signal in a digital mobile telephone of the SELP voice coding system, noise is superimposed.
By spreading the spectrum of the dual-tone signal and then coding it, good coding can be performed, and there is an effect that a dual-tone signal with little noise is obtained in the decoded speech signal.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of a dual tone signal transmitting apparatus according to the present invention.

FIG. 2 is a spectrum distribution analysis diagram of a transmission / reception experiment using the dual tone signal transmission device according to the present invention.

FIG. 3 is a spectrum distribution analysis diagram of a transmission / reception experiment using a conventional dual tone signal transmission device.

FIG. 4 is a basic configuration diagram of the VSELP method.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Handset 1-1 Keypad 2 Dial detection means 3 Tone dialer 4 Noise superposition means 4-1 Noise source 4-2 Adder 5 VSELP code decoder 10 Mobile telephone main unit

Claims (1)

(57) [Claims] 1. A vector sum driven linear predictive coding (VSE)
In a dual tone signal transmitting device for a digital mobile phone of the LP) type, a dial detecting means for detecting dial information from a keypad, a tone dialer for converting dial information into a dual tone signal, and noise for adding noise to the dual tone signal And a noise superimposing means, the noise superimposing means having a higher frequency of the dual tone signal .
Frequency band higher than twice the frequency and dual tone
A dual tone signal transmitting device, which is means for superimposing noise on a band having a frequency equal to or less than 1/2 of a low frequency of a signal.