KR100902842B1 - Fading channel modeling equipment of communication system and channel modeling method thereof - Google Patents

Abstract

A fading channel modeling device of a communication system and a channel modeling method thereof are provided to reduce a calculation quantity for performing a simulation of the communication system by interpolating a channel gain when beginning the simulation after calculating the channel gain before the simulation. A random noise passes through a fading filter and a complex channel gain of a frequency domain is produced(S10). The complex channel gain adding the fading effect is produced by filtering the random noise in the fading filter according to a fading factor(S20). The complex channel gain of the frequency domain is inverse-discrete-Fourier-transformed and the complex channel gain of the outputted time domain is stored(S30). The complex channel gain of the stored time domain is interpolated according to the interpolation factor and is provided to the complex channel gain for the simulation(S40).

Description

FADING CHANNEL MODELING EQUIPMENT OF COMMUNICATION SYSTEM AND CHANNEL MODELING METHOD THEREOF

The present invention relates to a communication system, and more particularly, to a channel modeling apparatus of a communication system and a channel modeling method thereof for testing a mobile environment of a wireless transmission / reception system.

The present invention is derived from the research conducted as part of the IT growth engine technology development project of the Ministry of Information and Communication and the Ministry of Information and Telecommunications Research and Development. [Task management number: 2006-S-017-02, Title: Development of advanced technology for terrestrial DMB transmission] .

Due to the development of the communication industry and increasing user demand for packet data services, there is an increasing need for a communication system capable of efficiently providing a high speed packet data service. Existing communication networks have been developed mainly for voice services, which have disadvantages of relatively small data transmission bandwidth and high usage fee. As a representative example of a broadband wireless access method for solving such disadvantages, research on an orthogonal frequency division multiplexing (hereinafter, referred to as OFDM) system is rapidly progressing. The OFDM scheme is widely applied to digital transmission technologies that require high-speed data transmission such as wideband wireless Internet, digital multimedia broadcasting (DMB), and wireless local area network (WLAN).

Various simulation models are provided to support the research and development of the aforementioned communication systems. In particular, the proper provision of channel modeling in a mobile environment is crucial to the duration or efficiency of research and development. A factor to be considered in channel modeling in a mobile environment is a fading effect. Fading is one of the biggest causes that affect the performance of communication systems in mobile wireless environments such as digital multimedia broadcasting and wireless LAN. The fading phenomenon is largely due to the multipath effect caused by reflection, absorption, refraction and scattering by buildings or other obstacles during signal transmission, and the frequency shifting effect of the received signal caused by the speed change of the moving object. Due to the multipath effect, the signals received through each path have different magnitudes and phases. According to the speed of the moving object, the received signals are accompanied by a change in envelope by causing time-varying mutual interference. Fading occurs as the envelope changes.

1 is a block diagram schematically illustrating a channel model of a general communication system. Referring to FIG. 1, the transmission signal s (t) through the channel experiences distortion corresponding to the channel gain h (t) and the addition of noise n (t). According to the influence on the channel, the received signal x (t) may be represented by Equation 1 below.

Where * is the convolution integrator

The channel gain h (t) is generally modeled as follows. First, independent random Gaussian noises are generated. The random Gaussian noise described above passes through a fading filter in order to receive a fading effect corresponding to the speed of the moving object. The fading filter may be configured as a digital filter controlled to have various channel characteristics corresponding to multipath, refraction, attenuation, speed change, and the like. The channel gain h (t) outputted through the fading filter and the transmission signal s (t) are convolutional integrated to add a fading effect of the channel. Then, channel n is applied to the received signal x (t) as noise n (t) is added.

According to the channel modeling method described above, various processes such as generation of random Gaussian noise, fading filtering processing, and convolution multiplication with a transmission signal should be performed for every sample of the transmission signal s (t). Therefore, in order to apply to each sample of the transmission signal s (t) using the modeling of the channel and the modeled channel gain, excessive computation and the burden of the accompanying simulation system are inevitable.

Simulation for testing channel performance in most wireless communication systems requires a lot of time. In particular, simulations involving fading channel models require large amounts of computation and simulation time. Therefore, minimizing the time required for fading channel modeling and increasing the computational operation speed of the channel gain have a great influence on the channel performance simulation of the entire transmit / receive system. Therefore, there is an urgent need for a channel modeling technique capable of reducing the amount of computation while providing channel gain at high speed.

The present invention provides a fading channel modeling apparatus and a channel modeling method which can generate a channel gain with a relatively small amount of computation during a simulation in a communication system and provide the channel gain at a high speed.

A fading channel modeling method of a communication system of the present invention for achieving the above object comprises the steps of: generating a complex channel gain in the frequency domain; Storing the complex channel gain of the time domain outputted by inverse Discrete Fourier Transform (IDFT) of the frequency domain in the memory; And interpolating the complex channel gain of the time domain stored in the memory to the communication system during a simulation operation of the communication system.

In this embodiment, generating a complex channel gain in the frequency domain comprises: generating fading coefficients that provide frequency characteristics of the first random noise and the second random noise and the fading filter independent of each other; Filtering each of the first random noise and the second random noise according to the fading coefficients; And multiplexing the filtered first random noise and the second random noise as complex channel gains in a frequency domain.

In this embodiment, the fading coefficient is set to pass a frequency component below the maximum Doppler frequency caused by the movement of the radio receiver.

In this embodiment, the fading coefficient is set in consideration of an interpolation factor set in the interpolating step.

In this embodiment, the sampling frequency of the fading filter is set to be inversely proportional to the magnitude of the interpolation factor.

In this embodiment, in the inverse discrete Fourier transform, the point size of the inverse discrete Fourier operation is determined according to the interpolation factor size.

Simulation system of the communication system of the present invention for achieving the above object comprises a channel gain generator for generating a complex channel gain in the frequency domain; An inverse discrete Fourier transformer for converting and storing the complex channel gain of the frequency domain into the complex channel gain of the time domain; And an interpolation block for interpolating and providing a complex channel gain of a time domain stored in the inverse discrete Fourier transform unit during a simulation operation.

In this embodiment, the channel gain generator comprises: a random noise generator for generating random noises independent of each other in a frequency domain; A plurality of fading filters for providing a fading effect for each of the independent random noises to produce the complex channel gain in the frequency domain; And a fading coefficient generator for generating fading coefficients for determining frequency characteristics of the fading filter.

In this embodiment, the random noise generator includes a first random noise generator and a second random noise generator that independently generate white Gaussian noise.

In this embodiment, the plurality of fading filters filter independent white Gaussian noises output from each of the first random noise generator and the second random noise generator according to the fading coefficient.

In this embodiment, a multiplexer for orthogonalizing the independent channel gains output from the plurality of fading filters and providing the complex channel gain is provided.

In this embodiment, the fading coefficient is set to pass a frequency component below the maximum Doppler frequency caused by the movement of the radio receiver.

In this embodiment, the sampling frequency of the fading filter is set to be inversely proportional to the magnitude of the interpolation factor.

In this embodiment, the point size of the inverse discrete Fourier transform is varied according to the interpolation factor size.

In this embodiment, the inverse discrete Fourier transformer further includes a memory for storing the transformed complex channel gain of the time domain.

According to the fading channel modeling apparatus and the channel modeling method according to the present invention through the above configuration, it is possible to calculate the channel gain before the simulation and to provide the interpolated channel gain when the simulation starts. Therefore, it is possible to greatly reduce the amount of calculation required for the simulation of the communication system, and consequently to improve the simulation speed.

It is to be understood that both the foregoing general description and the following detailed description are exemplary, and that additional explanations of the claimed invention are provided. Reference numerals are shown in detail in preferred embodiments of the invention, examples of which are shown in the reference figures. In any case, like reference numerals are used in the description and the drawings to refer to the same or like parts. DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings so that those skilled in the art may easily implement the technical idea of the present invention.

2 is a block diagram briefly showing a fading channel modeling apparatus 100 according to the present invention. Referring to FIG. 2, the fading channel modeling apparatus 100 according to the present invention performs and stores an acquisition operation of a complex channel gain h (t) in a simulation preparation step of a communication system. In addition, the complex channel gain h (t) generated in the simulation execution step may be interpolated and provided. In more detail,

The first random noise generator 110 and the second random noise generator 120 independently generate white Gaussian noise. White Gaussian Noise refers to noise having a uniform power spectral density in all frequency bands. The white Gaussian noise may be expressed as a stationary random process, and has a time-independent average power and auto correlation function. Of course, for modeling the channel, the first and second random noise generators 110 and 120 of the present invention generate data having a power spectral density and a stochastic characteristic in the frequency domain. Will be. The impulse response of most wireless channels is modeled in consideration of the white Gaussian noise described above.

The fading coefficient generator 130 generates a fading coefficient that determines the filtering characteristics of the fading filters 140 and 150. In particular, the fading coefficient generator 130 according to the present invention is caused by the influence of a multipath or a mobile receiver on the random Gaussian noise generated by the first random noise generator 110 and the second random noise generator 120. A fading coefficient C (f) is generated to provide a fading effect. The fading coefficient generator 130 generates fading coefficients corresponding to various channel characteristics in order to provide various fading effects. The fading coefficient generator 130 should consider the frequency for interpolation that is performed later along with the fading effect on the random noise generated in the frequency domain. That is, the fading coefficient generator 130 generates a fading coefficient so as to have a spectral characteristic to which a fading effect and an interpolation factor are applied.

Fading filters 140 and 150 filter out random Gaussian noise according to the fading coefficients described above. The fading filters 140 and 150 may provide a fading effect according to the movement of a multipath or a receiver to random Gaussian noise provided from each of the first random noise generator 110 and the second random noise generator 120. Impose an effect. This effect is performed through filtering by a fading coefficient provided from the fading coefficient generator 130. Random noises to which the fading effect is imposed from the fading filters 140 and 150 are respectively complexed to have an orthogonal relationship. The random noise generated from the fading filter 150 is multiplied by a complex number (-j) by the mixer 160 to have a 90 ° phase with respect to the fading filter 140. In the time domain, for this manipulation, random noise will be multiplied by a complex number (j). Then, each random noise having an orthogonal relationship is added by the adder 170 and provided to the inverse discrete Fourier transform unit 180. Here, the sampling frequency fs of the fading filters 140 and 150 should be set in consideration of an interpolation factor of the interpolation block 190 which will be described later. A multiplier 160 and an adder 170 inserted for multiplexing with the first random noise generator 110 and the second random noise generator 120, the fading coefficient generator 130, and the fading filters 140 and 150. ) May be collectively referred to as a channel gain generator for generating a complex channel gain in the frequency domain. The multiplier 160 and the adder 170 for multiplexing signals output from the respective fading filters 140 and 150 and providing the inverse discrete Fourier transform unit 180 with a complex channel gain may be briefly represented as a multiplexer. .

The inverse discrete Fourier transformer 180 converts the filtered random noise in the frequency domain output from the adder 170 into a signal in the time domain. The complex channel gain generated from the inverse discrete Fourier transform unit 180 is stored in a memory (not shown). Here, the memory may be included in the inverse discrete Fourier transform unit 180. Alternatively, the memory (not shown) may be configured as a separate device for storing complex channel gains. In general, a Discrete Fourier Transform and an Inverse Fast Fourier Transform (IFFT) algorithm are used for fast computation by eliminating iterations. However, the algorithm of the inverse discrete Fourier transform unit 180 of the present invention is not limited thereto. In the case of providing a complex channel gain simultaneously with the simulation, a complex channel gain h (t) should be provided for each of the signal samples to be simulated. However, in the present invention, all the complex channel gains are calculated and stored before the simulation, and the complex channel gains that have already been generated are extended through the interpolation block while the simulation is in progress. Accordingly, the inverse discrete Fourier transform unit 180 of the present invention may not be a high performance algorithm or arithmetic means for the inverse Discrete Fourier transform calculation.

Interpolation block 190 provides the complex channel gain of the time domain stored in memory (not shown) during the simulation operation. Interpolation refers to an operation of increasing the sampling frequency by N times by inserting N prediction signals between the signals in the sampled time domain. Interpolation block 190 of the present invention is activated in synchronization with the progress of the simulation operation of the communication system. The interpolation block 190 performs the above-described interpolation operation on the already calculated complex channel gain h (t) and provides the channel gain for the sample of the signal to be simulated. If the complex channel gain h (t) stored in the memory 180 is 10, it is converted to a complex channel gain having a sampling frequency of 10N times through interpolation. And the interpolated complex channel gain is applied to each of the samples of the transmission signal of the simulated communication system, thereby providing the characteristics of the fading channel. Here, interpolation block 190 is performed for each of real and imaginary channel gains, and each of the interpolated channel gains is added to provide a complex channel gain h (t).

Here, generation of the complex channel gain in the time domain by the first and second random noise generators 110 and 120, the fading filters 140 and 150, and the inverse discrete Fourier transform unit 180 described above is necessary for simulation. It should be done in all cases. That is, since the generated complex channel gain should effectively represent the statistical characteristics for the fading channel, the memory requirement increases with the enormous capacity of the complex channel gain generated before the simulation. This problem has been solved through the provision of an interpolation block 190 in the present invention. This is because an interpolation technique can generate and store a small amount of data in a memory (not shown) than the required amount of complex channel gain.

In addition, the method of generating the complex channel gain through the generation of random noise in the frequency domain, the insertion of the fading effect by the fading filters 140 and 150, and the inverse discrete Fourier transform according to the above-described scheme is linear. Based on the property. That is, since the statistical characteristics of the signal generated by the fading filtering and the inverse fourier transform do not change even after being transformed from the frequency domain to the time domain.

Through the above configurations, the fading channel modeling apparatus 100 according to the present invention calculates a complex channel gain before the simulation operation is performed, and the calculated complex channel gain is stored in a memory (not shown). In the simulation operation, the complex channel gain stored in the memory (not shown) is interpolated and provided for the simulation operation. This configuration can greatly reduce the amount of calculation required for the simulation operation.

FIG. 3 is a diagram briefly showing a frequency spectrum of the fading coefficient C (f) generated by the fading coefficient generator 130 of FIG. 2. Referring to FIG. 3, the fading coefficient C (f) for providing band-specific weights for filtering random noise in the frequency domain has a parameter corresponding to the fading effect by the mobile receiver. The fading coefficient C (f) includes the sampling frequencies fs of the fading filters 140 and 150 in consideration of interpolation. In more detail,

By the fading coefficient C (f), the fading filters 140 and 150 filter random Gaussian noise input with different weights according to the magnitude of the frequency. For example, the fading filters 140 and 150 function as a low pass filter that passes only frequencies of magnitude corresponding to the Doppler frequency in order to provide fading effects caused by the Doppler effect. Has The setting of the fading coefficient C (f) is not limited to only considering the fading effect. The setting of the fading coefficient C (f) should take into account the Inverse Discrete Fourier Transform or Interpolation operation.

First, look at the conditions of the fading coefficient C (f) to provide a fading effect. In general, fading effects are caused by the movement of a multipath or wireless receiver. A typical effect of such fading is the Doppler effect. The frequency change on the transmission signal through the Doppler effect is expressed by Equation 2 below.

f _{D , max} : maximum Doppler shift frequency

f _C Carrier frequency

c: luminous flux

v: speed of mobile receiver

As shown in Equation 1, the fading effect is determined by the carrier frequency and the moving speed of the receiver. In general, in a system using a band of several hundred MHz to several GHz as the carrier frequency, the moving speed of the receiver is extremely low compared to the luminous flux c. And the moving speed of the receiver is negligibly low compared to the speed of light. Thus, the fading coefficient C (f) will have a very low maximum Doppler shift frequency f _{D , max} . In other words, in order to pass only the random noise in which fading is performed by the Doppler frequency _, a low pass filter that passes only the random noise corresponding to the maximum Doppler moving frequencies f _{D and max} is performed. The fading coefficient C (f) has the maximum Doppler shift frequency as the cutoff frequency.

Referring back to FIG. 3, the spectrum of the fading coefficient C (f) should be considered in advance for the sampling frequency of the fading filters 140, 150 for interpolation in addition to the manipulation for providing a fading effect on random noise. Accordingly, sampling frequencies of the fading filters 140 and 150 are set in consideration of the point of the inverse discrete Fourier transform and the interpolation factor of the interpolation operation performed by the interpolation block 190. This setting is shown in Equation 3 below.

fs _: sampling frequency of the fading filter

Fs: sampling frequency of the original signal

L: Interpolation factor (Output rate / Input rate)

As a result, the sampling frequency fs of the fading filters 140 and 150 and the point setting of the inverse discrete Fourier transform unit 180 change according to the magnitude of the interpolation factor.

According to the above conditions, the fading coefficient C (f) is applied to the random noise by the fading filters 140 and 150. In the frequency domain, filtering is generally expressed as the product of the Fourier transformed signals of the input signal and the impulse response. Therefore, the filtering operation is performed by multiplying the random noise by the aforementioned fading coefficient C (f). The fading filtering considering the fading effect and interpolation is completed by multiplying the random noise by the fading coefficient C (f).

An example of the sampling frequency (fs) setting of the fading filters 140 and 150 and the point setting of the inverse discrete Fourier transform unit 180 according to the magnitude of the interpolation factor will be briefly described as follows. Assume that the amount of data of the complex channel gain h (t) required for the simulation of the communication system is 1 million samples. In this case, in order to provide a complex channel gain simultaneously with simulation without interpolation, random noise generation and fading filtering should be performed for each one million samples of each transmission signal. However, according to the present invention, a complex channel gain of a size corresponding to (1 million × 1 / L, L is an interpolation factor) is generated before performing a simulation operation by generating random noise, fading filtering, and inverse discrete Fourier transform. Save it to memory. When the simulation is started, a complex channel gain having a size of 1 million × 1 / L, which is already calculated, may be provided by performing an interpolation operation with the size of the interpolation factor L. Thus, once the simulation of the communication system is started, it is not necessary to perform computational calculations such as random noise generation, fading filtering and inverse discrete Fourier operations.

4 is a flowchart briefly illustrating a channel modeling method in the above-described manner. Referring to FIG. 4, a complex channel gain of a time domain is generated by performing an inverse discrete Fourier transform by multiplying a random noise generated in a frequency domain by a fading coefficient considering a fading effect and an interpolation factor. However, such complex channel gains may be stored in memory (not shown) and continuously provided by interpolation in the simulation of the communication system. Hereinafter, this operation will be described with reference to FIGS. 2 and 3.

When the separate channel modeling operation according to the present invention is started prior to the simulation of the wireless communication system, fading coefficients and random Gaussian noise applied to the fading filter are generated. Here, fading coefficient C (f) and random noise are generated in the frequency domain. Thus, random noise will be generated to have a constant power spectral density over the entire frequency range. The random noise may be generated through two independent random noise generators 110 and 120, respectively. This is because a complex channel gain including a real part and an imaginary part can be derived by combining channel gains generated through independent random noises. The fading coefficient C (f) will be set to have the frequency characteristic shown in FIG. That is, the random noise is generated in consideration of the fading effect and the interpolation factor of the interpolation block performed at the time of performing the simulation operation (S10).

When generation of the random noise and fading coefficient C (f) is completed, the filtering is performed by the fading filters 140 and 150. The sampling frequency fs of the fading filters 140 and 150 is determined in consideration of the interpolation factor of the interpolation block 190. Interpolation factor means the length of the output data with respect to the length of the input data of the interpolation block (S20).

Each of the random noises independently filtered by the fading filters 140, 150 are added as complex random noise to be orthogonal. The added complex random noise is converted into time domain data by the inverse discrete Fourier transformer 180. The complex random noise of the inverse discrete Fourier transform time domain is then stored in the memory as a complex channel gain (S30).

At the time of simulation of the wireless communication system, the complex channel gain stored in the memory (not shown) is extended by the interpolation block 190 to the sufficient channel gain required for the simulation. During the simulation operation of the communication system, the interpolation block 190 extends the complex channel gain h (t) to an output length corresponding to the interpolation factor L described above. Therefore, the computation of the generation of the complex channel gain h (t) and the inverse discrete Fourier transform need not be performed at the time when the simulation operation is performed. Only, the small sized complex channel gain h (t) already generated and stored is provided by increasing the sampling frequency by the length of the interpolation factor L. As a result, it is possible to reduce the amount of calculation required for the simulation, and high-speed simulation is possible when dedicating the reduced resources to the simulation operation (S40).

According to the fading channel model according to the present invention, the process from the generation of the white Gaussian noise to the inverse discrete Fourier transform is performed once before the simulation and the generated complex channel gain is stored in the memory. While the actual simulation is performed, only the interpolation operation on the complex channel gain stored in the memory is performed to generate the complex channel gain used for the simulation. Thus, it is possible to improve the speed of the simulation by reducing the calculation amount of the entire simulation system during the simulation operation.

Meanwhile, in the detailed description of the present invention, specific embodiments have been described, but various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the above-described embodiments, but should be defined by the equivalents of the claims of the present invention as well as the following claims.

1 is a block diagram schematically illustrating a channel model of a communication system;

2 is a block diagram showing an apparatus for fading channel modeling according to the present invention;

3 shows a simplified representation of the frequency characteristics of fading coefficients; And

4 is a flow chart briefly describing a channel modeling method of the present invention.

* Description of the symbols for the main parts of the drawings *

10: channel impulse response 20, 170: adder

100: fading channel modeling device

110 and 120: first and second random noise generator

130: fading coefficient generator 140, 150: fading filter

160: multiplier 180: inverse discrete Fourier transform unit (IDFT)

190: interpolation block

Claims (15)

In the channel modeling method of a communication system: Passing random noise through a fading filter to produce a complex channel gain in the frequency domain; Storing the complex channel gain of the time domain output by inverse Discrete Fourier Transform (IDFT) of the complex channel gain of the frequency domain; And Interpolating the stored complex channel gain of the time domain according to an interpolation factor. The method of claim 1, Generating the complex channel gain of the frequency domain, Generating a fading coefficient that provides first and second random noise independent of each other and frequency characteristics of the fading filter; Filtering each of the first random noise and the second random noise with the fading filter set to the fading coefficients; And Multiplexing the filtered first random noise and the second random noise and providing them as a complex channel gain in the frequency domain. The method of claim 2, And wherein the fading coefficient is set to pass a frequency component below a maximum Doppler shift frequency caused by the movement of the radio receiver. The method of claim 2, And the fading coefficient is set in consideration of the interpolation factor. The method of claim 4, wherein The sampling frequency of the fading filter is set to be inversely proportional to the magnitude of the interpolation factor. The method of claim 1, In the inverse discrete Fourier transform step, the point of the inverse discrete Fourier operation is determined according to the size of the interpolation factor (Interpolation Factor). In a fading channel modeling apparatus of a communication system: A channel gain generator for generating complex channel gains in a frequency domain by passing random noises independent from each other through a plurality of fading filters; An inverse discrete Fourier transformer for converting and storing the complex channel gain of the frequency domain into the complex channel gain of the time domain; And And an interpolation block for interpolating and outputting the complex channel gain of the time domain stored in the inverse discrete Fourier transform unit according to an interpolation factor. The method of claim 7, wherein The channel gain generator, A random noise generator for generating random noises independent of each other in a frequency domain; A plurality of fading filters that provide a fading effect for each of the independent random noises to produce a complex channel gain in the frequency domain; And And a fading coefficient generator for generating fading coefficients for determining frequency characteristics of the fading filters. The method of claim 8, The random noise generator comprises a first random noise generator and a second random noise generator for generating white Gaussian noise independently. The method of claim 9, And a plurality of fading filters filter independent white Gaussian noises output from each of the first random noise generator and the second random noise generator according to the fading coefficients. The method of claim 8, And a multiplexer for orthogonalizing the independent channel gains output from the plurality of fading filters and providing the complex channel gain as the complex channel gain. The method of claim 8, And the fading coefficient is set to pass a frequency component below a maximum Doppler shift frequency caused by the movement of the radio receiver. The method of claim 8, And a sampling frequency of the fading filter is set in inverse proportion to the magnitude of the interpolation factor. The method of claim 8, The point of the inverse discrete Fourier transform operation performed by the inverse discrete Fourier transform unit is variable according to the interpolation factor size. The method of claim 8, The inverse discrete Fourier transformer further includes a memory for storing the transformed complex channel gain of the time domain.

Images