JPH05300091A - Voice signal coding and decoding device in mobile body communication equipment - Google Patents

Abstract

PURPOSE:To transmit an excellent voice signal without discontinuity when a code transmission line is switched on the way of communication in the mobile body communication equipment. CONSTITUTION:A control circuit 6 generates a control signal in response to a switching timing of a code transmission line and feeds it to an initializing circuit 7. The initializing circuit 7 initializes a memory content in a voice signal coding circuit 5 in response to the control signal. Therefore since the memory content in the coding circuit is initialized by the transmission line switching timing, the influence of a preceding voice code is eliminated and discontinuity is eliminated.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice signal coding / decoding device in a mobile communication device, and more particularly to a voice signal in a mobile communication device used in a cell type automobile telephone system in which switching of a transmission path is inevitable during communication. The present invention relates to a signal encoding / decoding device.

[0002]

2. Description of the Related Art An outline of a cell type automobile telephone system of this type will be described with reference to FIG. For the cell type mobile telephone system, reference can be made to "Automobile Telephone", supervised by Moruji Kuwahara, published by IEICE (Reference 1).

First, it is assumed that the mobile unit 1 is performing voice communication via the code transmission path A and the wireless station 2. The input voice from the voice input terminal IN is encoded in the voice signal encoding device 11 in the mobile unit 1 and transmitted to the wireless station 2 via the wireless device 12.

As this speech signal coding apparatus, "Cell
ular System Dual-Mode Mobile Station-Base Station
Compatibility Standard, EIA / TIA Interim Standard
54 ”section 2.1.3.3.2 (reference 2) and“ GSM FUL
L RATE SPEECH TRANSCODING ”(CEPT / CCH / GSM RECOMMEN
DATION 06.10,13 JULY 1989) (reference 3) can be referred to.

The voice code received by the radio 22 in the radio station 2 is decoded by the voice signal decoding device 22 into a voice signal, which is sent to the exchange 3. For the audio signal decoding device 22, refer to Section 2.2 of Document 2.

On the other hand, the voice signal sent from the exchange 3 is encoded by the voice signal encoder 23 in the radio station 2 and transmitted by the radio 21. The voice coded signal transmitted to the mobile unit 1 via the transmission path A is received by the wireless device 12, decoded by the voice signal decoding device 13 and becomes a reproduced voice signal, which is output from the voice output terminal OUT.

At this time, it is assumed that the mobile unit 1 moves to another place and reaches the position of the mobile unit 1 '. In this case, the code transmission path is switched so that the mobile body 1 performs the voice communication through the code transmission path B from the state where the voice communication is performed through the transmission path A.

The switching source radio station 2 constantly monitors the state of radio waves from the mobile unit 1. The state of the radio wave is generally monitored by monitoring the electric field strength, but other indicators such as the fading variation rate may be used.

When the condition of this radio wave deteriorates, the wireless station 2 requests the exchange 3 to switch from the current transmission path A to another transmission path. Upon receiving this request, the exchange 3 commands the radio station 4 to monitor the state of the radio wave from the mobile unit 1 (1 '). When the radio station 4 determines that the radio wave condition is better than that of the radio station 2, the exchange 3 generates a switching control signal for switching the channel to the mobile body via the radio station 2.

This control signal is detected by the control unit 14,
The wireless device 12 switches to a channel corresponding to the wireless station 4 by switching the wireless frequency. This wireless station 4 of the switching destination has the same configuration as the wireless station 2, and starts to send a signal corresponding to silence to the exchange 3 from a predetermined period before the switching timing at which the switching is performed, and the switching is performed and the voice code is wirelessly transmitted. When the voice signal is received by the machine 41, the reproduced voice signal is decoded by the voice signal decoding device 43 and sent to the exchange 3.

The voice signal decoding device 22 in the radio station 2 immediately stops the decoding operation and sends a signal corresponding to silence to the exchange 3 when the transmission path is switched and the voice code is not input.

In the exchange 3, the voice signal from the wireless station 2 is first selected and transmitted to the transmission line output terminal OUT, but the voice signal from the wireless station 2 is output for a certain period before and after the transmission line is switched. The signal and the audio signal from the wireless station 4 are combined and output to the transmission line output terminal OUT. Thereafter, the exchange 3 selects the signal from the wireless station 4 and selects the transmission line output terminal O
Output to UT.

The exchange 3 initially sends the voice signal input from the transmission line input terminal IN to the wireless station 2 which is the switching source, but operates to send it to the wireless station 4 to be switched.

In such a conventional system, the switching of the code transmission path is performed during the call regardless of the operation of the voice signal coding / decoding device, so that the voice signal is interrupted and the voice quality is deteriorated. Is effective.

[0015]

It is an object of the present invention to provide a voice signal coding / decoding device in a mobile communication device which can obtain a good voice signal even if the code transmission path is switched during a call. is there.

[0016]

Principle of the invention The audio signal decoded by the audio signal decoding device not only determines the reproduced audio signal depending only on the currently received encoded signal, but also reflects the influence of past received codes. .. The voice code encoded by the mobile voice signal encoding device is decoded by different voice signal decoding devices before and after the transmission path switching. At this time,
Since the voice signal decoding device at the switching destination cannot reflect the influence of the past received code, the voice signal becomes discontinuous. Therefore, when calculating the code to be sent to the new audio signal decoding device by switching the communication path, the mobile audio encoding device should be initialized in order to eliminate the influence of the past audio code. Encoding is performed.

Further, since the mobile voice signal decoding apparatus decodes the voice code encoded by the different voice signal encoding apparatus by switching the communication path, the influence of the past received code must be reflected correctly. But not. After the transmission path is switched, in order to artificially generate the influence of the past speech coding, the speech signal is extrapolated using the past reception code using the extrapolation circuit. .. Furthermore,
The extrapolated signal of the extrapolation circuit and the reproduced signal are weighted and combined to improve the perceptual quality.

[0018]

Based on this principle, the following structure can be obtained. First, a speech signal coding apparatus according to the present invention is a speech signal coding apparatus in a mobile communication device for coding an input speech signal according to a predetermined coding method and transmitting the speech coded signal to a transmission line, It is characterized by having an initialization means for initializing a memory circuit for voice encoding in response to the switching of the transmission path.

A voice signal decoding device according to the present invention is a voice signal decoding device for a mobile communication device, which receives coded voice information from a transmission path, decodes it according to a predetermined decoding method, and outputs a reproduced voice signal. An initialization device that initializes a memory circuit for voice decoding in response to the switching of the transmission line, and a past reception code in response to the switching of the transmission line. An extrapolation means for extrapolating a voice signal, and a synthesizing means for weighting and synthesizing an output of a decoding circuit for voice decoding and an extrapolation signal by the extrapolation means by predetermined weighting respectively. Characterize.

[0020]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic block diagram of an embodiment of a speech signal coding apparatus according to the present invention. Speech signal coding circuit 5
Receives a discrete voice signal from the input terminal 51, encodes it using a predetermined encoding method, and outputs a voice code to the output terminal 52. The encoding circuit 5 has the same configuration as the conventional encoding device 11 (see FIG. 3), and the configurations shown in the above-mentioned documents 2 and 3 can be used.

The control circuit 6 receives the code transmission line switching control signal from the control input terminal 53, and outputs the initialization control signal to the initialization circuit 7 at the timing when the code transmission line is switched. Upon receiving the initialization control signal, the initialization circuit 7 initializes a predetermined memory circuit in the audio signal encoding circuit 5.

The coding apparatuses in Documents 2 and 3 perform processing for each frame of a fixed time length, and send the coding result. The code is not determined only by the input audio signal in the current frame, but is affected by the processing result of the past frame. Therefore, the contents of the memory circuit for reflecting the processing result of the past frame are initialized.

Examples of this memory circuit include a memory for the short-term prediction filter and the short-term analysis filter, a memory for the long-term prediction filter and the long-term analysis filter, and a memory for storing the code of the previous frame. , These memory contents are initialized.

FIG. 2 is a schematic block diagram of an embodiment of a speech signal decoding apparatus according to the present invention. Audio signal decoding circuit 8
Outputs the reproduced audio signal to the weighting synthesis circuit 20 by decoding the received code input from the receiving terminal 81 using a predetermined decoding method. This audio signal decoding circuit 8 has the same structure as the conventional audio signal decoding device 13 (see FIG. 3), and the structures shown in the above-mentioned documents 2 and 3 can be used.

The control circuit 30 receives the code transmission line switching control signal from the control input terminal 83 and sends the extrapolation control signal to the extrapolation circuit 9 for a predetermined period of switching the code transmission line. The extrapolation circuit 9 calculates the extrapolated voice based on the past received code while the extrapolation control signal is input, and outputs the extrapolated voice to the weighting synthesis circuit 20.

Various widely known methods can be used as the extrapolation method of the audio signal. For example, “LOST F
RAME SUBSTITUTION AND MUTING (GSM FULL RATE SPEE
CH TRANSCODING) ”(CEPT / CCH / GSM RECOMMENDATI
The extrapolation method disclosed in ON 06.11,5 JULY 1989) (reference 4) can be adopted.

At the switching timing of the code transmission path,
The initialization control signal is output to the initialization circuit 10. Upon receiving the initialization control signal, the initialization circuit 10 initializes a predetermined memory circuit in the audio signal decoding circuit 8.

In the decoding device in Documents 2 and 3, the processing is performed for each frame of a fixed time length, and the contents of the memory storing the decoding data related to the past frame are initialized. ..

In the weighting synthesis circuit 20, the reproduced voice signal of the decoding circuit 8 is directly output to the output terminal 82 when the extrapolation control signal is not generated, and from the extrapolation circuit 9 during the generation period of the extrapolation control signal. The extrapolated signal and the reproduced audio signal from the decoding circuit 8 are weighted and combined and output to the output terminal 82.

Now, it is assumed that the reproduced voice signal from the decoding circuit 8 is x (n), the extrapolation signal from the extrapolation circuit 9 is y (n), and the output signal of the weighting synthesis circuit 20 is z (n). These x (n), y (n), z (n) are discrete signals,
n is an integer, and when the sampling period is T, the current time t is t = nT.

In the weighting synthesis circuit 20, the control circuit 30
When the extrapolation control signal is not generated, the reproduction signal x (n) of the decoding circuit 8 is output as it is as z (n) = x (n).

When the extrapolation control signal is generated, x (n) and y (n) are input, and the function z (n) shown below is output. That is, z (n) = f (n-n0) x (n) + {1-f (n-n0)} y (n) is a function, and the interpolation control signal is input to n = n0. It is assumed that

Here, f (n) represents a predetermined function, for example, n <0 ... f (n) = 0 0 ≦ n ≦ N ... f (n) = n / N n <n The function is such that f (n) = 1. N is a constant and indicates the time length of the weighted combination.

In the above embodiment, the extrapolation signal of the voice is obtained by the extrapolation circuit 9 by directly calculating from the past received code. Since it can be calculated using the intermediate result obtained in the voice decoding process or the reproduced voice signal, the data necessary for extrapolation may be sent from the voice signal decoding circuit 8 to the extrapolation circuit 9.

[0036]

As described above, according to the present invention, even if the code transmission path is switched during a call during the voice coding process of the mobile unit, the influence of the past voice code can be eliminated. Therefore, there is an effect that the quality of the audio signal does not deteriorate.

Further, even if the code transmission path is switched during the call during the voice decoding process of the mobile unit, the influence of the past voice code is artificially generated, so that the quality of the voice signal is improved. Has the effect of not decreasing.

[Brief description of drawings]

FIG. 1 is a block diagram of an audio signal encoding device according to an embodiment of the present invention.

FIG. 2 is a block diagram of an audio signal decoding device according to an embodiment of the present invention.

FIG. 3 is a block diagram of a mobile communication system to which an embodiment of the present invention is applied.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Mobile 2, 4 Radio station 3 Switch 5 Voice signal coding circuit 6,30 Control circuit 7, 10 Initialization circuit 8 Voice signal decoding circuit 9 Extrapolation circuit 20 Weighting synthesis circuit

Claims (2)

[Claims] 1. A voice signal encoding device in a mobile communication device for encoding an input voice signal according to a predetermined encoding method and transmitting the voice code signal to a transmission line, which is responsive to switching of the transmission line. Then, the speech signal coding apparatus is provided with an initialization means for initializing a memory circuit for speech coding. 2. A voice signal decoding device in a mobile communication device, which receives encoded voice information from a transmission path, decodes the encoded voice information according to a predetermined decoding method, and outputs a reproduced voice signal, wherein the transmission is performed. Initialization means for initializing a memory circuit for speech decoding in response to the switching of the channel; and extrapolation for extrapolating the speech signal using past received codes in response to the switching of the transmission channel. And a synthesizing means for synthesizing the output of the decoding circuit for speech decoding and the extrapolated signal by the extrapolating means by weighting by predetermined weighting, respectively. ..