KR101593185B1 - Codebook design method and apparatus - Google Patents

Abstract

The present invention relates to a method and a device for ensuring reliable data transmission by a code book design optimized for a WCDMA voice channel. The method includes the steps of: generating a code book including multiple symbols; selecting the symbol corresponding to an input signal; executing multiple transcoding operations for the selected symbol; estimating the input signal by obtaining the correlation between the symbols of the code book and the symbols having gone through the multiple transcoding operations; calculating an error rate of the symbols by comparing the input signal and the estimated signal; and executing a symbol adjustment operation by minimizing the symbol error rate.

Description

{Codebook design method and apparatus}

The present invention relates to a codebook design method and apparatus, and more particularly, to a method and apparatus for designing a codebook in a data transmission system for transmitting data through a WCDMA voice channel.

Data transmission technology through voice channel of mobile communication network has been studied for the purpose of secure communication, text service, user authentication, and so on.

Particularly, in the case of a voice channel using a speech coder, it is difficult for a general side / demodulation signal to pass normally due to the characteristics of the speech coder.

Recently, a method of constructing a transmission signal by mapping data to be transmitted to a codeword index in a codebook has been proposed. The key of this scheme is to find an optimal codebook composed of symbols that can pass through the voice channel to which the speech encoder is applied as an unknown system.

Here, the codebook search can be performed in various ways, ranging from the whole number search method for finding the optimal codebook to the pattern search and the genetic algorithm by searching all possible search intervals.

Although it is the most intuitive method, the size of the search space grows exponentially as the size of the codebook increases.

On the other hand, the pattern search method adopts the experience that it is effective to search the optimal solution while maintaining a certain pattern. The pattern search method is relatively short in search time compared to the random search and search based optimization method, and is more likely to find the global optimal solution than other optimization methods using the slope information.

The genetic algorithm is a computational model based on natural evolutionary processes and is a global optimization technique developed by John Holland in 1975. Genetic algorithms are known to be able to quickly find approximate solutions by mimicking the process of biological evolution.

The pattern search method and the genetic algorithm have a disadvantage that they can converge to the local minima instead of the global minima for the nonlinear dominant optimization problems such as the voice channel in which the speech coder exists.

In addition, since all of the existing studies focus on GSM speech coder, a codebook design method optimized for WCDMA voice channel is required to be applied to domestic mobile communication network.

In the related art, after the non-communication dongle is connected to the portable terminal, authentication between the transmission and reception side communication communication dongles is performed in a state in which the communication line is formed, and when the authentication is successful, the transmitting and receiving side communication dongles And the contents to perform decryption on the received data to provide non-communication via the non-communication communication dongle are described in Korean Patent Laid-Open Publication No. 2011-0085282 (apparatus and method for providing non-communication).

Other related prior art disclosures describing a method for transmitting data to a voice channel of a GSM network are described in IEEE Transactions on Vehicular Technology vol 57/4, pp 2205-2218, 2008 (entitled A Data Modem for GSM Voice Channel ).

SUMMARY OF THE INVENTION It is an object of the present invention to provide a method and apparatus for guaranteeing reliable data transmission through a codebook design optimized for a WCDMA voice channel.

According to an aspect of the present invention, there is provided a method of designing a codebook, the method comprising: generating a codebook including a plurality of symbols; The modulating unit selecting a symbol corresponding to the input signal; The speech encoder performing a plurality of transcoding operations on the selected symbol; Estimating the input signal by obtaining a degree of correlation between symbols of the plurality of times of transcoding and symbols of the codebook; Calculating a symbol error rate; calculating a symbol error rate by comparing the input signal and the estimated signal; And a symbol coordinating unit, performing symbol coordination so that the symbol error rate is minimized.

The step of estimating the input signal may estimate the input signal from an index of a symbol having a maximum degree of correlation among the degrees of correlation between the plurality of transcoded symbols and the symbols of the codebook.

Wherein the step of calculating the symbol error rate comprises the steps of:

(Where, v is a vector, and, N is a constant that indicates the size of the code book, is the i-th L _err (i) is the number of i-th symbol error, and L _tot (i) represents the component of the entire code book Symbol denotes the total number of symbols, and ∥ · ∥ denotes an L _∞ norm), the symbol error rate can be calculated.

In the step of performing symbol adjustment, minimax optimization based on SOCP may be performed so that the symbol error rate is minimized.

The minimax optimization is performed using the following equation

(Where L denotes the frequency spectrum of the symbol, L _err (i) denotes the number of errors of the i-th symbol, and L _tot (i) denotes the total number of the i-th symbol) have.

The step of performing the symbol adjustment may find a symbol pair having the largest symbol error rate among the symbols of the codebook and replace the symbol with another value within the search space limited to the voice band.

The transcoding may occur twice in the WCDMA voice channel to which the speech coder is applied.

Meanwhile, an apparatus for designing a codebook according to a preferred embodiment of the present invention includes: a codebook generator for generating a codebook including a plurality of symbols; A modulator for selecting a symbol corresponding to an input signal; A voice encoder for performing a plurality of transcoding operations on the selected symbol; A demodulator for estimating the input signal by obtaining a degree of correlation between a symbol passed through the speech encoder and symbols of the codebook; A symbol error rate calculator for calculating a symbol error rate by comparing the input signal with a signal estimated by the demodulator; And a symbol adjuster for performing symbol adjustment so that the symbol error rate is minimized.

According to the present invention having such a configuration, a codebook design robust to a WCDMA voice channel is possible.

The codebook designed according to the present invention is robust to a WCDMA voice channel having nonlinear characteristics. Therefore, the data transmission system using the codebook according to the present invention can guarantee reliable data transmission over the voice channel in the domestic communication network as well as the foreign communication network using the AMR encoder.

In addition, it is possible to provide a reliable encrypted communication by encrypting a voice signal and converting it into a data form, and then using the data transmission method according to the present invention.

FIG. 1 is a diagram illustrating a process in which a mobile communication provider of a transmitting side and a receiving side perform voice communication in different situations.
2 is a diagram for explaining a configuration of a codebook designing apparatus according to an embodiment of the present invention.
3 is a flowchart illustrating a codebook design method of a data transmission system according to an embodiment of the present invention.
4 is a diagram for explaining a data transmission process using an optimal codebook according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail.

It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the relevant art and are to be interpreted in an ideal or overly formal sense unless explicitly defined in the present application Do not.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In order to facilitate the understanding of the present invention, the same reference numerals are used for the same constituent elements in the drawings and redundant explanations for the same constituent elements are omitted.

First, a case where the method of the present invention is not applied will be described with reference to FIG. FIG. 1 is a diagram illustrating a process in which a mobile communication provider of a transmitting side and a receiving side perform voice communication in different situations.

In a voice call, the voice signal of the transmitting terminal 1 is encoded through an AMR (Adaptive Multi Rate) encoder 1a and transmitted to the base station 2. The base station 2 decodes the signal by the AMR decoder 2a and performs pulse code modulation (PCM) or adaptive differential pulse code modulation (DPC) in the converter 2b to transmit the digitized voice signal to the other base station 3 adaptive differential pulse code modulation (ADPCM).

The signal thus transmitted is encoded again by the base station 3 through the converter 3a and the AMR encoder 3b and transmitted to the receiving terminal 4. [ Then, it is decoded by the AMR decoder 4a of the receiving terminal 4 to output the voice of the sender.

In Figure 1, two transcoding steps are performed until the voice of the sender is heard at the end recipient. Here, the two transcoding means transcoding through the AMR encoder 1a and the AMR decoder 2a, and transcoding through the AMR encoder 3b and the AMR decoder 4a.

At this time, the quality of the voice heard by the sender and the receiver is not greatly different before and after the transcoding, but there are many differences when comparing the waveforms of the voice. For this reason, there is a problem in data transmission through a voice channel of a heterogeneous network in which transcoding occurs.

In order to solve such a problem, the present invention proposes a new codebook designing method in order to transmit reliable data through a voice channel.

2 is a diagram for explaining a configuration (framework) of a codebook designing apparatus according to an embodiment of the present invention.

A framework (basic concept structure) for an optimal codebook design includes a codebook generator 10, a modulator 12, AMR encoders 14 and 16, a demodulator 18, a symbol error rate calculator 20, And a symbol adjusting unit 22.

The codebook generator 10 can generate a codebook having N symbols of M sample length. Thus, assuming that the codebook has N symbols of M sample length, the size of the codebook is N × M. In the embodiment of the present invention, in order to reduce the amount of computation in the optimization process, a method of designing in the frequency domain and then converting it into the time domain is adopted. If the codebook design is performed in the time domain, the size of the search space becomes N × M, and if N is large, the search space becomes very large, so that it takes much time to design the optimum codebook. Assuming that a code word having M sample lengths is converted into a frequency domain and that the number of frequency bins (data storage spaces) corresponding to 300 to 3400 Hz is K, a random vector for a codebook design in a frequency domain The size is 2NK x 1. In order to generate a real codebook in the time domain from this vector, a complex conjugate is performed to fill the remaining frequency bin and an inverse fast Fourier transform (IFFT) is used to generate a time-domain real codebook.

The codebook generator 10 transmits the generated codebook to the modulator 12.

The modulator 12 selects a symbol corresponding to the input bit stream. Here, the input bitstream is randomly generated and input by the codebook design algorithm. On the other hand, the size of the codebook can be changed according to the target data rate, and the bit processing unit can be determined according to the selected codebook.

The AMR encoders 14 and 16 pass the symbol selected by the modulator 12. That is, since there are two AMR encoders, the symbol selected by the modulator 12 passes through the AMR encoder twice. In the embodiment of the present invention, since the mobile communication is assumed to be a general heterogeneous network in which the transcoding occurs twice, the input symbol is processed so as to pass through the AMR encoder twice. Here, it can be understood that the AMR encoders 14 and 16 are applied to a voice channel. The AMR encoder 14 includes the AMR encoder 1a and the AMR decoder 2a of FIG. 1 and the AMR encoder 16 includes the AMR encoder 3b and the AMR decoder 4a of FIG. 1 can see. The AMR encoders 14 and 16 may be called voice encoders, respectively.

The demodulator 18 calculates the correlation between the symbol passing through the AMR encoders 14 and 16 and the symbols of the codebook, as shown in Equation (1) below, and estimates the input signal from the index of the symbol having the maximum correlation.

Here, y (s _i ) denotes the signal after the symbol s _i has passed through the AMR encoder twice, and <.≫ denotes the inner product of the vector.

The symbol error rate calculator 20 compares the input signal (input bit stream) with the signal estimated by the demodulator 18 to calculate a symbol error rate.

The symbol adjusting unit 22 finds a symbol pair having the largest symbol error rate and adjusts the error rate of the corresponding symbol to be minimized.

In order to design a symbol constituting a codebook in a codebook designing framework, in the embodiment of the present invention, attention is paid to the fact that a symbol error rate increases as an overlapping portion between received symbols increases, I have chosen to find the symbol and minimize it. Equation (2) below is a cost function to achieve the above object.

Here, v is a vector representing a component of the entire codebook, N is a constant indicating the size of a codebook, L _err (i) denotes the number of errors of the i-th symbol, and L _tot (i) And ∥ · ∥ denotes the L _∞ norm.

The codebook designing method according to the embodiment of the present invention calculates the symbol error rate from the initial codebook value, optimizes the method by selecting the symbols having the largest error among them, and minimizing the symbols. This will be described in more detail with reference to FIG.

3 is a flowchart illustrating a codebook design method of a data transmission system according to an embodiment of the present invention.

First, the codebook generator 10 generates a codebook (S10). At this time, assuming that the generated codebook has N symbols of M sample length, the size of the codebook will be N × M. On the other hand, in order to reduce the amount of computation in the optimization process, the codebook is designed in the frequency domain and then transformed into the time domain. Then, the generated codebook is provided to the modulator 12. [

Next, the modulator 12 selects a symbol corresponding to the input signal (input bit stream) (S20).

The selected symbol passes through the AMR encoder twice. That is, the selected symbol passes through the AMR encoders 14 and 16 in order, and is encoded and decoded twice (S30).

The symbols passed through the AMR encoders 14 and 16 are transmitted to the demodulator 18. Accordingly, the demodulator 18 obtains the degree of correlation between the symbol passing through the AMR encoders 14 and 16 and the symbols of the codebook, and estimates the input signal from the index of the symbol having the maximum correlation (S40).

Then, the symbol error rate is calculated (S50). That is, the symbol error rate calculator 20 calculates the symbol error rate by comparing the input signal (input bit stream) with the signal estimated by the demodulator 18 (that is, the index of the symbol with the maximum correlation). As described above, the symbol error rate calculation unit 20 can calculate the symbol error rate through symbol decoding on the input signal continuously input.

Finally, the symbol adjusting unit 22 finds the symbol pair having the largest symbol error rate and adjusts the error rate of the corresponding symbol to be minimized (S60). In other words, the L _∞ norm of the symbol error rate calculated according to the cost function of Equation (2) is calculated, and the minimax optimization based on the second order cone programming (SOCP) is performed to minimize the L _∞ norm. Equation (3) below shows this.

Here, L denotes a frequency spectrum of a symbol, L _err (i) denotes a number of errors of an i-th symbol, and L _tot (i) denotes a total number of i-th symbols.

That is, the symbol adjusting unit 22 finds a symbol pair having the largest symbol error rate among the symbols of the codebook, and replaces the symbol with another value within the search space limited to the voice band. This process is repeated until the desired performance is reached or the maximum number of iterations is reached.

4 is a diagram for explaining a data transmission process using an optimal codebook according to an embodiment of the present invention.

The modulation section 12 of the waveform mapping system converts an input bit stream into a waveform codebook signal having characteristics robust to a voice channel. This method is a method of preliminarily configuring N symbols or waveforms, dividing an input bitstream into a predetermined size, and mapping the index of the input bitstream with the index.

The transmitting end transmits a symbol corresponding to the input bitstream according to a data rate.

The transmitted symbols are received at the receiving end (in particular, the demodulating unit 18) through the voice channel. Here, the voice channel considered in the present invention is a WCDMA voice channel using the AMR coders 14 and 16 used in the domestic mobile communication network.

The demodulator 18 finds a symbol having the highest correlation with the symbol received through the voice channel using a matched filter, and finds an index of the symbol and restores the bitstream.

As described above, an optimal embodiment has been disclosed in the drawings and specification. Although specific terms have been employed herein, they are used for purposes of illustration only and are not intended to limit the scope of the invention as defined in the claims or the claims. Therefore, those skilled in the art will appreciate that various modifications and equivalent embodiments are possible without departing from the scope of the present invention. Accordingly, the true scope of the present invention should be determined by the technical idea of the appended claims.

10: codebook generator 12:
14, 16: AMR encoder 18: demodulator
20: Symbol error rate calculation unit 22: Symbol adjustment unit

Claims (14)

A codebook generating section generates a codebook including a plurality of symbols;
The modulating unit selecting a symbol corresponding to the input signal;
The speech encoder performing a plurality of transcoding operations on the selected symbol;
Estimating the input signal by obtaining a degree of correlation between symbols of the plurality of times of transcoding and symbols of the codebook;
Calculating a symbol error rate; calculating a symbol error rate by comparing the input signal and the estimated signal; And
And performing symbol adjustment so that the symbol error rate is minimized. The method according to claim 1,
Wherein the step of estimating the input signal estimates the input signal from an index of a symbol having a maximum degree of correlation among the degrees of correlation between the symbols subjected to the plurality of times of transcoding and the symbols of the codebook, . The method according to claim 1,
Wherein the step of calculating the symbol error rate comprises the steps of:

(Where, v is a vector, and, N is a constant that indicates the size of the code book, is the i-th L _err (i) is the number of i-th symbol error, and L _tot (i) represents the component of the entire code book Denotes the total number of symbols, and ∥ · ∥ denotes the L _∞ norm)
And the symbol error rate is calculated by the symbol error rate. The method according to claim 1,
Wherein the step of performing the symbol adjustment comprises minimizing optimization based on the SOCP so that the symbol error rate is minimized. The method of claim 4,
The minimax optimization is performed using the following equation

(Where L denotes the frequency spectrum of the symbol, L _err (i) denotes the number of errors of the i-th symbol, and L _tot (i) denotes the total number of the i-th symbol)
Is defined as a codebook designing method. The method according to claim 1,
Wherein the step of performing the symbol adjustment is to find a symbol pair having the largest symbol error rate among the symbols of the codebook and to replace the symbol with another value within the search space limited to the voice band. The method according to claim 1,
Wherein the transcoding occurs twice in a WCDMA voice channel to which the speech encoder is applied. A codebook generator for generating a codebook including a plurality of symbols;
A modulator for selecting a symbol corresponding to an input signal;
A voice encoder for performing a plurality of transcoding operations on the selected symbol;
A demodulator for estimating the input signal by obtaining a degree of correlation between a symbol passed through the speech encoder and symbols of the codebook;
A symbol error rate calculator for calculating a symbol error rate by comparing the input signal with a signal estimated by the demodulator; And
And a symbol adjuster for performing symbol adjustment so that the symbol error rate is minimized. The method of claim 8,
Wherein the demodulator estimates the input signal from an index of a symbol having a maximum correlation among degrees of correlation between a symbol passed through the speech encoder and symbols of the codebook. The method of claim 8,
The symbol error rate calculator calculates the symbol error rate

(Where, v is a vector, and, N is a constant that indicates the size of the code book, is the i-th L _err (i) is the number of i-th symbol error, and L _tot (i) represents the component of the entire code book Denotes the total number of symbols, and ∥ · ∥ denotes the L _∞ norm)
And the symbol error rate is calculated by the symbol error rate calculating unit. The method of claim 8,
Wherein the symbol adjuster performs minimax optimization based on the SOCP so that the symbol error rate is minimized. The method of claim 11,
The minimax optimization is performed using the following equation

(Where L denotes the frequency spectrum of the symbol, L _err (i) denotes the number of errors of the i-th symbol, and L _tot (i) denotes the total number of the i-th symbol)
Wherein the codebook designing unit is defined as a codebook designing unit. The method of claim 8,
Wherein the symbol adjusting unit finds a symbol pair having the largest symbol error rate among the symbols of the codebook and replaces the symbol with another value within the search space limited to the voice band. The method of claim 8,
Wherein the transcoding occurs twice in a WCDMA voice channel to which the speech encoder is applied.

Images