KR20050119592A - Apparatus and method for channel estimation in a mobile communication system using an frequency hopping - orthogonal frequency division multipl access scheme - Google Patents

Abstract

The present invention divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted. In a mobile communication system using a frequency hopping-orthogonal frequency division multiple access scheme, a fast Fourier transform of a signal received in a first predetermined number of symbol time periods including a symbol time period to estimate the channel Detecting reference signals in each of the first number of symbol time periods; Generating a first channel estimate by interpolating reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals; Generating a second channel estimate by smoothing the first channel estimate in the time domain for each of the subcarriers in a preset third number of symbol time periods; The second channel estimate is smoothed in a frequency domain in units of a fourth predetermined number of subcarrier bands for each symbol time period, and is generated as a final channel estimate of the symbol time period to be estimated.

Description

FIELD AND METHOD FOR CHANNEL ESTIMATION IN A MOBILE COMMUNICATION SYSTEM USING AN FREQUENCY HOPPING-ORTHOGONAL FREQUENCY DIVISION MULTIPL ACCESS SCHEME

The present invention is a mobile communication using a frequency hopping (FH)-Orthogonal Frequency Divisional Multiple Access (OFDMA) scheme The present invention relates to a system (hereinafter referred to as an "FH-OFDMA mobile communication system"), and more particularly, to a multi-stage channel estimation apparatus and method in an FH-OFDMA mobile communication system.

In the 4th Generation (hereinafter, referred to as '4G') communication system, users of services having various quality of service (hereinafter referred to as 'QoS') having a high transmission rate are used by users. Active research is underway to provide it. In particular, in 4G communication systems, broadband wireless such as a wireless local area network (hereinafter, referred to as a 'LAN') system and a wireless metropolitan area network (hereinafter, referred to as a 'MAN') system are used. Researches are being actively conducted to support high-speed services in a form of guaranteed mobility and QoS in a broadband wireless access (BWA) communication system.

Therefore, the 4G communication system is actively studying orthogonal frequency division multiplexing (OFDM) scheme as a method useful for high-speed data transmission in wired and wireless channels. The method uses a multi-carrier to transmit data. A plurality of sub-carriers each having orthogonality to each other by converting serially input symbol strings in parallel. It is a kind of multi-carrier modulation (MCM) that modulates and transmits data.

Broadband spectrum resources are required for the 4G communication system to provide high speed, high quality wireless multimedia services. However, when the broadband spectrum resource is used, fading effects on the radio transmission path due to multipath propagation become serious and frequency selective fading also occurs within the transmission band. Therefore, the OFDM scheme, which is robust against frequency selective fading, has greater gain than the code division multiple access (CDMA) scheme for high-speed wireless multimedia services. There is a trend of being actively used in communication systems.

Meanwhile, the OFDMA scheme is a scheme in which a plurality of terminals use all sub-carriers in a multiple access scheme based on the OFDM scheme. The OFDMA method does not require a spreading sequence for spreading, but since the subchannel allocated to a specific terminal is kept fixed, the allocated subchannel is continuously fading. In this case, the transmission efficiency is reduced. Here, the subchannel refers to a channel composed of a plurality of subcarriers.

Accordingly, a frequency diversity gain may be obtained by dynamically changing a set of subcarriers allocated to a specific terminal, that is, a subchannel according to a fading characteristic of a wireless transmission path, and obtaining the obtained frequency diversity gain. This can increase the transmission efficiency. In this way, a method of dynamically changing a subchannel allocated to a specific terminal is called a 'dynamic resource allocation' method, and a representative method of the dynamic resource allocation method is the FH method.

As a result, the FH-OFDMA scheme is a multiple access scheme having robustness to frequency selective fading through the characteristics of the OFDMA scheme and the FH scheme. In the FH-OFDMA scheme, a subchannel frequency allocated to each terminal is leaped using a predetermined FH code to simultaneously acquire effects due to the FH scheme as well as the OFDMA scheme. In this case, a Latin square code or the like is used as the FH code. The Latin square code has an advantage of reducing cell division as well as inter-cell interference (ICI) in a multi-cell environment.

On the other hand, unlike a wired channel environment, a wireless channel environment exists in a mobile communication system due to multipath interference, shadowing, propagation attenuation, time-varying noise, and delay spread. Various factors, such as inter-symbol interference (ISI) and frequency selective fading, cause unavoidable errors and loss of information data. The loss of the information data causes severe distortion in the actual transmission signal, thereby acting as a factor that degrades the overall performance of the mobile communication system.

Accordingly, in the OFDMA communication system, not only the multiple orthogonal subcarriers are used to overcome the inter-symbol interference and frequency selective fading, but also the last constant samples of the OFDM symbols in the time domain are copied to the effective OFDM. The OFDM symbols are transmitted by inserting a guard interval of a 'cyclic prefix' method of inserting a symbol or a 'cyclic postfix' method of copying first predetermined samples of an OFDM symbol in a time domain and inserting the same in a valid OFDM symbol.

Accordingly, the frequency selective fading is uniformly distributed for each subcarrier, so that a receiver can receive a signal even if only a single tap equalizer is provided for each subcarrier. In addition, the time change of channel state information (CSI: CSI) according to the Doppler phenomenon generates inter-carrier interference (ICI) to generate the OFDMA communication. It results in performance degradation of the system.

As described above, accurate CSI detection is essential in a communication system using the OFDM scheme, and pilot is used for channel estimation. On the other hand, when the Latin Square code is used as the FH code, when a periodic FH code pilot is used, a constant pilot interval can be maintained in the frequency domain but constant in the time domain. Since the pilot interval cannot be maintained, the density of the FH code pilot must be increased for accurate channel estimation. However, when the density of the FH code pilot is increased in the FH-OFDMA communication system, data transmission efficiency is deteriorated.

Accordingly, there is a need for an effective channel estimation method in a time-varying frequency selective fading channel while using a minimum pilot in an FH-OFDMA communication system using a Latin square code as the FH code.

Accordingly, an object of the present invention is to provide an apparatus and method for multi-stage channel estimation in an FH-OFDMA mobile communication system.

Another object of the present invention is to provide an apparatus and method for estimating a channel using interpolation in an FH-OFDMA mobile communication system.

It is still another object of the present invention to provide an apparatus and method for estimating a channel using a curve splicing method in an FH-OFDMA mobile communication system.

Another object of the present invention is to provide an apparatus and method for estimating a channel using an interpolation scheme and a curve joint scheme in an FH-OFDMA mobile communication system.

The method of the present invention for achieving the above objects; A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimating method of a mobile communication system using a hopping-orthogonal frequency division multiple access scheme, the fast Fourier performing a signal received in a first predetermined number of symbol periods including a symbol time period to estimate the channel. Converting and detecting reference signals in each of the first number of symbol time periods, and interpolating the reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals Generating a first channel estimate, generating the first channel estimate, Generating a second channel estimate by smoothing the first channel estimate in the time domain for each of the subcarriers in a set number of symbol time periods, generating the second channel estimate, and then generating the second channel estimate for each symbol time period. And a second channel estimate is smoothed in a frequency domain in units of fourth preset subcarrier bands to generate a final channel estimate of the symbol time period to be estimated.

Another method of the present invention for achieving the above objects is; A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimating method of a mobile communication system using a hopping-orthogonal frequency division multiple access scheme, comprising: fast Fourier transforming a received signal to detect reference signals, and using the reference signal borrowing scheme and frequency domain interpolation scheme. Generating a first channel estimate, generating a second channel estimate using a smoothing method in a time domain after generating the first channel estimate, and generating the second channel estimate After the second channel estimate in the frequency domain using a smoothing method It characterized in that it comprises the step of generating politics.

The apparatus of the present invention for achieving the above objects; A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimating apparatus of a mobile communication system using a hopping-orthogonal frequency division multiple access scheme, the fast Fourier receiving signal received in a first predetermined number of symbol periods including a symbol time period to estimate the channel. A fast Fourier transformer for converting and detecting reference signals in each of the first number of symbol time periods, and reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals Generate a first channel estimate by interpolation, and set a third predetermined number of symbols After generating the second channel estimate by smoothing the first channel estimate in the time domain for each of the subcarriers in intervals, the second channel estimate is preset in units of a fourth number of subcarrier bands for each symbol time period. And a channel estimator for smoothing in the frequency domain and estimating the final channel estimate of the symbol time period to be estimated.

Another apparatus of the present invention for achieving the above objects; A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimating apparatus of a mobile communication system using a hopping-orthogonal frequency division multiple access scheme, comprising: a fast Fourier transformer for detecting a reference signals by fast Fourier transforming a received signal, and using the reference signal borrowing scheme and frequency domain interpolation A first channel estimate is generated using the method, a second channel estimate is generated using the smoothing method in the time domain, and then the second channel estimate is finalized using the smoothing method in the frequency domain. And a channel estimator for estimating the channel estimate.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be noted that in the following description, only parts necessary for understanding the operation according to the present invention will be described, and descriptions of other parts will be omitted so as not to distract from the gist of the present invention.

The present invention is a mobile communication using a frequency hopping (FH)-Orthogonal Frequency Divisional Multiple Access (OFDMA) scheme An apparatus and method for estimating a channel using an interpolation method and a curve fitting method in a system (hereinafter referred to as an 'FH-OFDMA mobile communication system') are proposed. In particular, the present invention uses a Latin-square code as an FH code in an FH-OFDMA mobile communication system, and uses only a minimum pilot, but efficiently performs three-step channel estimation in a frequency selective fading channel. An apparatus and method are proposed. In addition, the present invention considers channel estimation in an FH-OFDMA mobile communication system having a multi-cell structure.

1 is a diagram illustrating an internal structure of a transmitter of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention.

Referring to FIG. 1, first, the transmitter, that is, the base station, includes an encoder 111, an interleaver 113, a symbol mapper 115, and a serial / parallel converter. to parallel converter (417), pilot symbol inserter (419), frequency hopping (FH unit) 419, and Inverse Fast Fourier Transform (IFFT) 123, a parallel to serial converter 125, a guard interval inserter 127, and a digital to analog converter 129 and a radio frequency (RF) processor 131.

First, when data is input, the encoder 111 encodes the input data using a predetermined coding scheme and outputs the encoded data to the interleaver 113. Here, the coding scheme may be a turbo coding scheme or a convolutional coding scheme having a predetermined coding rate. The interleaver 113 inputs the signal output from the encoder 411, interleaves the interleaving method in a predetermined interleaving manner, and outputs the signal to the symbol mapper 115. In this case, the interleaving scheme may be a random interleaving scheme.

The symbol mapper 115 modulates the coded bits output from the interleaver 113 using a predetermined modulation scheme to generate modulation symbols, and then converts the coded bits to the serial / parallel converter 117. Output The modulation scheme may be a quadrature phase shift keying (QPSK) scheme or a quadrature amplitude modulation (16QAM) scheme.

The serial / parallel converter 117 inputs the serial modulation symbols output from the symbol mapper 115 and converts them in parallel and outputs them to the pilot symbol inserter 119. The pilot symbol inserter 119 inserts pilot symbols into the parallel-converted modulated symbols output from the serial / parallel converter 117 and then outputs the pilot symbols to the frequency hop unit 121. The frequency hopping unit 121 inputs the signal output from the pilot symbol inserter 119 to frequency hopping with a preset FH code, that is, a Latin square code, and outputs the frequency to the IFFT unit 123. Here, the signal output from the frequency hopping unit 121, that is, one OFDM symbol ( Let's assume.

The IFFT device 123 is a signal in the frequency domain output from the frequency hopping device 121, that is, N-point IFFT is input to generate a signal in the time domain, and then output to the parallel / serial converter 125. Here, it is assumed that one OFDM symbol in the time domain output from the IFFT unit 123 is x (n), and x (n) may be represented by Equation 1 below.

In Equation 1, N represents the total number of subcarriers used in the FH-OFDMA mobile communication system.

The parallel / serial converter 125 inputs the signal output from the IFFT device 123, converts the signal in series, and outputs the serial signal to the guard interval inserter 127. The guard interval inserter 127 inputs a signal output from the parallel / serial converter 125 to insert a guard interval signal and then outputs the guard interval signal to the digital / analog converter 129. Here, the guard interval is inserted to remove inter-symbol interference (ISI) between the OFDM symbol transmitted at the previous OFDM symbol time and the current OFDM symbol transmitted at the current OFDM symbol time when transmitting the OFDM symbol. In addition, the guard interval is a 'cyclic prefix' scheme in which the last constant samples of the OFDM symbols in the time domain are copied and inserted into a valid OFDM symbol, or the first constant samples of the OFDM symbols in the time domain. It can be used as a 'cyclic postfix' method of copying and inserting into a valid OFDM symbol. The length of the guard interval is set longer than the maximum delay spread length allowable in the FH-OFDMA mobile communication system.

The digital-to-analog converter 129 inputs the signal output from the guard interval inserter 127 and performs analog conversion, and then outputs the signal to the RF processor 131. The RF processor 131 may include components such as a filter and a front end unit, and may transmit a signal output from the digital-to-analog converter 129 on an actual air. After the RF process, the transmission is performed on the air through a Tx antenna.

In FIG. 1, a transmitter structure of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention has been described. Next, an FH for performing a function in an embodiment of the present invention will be described with reference to FIG. The receiver structure of the -OFDMA mobile communication system will be described.

2 is a diagram illustrating an internal structure of a receiver of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention.

Referring to FIG. 2, first, a receiver of a FH-OFDMA mobile communication system, that is, a terminal, includes an RF processor 211, an analog / digital converter 213, and a guard interval remover. 215, a serial / parallel converter 217, a Fast Fourier Transform (FFT) 219, an equalizer 221, A pilot symbol extractor 223, a channel estimator 225, a frequency de-hopping unit 227, a parallel / serial converter 229, and a symbol D. It consists of a symbol demapper 231, a deinterleaver 233, and a decoder 235.

First, a signal transmitted from a transmitter of an IFH-OFDMA mobile communication system as described with reference to FIG. 1 undergoes a time-varying frequency selective fading channel and a noise component is added to the reception antenna of the receiver of the FH-OFDMA mobile communication system (Rx). received via an antenna. Herein, it is assumed that the signal received through the reception antenna is y (n), and the reception signal y (n) may be represented by Equation 2 below.

In Equation 2, M represents the number of multipaths, Denotes the channel impulse response (CIR) of the l-th path, and w (n) denotes a noise component, that is, white additive Gaussian noise (AWGN). Indicates.

The signal received through the receive antenna is input to the RF processor 211, and the RF processor 211 down converts the signal received through the receive antenna to an intermediate frequency (IF) band. And then output to the analog-to-digital converter 13. The analog / digital converter 213 digitally converts the analog signal output from the RF processor 211 and outputs the digital signal to the guard interval remover 215.

The guard interval remover 215 removes the guard interval signal by inputting the signal output from the analog / digital converter 213 and outputs the signal to the serial / parallel converter 217. The serial / parallel converter 217 inputs a serial signal output from the guard interval remover 215 to perform parallel conversion and outputs the serial signal to the FFT unit 219. The FFT 219 outputs the signal output from the serial / parallel converter 217 to the equalizer 221 and the pilot symbol extractor 223 after performing an N-point FFT. The equalizer 221 inputs the signal output from the FFT 219 to channel equalize and outputs the signal to the frequency reverse hop 227. The frequency reverse hop unit 227 frequency-hops using the same FH code as the FH code applied by the transmitter of the FH-OFCDMA mobile communication system and outputs the same to the parallel / serial converter 229. The parallel / serial converter 229 inputs the signal output from the frequency reverse hop 227, converts the signal in series, and outputs the serial signal to the symbol demapper 231.

On the other hand, the signal output from the FFT 219 is input to the pilot symbol extractor 223, the pilot symbol extractor 223 detects pilot symbols from the signal output from the FFT 219, The detected pilot symbols are output to the channel estimator 225. The channel estimator 225 performs channel estimation using the pilot symbols output from the pilot symbol extractor 223 and outputs the channel estimation result to the equalizer 221.

The symbol demapper 231 demodulates a signal output from the parallel / serial converter 229 in a demodulation scheme corresponding to a modulation scheme applied by a transmitter of the FH-OFDMA mobile communication system, and then deinterleaver 233. Will output The deinterleaver 233 deinterleaves the signal output from the symbol demapper 231 in a deinterleaving scheme corresponding to an interleaving scheme applied by a transmitter of the FH-OFDMA mobile communication system, and then decodes the decoder (deinterleaving). 235).

The decoder 235 decodes and outputs a signal output from the deinterleaver 233 using a decoding method corresponding to a coding method applied by a transmitter of a transmitter of the FH-OFDMA mobile communication system (output data).

2 illustrates a receiver structure of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention. Next, an FH-OFDMA mobile communication system according to an embodiment of the present invention will be described with reference to FIG. In the following description, a structure in which pilot signals are transmitted and a structure in which pilot signals are borrowed will be described.

3 is a diagram schematically illustrating a structure in which pilot signals are transmitted and a pilot vehicle structure in an FH-OFDMA mobile communication system according to an embodiment of the present invention.

As described above, since the FH-OFDMA mobile communication system has a multi-cell structure as described above, FH capable of cell division in order to minimize inter-cell interference (ICI) between the multiple cells is possible. The code set must be used, and the influence of the selection of the FH code set on the system performance improvement of the FH-OFDMA mobile communication system is very large. Accordingly, in the present invention, the Latin square code is used as the FH code, and the Latin square code matrix is composed of p code sets of length p, and each of the codes constitutes a number from 0 to p-1. It has an element. In this case, the rows and columns of the Latin square code matrix are configured to have the numbers 0 to p-1 only once.

Then, the following description about the Latin square code generation process is as follows.

First, the first row of the Latin square code matrix is sequentially arranged with numbers from 0 to p-1, and the second row is generated by cyclic shifting by a in the first row. The third row is generated by cyclically shifting the second row by a, and the remaining rows are generated in the same manner as the first to third rows. In addition, p-1 different Latin square code sets may be generated while varying a corresponding to the slope from 1 to p-1. When the code length is a prime number, the Latin square code generation process may be expressed by Equation 3 below.

In Equation 3, {a} represents the Latin square code matrix, i represents a row of the Latin square code matrix, j represents a column of the Latin square code matrix, i, j = 0, 1 , ..., p-1. The FH code corresponds to each column of the Latin square code matrix. Therefore, each of the columns of the Latin square code matrix is allocated to each of the terminals, and frequency hopping is performed at a frequency corresponding to the number constituting the columns of the Latin square code matrix every cycle. On the other hand, the slope of the FH code set is also used to distinguish the cells, of course.

3 illustrates a structure in which pilot signals are transmitted when the variable representing the FH code set, that is, the slope a is 6, the FH-code length p is 1637, and the pilot interval L is 10. FIG. As shown in FIG. 3, when the Latin square code is used as the FH code, the position where the pilot is transmitted is irregular. That is, when the Latin square code is selected as the FH code, pilot borrowing is possible by accurately setting the positions of the subcarriers to which pilots are transmitted in the vicinity of the OFDM symbol to be estimated, so that accurate estimation is possible. Do. As described in the prior art, when the Latin square code is used as the FH code, the pilot density must be increased for accurate channel estimation. In the present invention, the accurate channel estimation can be performed while maintaining the pilot density.

Next, a multi-stage channel estimation process according to an embodiment of the present invention will be described with reference to FIG. 4.

4 is a flowchart illustrating a multi-step channel estimation process according to an embodiment of the present invention.

Before describing FIG. 4, the multi-stage channel estimation process proposed by the present invention includes a first stage channel estimation using a pilot borrowing scheme and a frequency domain interpolation scheme, and a second stage using a time domain smoothing scheme. Channel estimation and a third stage channel estimation using a frequency domain smoothing scheme.

Referring to FIG. 4, in step 411, a receiver of the FH-OFDMA mobile communication system extracts pilot symbols from a received signal and then proceeds to step 413. In step 413, the receiver performs a first channel estimation operation, that is, performs a channel estimation operation using the pilot borrowing method and the frequency domain interpolation method, and then proceeds to step 415. The first step channel estimation process will now be described.

First, in order to increase the data rate in the FH-OFDMA mobile communication system, the pilot interval, that is, the pilot density, must be kept as low as possible. For the channel estimation in the frequency domain, if the frequency selective fading of the channel is relatively severe, if the pilot density is too low, the accuracy of the channel estimation is degraded and the performance of the FH-OFDMA mobile communication system is degraded. Accordingly, in the present invention, the channel state information (CSI) in the time domain is referred to as shown in FIG. 5 by using a point that the change is relatively smaller than the CSI change in the frequency domain. As described above, the channel estimates are obtained by using pilot signals distributed in OFDM symbols adjacent to the current OFDM symbol, and then the channel is estimated using an interpolation scheme, for example, an n-th order Lagrange interpolation scheme. If the channel estimate detected through the first channel estimation process is represented, it may be expressed as Equation 4 below.

In Equation 4, Denotes a channel estimate obtained through the first step of channel estimation of the k-th subcarrier, Represents an LS (Least Square) channel estimate of the borrowed pilot signals, i.e. Denotes a value obtained by simply dividing a received signal by a pilot symbol. Also, k _b represents the number of subcarriers of pilot symbols including the borrowed pilot signals.

Next, the first step channel estimation process will be described with reference to FIG. 5.

FIG. 5 is a diagram schematically illustrating an operation in the first stage channel estimation process of FIG. 4.

Referring to FIG. 5, channel estimates are obtained by borrowing pilot signals distributed over three OFDM symbols and five OFDM symbols in a time domain. Then, the interpolation operation is performed in the frequency domain by using the secondary Lagrangian interpolation method. Here, the second Lagrangian interpolation method uses three pilot signals in the frequency domain and averages data signals that are not pilot signals when duplicated.

In addition, in step 415 of FIG. 4, the receiver performs a second channel estimation operation, that is, performs a channel estimation operation using a time domain smoothing method, and then proceeds to step 417. Next, the second step channel estimation process will be described.

First, channel estimates obtained in the first step channel estimation process include the noise component, frequency selective fading, inter-carrier interference (ICI) according to CSI time variation, and the first step channel. An error component of the pilot borrowing method and the interpolation method is included in the estimation process. Accordingly, an error included in channel estimates obtained in the first step channel estimation process may be reduced by using a time domain smoothing method using a time domain curve bonding method for each subcarrier as shown in Equation 5 below.

In Equation 5, i represents an OFDM symbol index, Denotes the η-th polynomial coefficient of the k-th subcarrier, η = 0, 1, ..., n, wherein Is determined as a value capable of minimizing the mean square error (MSE), which can be expressed as in Equation 6 below.

In Equation 6 Denotes the number of OFDM symbols corresponding to the curved junction interval, Has a value greater than or equal to η +1 ( ), Denotes a channel estimate of the i-th OFDM symbol of the k-th subcarrier obtained in the first step of channel estimation.

Next, the second step channel estimation process will be described with reference to FIG. 6.

FIG. 6 is a diagram schematically illustrating an operation in the second stage channel estimation process of FIG. 4.

Referring to FIG. 6, a plurality of OFDM symbols in a time domain, for example, in a 50 [km / h] environment, channel carriers are estimated using a second LS curve joint scheme on subcarriers during 47 OFDM symbol intervals. For example, in a 150 [km / h] environment, subcarriers for 15 OFDM symbol intervals are acquired using a second LS curve joint scheme to obtain a channel estimate. Then, the channel estimate obtained in the first step channel estimation process is re-estimated using the polynomial as described in Equations 5 and 6 to obtain a channel estimate.

In addition, in step 417 of FIG. 4, the receiver terminates after performing the third channel estimation operation, that is, performing the channel estimation operation using the frequency domain smoothing method. Next, the third step channel estimation process will be described.

First, the channel estimates acquired in the second channel estimation process may reduce the estimation error by applying a curve joint method again in the frequency domain. Here, the n-th order curved joint polynomial of the frequency domain may be expressed by Equation 7 below.

Next, the third step channel estimation process will be described with reference to FIG. 7.

FIG. 7 is a diagram schematically illustrating an operation in the third stage channel estimation process of FIG. 4.

Referring to FIG. 7, channel estimates are obtained for each OFDM symbol using the second order LS curve joint scheme. That is, the final channel estimate is obtained by re-estimating the channel estimate obtained in the second step channel estimation process using the polynomial as described in Equation 7. In the third step of channel estimation, a second LS curve joint scheme is applied in units of five subcarriers, and the overlapping signals are averaged.

Next, the performance characteristics of the three-stage channel estimation process proposed by the present invention will be described with reference to FIGS. 8 to 10.

Prior to describing the performance characteristics of the three-stage channel estimation process, the performance characteristics use quadratic curve-junction polynomials in both the time domain and the frequency domain, and use 47 OFDM symbols in a 50 [km / h] environment. In the case of 150 [km / h], the performance characteristics obtained when 15 OFDM symbols are used. Here, the reason for using the quadratic curve joint polynomial is to consider the complexity of the operation and the delay of the OFDM symbols according to the application of the algorithm, and the number of optimal OFDM symbols in the respective speed environments is the second stage channel estimation. It is determined by considering the time domain polynomial approximation in the process.

The performance characteristic is that the center frequency is 5.8 [GHz], uses 100 [MHz] bandwidth, uses 2048 subcarriers, the guard interval length is 512 samples, and the modulation scheme applies the QPSK scheme. The pilot spacing L is 10 (10%), 12 (8.3%), and 15 (6.3%), and the error variance of the LS channel estimates by the received pilot signal is 0.1 (10 [dB]), 0.032 (15 [ dB]), and the borrowing range B of the pilot symbol is a performance characteristic when 3, 5, and 7 are applied. The performance characteristic is a coding rate 1/2 and a length of a coded block. The 1200-bit zigzag code is used as the channelization code, and the number of repeated decoding is a performance characteristic when 8 is applied. In addition, the performance characteristics assume 50 independent multipaths in which the average power of each of the multipaths decreases exponentially, the root mean square (RMS) delay spread is 1.5 [μs], and the maximum delay spread is 5 [μs]. Considering the Jakes model, each of which consists of the sum of 50 independent propagation lines, the maximum Doppler frequency is 268.5 [Hz] and 150 [km / h] in a 50 [km / h] environment. Performance characteristics when considering 805.6 [Hz] in the environment.

8 illustrates a bit error rate according to a maximum Doppler frequency of 268.5 [Hz] in a 50 [km / h] environment and a pilot signal interval and the number of OFDM symbols to be borrowed when using accurate CSI at a pilot signal position (BER: Bit) Error Rate, hereinafter referred to as BER). A graph showing performance.

Referring to FIG. 8, when the pilot signal interval is 12 and the number of borrowed OFDM symbols is seven, the performance of the three-stage channel estimation scheme proposed by the present invention is almost compared with the theoretical perfect channel estimation performance. It can be seen that the same. In the radio channel environment as shown in FIG. 8, since the time variation of the CSI is relatively small, the performance is improved as the number of OFDM symbols borrowed increases. In FIG. 8, when the pilot signal interval is 15, when the number of borrowed OFDM symbols is 5, when using 15 OFDM symbols in the second channel estimation process, about 0.2 [dB] is used. Only the difference in degree is shown.

8 illustrates the BER performance according to the maximum Doppler frequency of 268.5 [Hz] in a 50 [km / h] environment and the pilot signal spacing and the number of OFDM symbols borrowed when the correct CSI is used at the pilot signal position. Next, referring to FIG. 9, according to the maximum Doppler frequency of 805.6 [Hz] in the 150 [km / h] environment and the pilot signal interval and the number of borrowed OFDM symbols when there is no error in the channel estimate at the pilot signal position. BER performance will be described.

9 illustrates BER performance according to a maximum Doppler frequency of 805.6 [Hz] in a 150 [km / h] environment and a pilot signal interval and the number of borrowed OFDM symbols when there is no error in the channel estimate at the pilot signal position. It is a graph shown.

9, it can be seen that the smaller the pilot signal interval, the BER performance is improved, and the highest performance is obtained when the number of borrowed OFDM symbols is five. This is because the higher the Doppler frequency, the larger the time variation of the CSI, thereby increasing the probability of error due to the pilot borrowing method. In addition, it can be seen that the performance when the pilot signal interval is 12 and the number of borrowed OFDM symbols is 5 shows excellent performance within about 0.2 [dB] compared to the ideal channel estimation performance. In addition, when the pilot signal interval is increased to 15 and the number of borrowed OFDM symbols is used as 5, the BER performance is about 1 [dB] deteriorated compared to the ideal channel estimation performance. Can be.

9 illustrates a maximum Doppler frequency of 805.6 [Hz] in a 150 [km / h] environment, and BER performance according to the pilot signal interval and the number of borrowed OFDM symbols when there is no error in the channel estimate at the pilot signal position. Next, BER performance according to pilot signal interval and pilot channel estimation error will be described with reference to FIG. 10 when applying a maximum Doppler frequency of 805.6 [Hz] in a 150 [km / h] environment.

10 is a graph illustrating BER performance according to pilot signal spacing and pilot channel estimation error when a maximum Doppler frequency of 805.6 [Hz] is applied in a 150 [km / h] environment.

Referring to FIG. 10, it can be seen that when the maximum Doppler frequency of 805.6 [Hz] is applied in the 150 [km / h] environment, the performance difference is more sensitive to errors in the pilot channel estimate than the pilot signal interval. . In particular, the performance when the pilot signal interval is 12, 5 OFDM symbols are borrowed, and the error variance of the pilot estimate is 0.032 (15 [dB]) is approximately 1.5 [dB at ideal channel estimation performance and BER 10 ^-4 . ], It can be seen that there is a difference in pilot channel estimation error, and also sensitive to pilot channel estimation error.

As described above with reference to FIGS. 8 to 10, it can be seen that the error of the LS channel estimate has a greater influence on the channel estimation performance than the pilot signal interval. In addition, even if the three-stage channel estimation process proposed by the present invention is not necessarily performed, that is, even if only the second-stage channel estimation process is performed, a performance degradation of about 0.2 [dB] occurs, which is superior to the general channel estimation method. You can see it. In particular, when the Doppler frequency is low, a sufficient pilot signal can be secured by performing only the first stage channel estimation process, and the time domain smoothing interval can be set relatively large in the second stage channel estimation process, so that only the second channel estimation process It can be seen that the performance is superior to that of the general channel estimation method.

Meanwhile, in the detailed description of the present invention, specific embodiments have been described, but various modifications are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the scope of the following claims, but also by the equivalents of the claims.

As described above, the present invention has the advantage that the channel estimation performance can be improved while using a minimum pilot signal by newly proposing a three-stage channel estimation scheme in an FH-OFDMA mobile communication system.

1 is a diagram illustrating an internal structure of a transmitter of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention.

2 is a diagram illustrating an internal structure of a receiver of an FH-OFDMA mobile communication system for performing a function in an embodiment of the present invention.

3 schematically illustrates a structure in which pilot signals are transmitted and a structure for pilot borrowing in an FH-OFDMA mobile communication system according to an embodiment of the present invention;

4 is a flowchart illustrating a multi-step channel estimation process according to an embodiment of the present invention.

FIG. 5 is a diagram schematically illustrating an operation in a first stage channel estimation process of FIG. 4.

FIG. 6 is a diagram schematically illustrating an operation in a second stage channel estimation process of FIG. 4.

FIG. 7 is a diagram schematically illustrating an operation in a third stage channel estimation process of FIG. 4.

8 shows a bit error rate according to a maximum Doppler frequency of 268.5 [Hz] in a 50 [km / h] environment and a pilot signal interval and number of OFDM symbols borrowed when using accurate CSI at a pilot signal position (BER) Rate, hereinafter referred to as BER) Graph showing performance

FIG. 9 shows BER performance according to a maximum Doppler frequency of 805.6 [Hz] in a 150 [km / h] environment and a pilot signal interval and the number of OFDM symbols borrowed when there is no error in the channel estimate at the pilot signal position. A graph

10 is a graph showing BER performance according to pilot signal interval and pilot channel estimation error when applying a maximum Doppler frequency of 805.6 [Hz] in a 150 [km / h] environment

Claims (34)

A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimation method of a mobile communication system using a hop-orthogonal frequency division multiple access scheme, Detecting the reference signals in each of the first symbol time periods by performing Fast Fourier transform on a signal received in the first predetermined number of symbol time periods including a symbol time period for which the channel is to be estimated; Generating a first channel estimate by interpolating reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals; Generating the second channel estimate by generating the first channel estimate and smoothing the first channel estimate in the time domain for each of the subcarriers in a third predetermined number of symbol time periods; After generating the second channel estimate, the second channel estimate is smoothed in a frequency domain in units of a fourth predetermined number of subcarrier bands for each symbol time interval to generate a final channel estimate of the symbol time interval to be estimated. The method comprising the step of. The method of claim 1, The generating of the first channel estimate may be performed by using the n-th order Lagrangian interpolation method of Equation 8, based on the reference signals detected in each of the first number of symbol time periods, in units of the second number of reference signals. Generating a first channel estimate. In Equation 8, Denotes the first channel estimate, k denotes the index of the subcarrier, Is an LS (Least Square) channel estimate of reference signals borrowed in time intervals other than the current symbol time interval, and k _b represents the number of subcarriers of the symbols including the borrowed reference signals. The method of claim 2, The generating of the second channel estimate may include generating the second channel estimate by using a curve joint method in the time domain for each of the subcarriers in the third number of symbol time periods. Said method. The method of claim 3, The curve joining method in the time domain is expressed by Equation 9 below. In Equation 9, i represents the symbol index, Denotes the η-th polynomial coefficient of the k-th subcarrier, η = 0, 1, ..., n, wherein Is expressed as in Equation 10 below. In Equation 10, Denotes the number of symbols to which the curved joint method is applied ( ), Denotes the first channel estimate of the i th symbol of the k th subcarrier. The method of claim 4, wherein The generating of the final channel estimate may include generating the final channel estimate by using the curve joint scheme in the frequency domain in the unit of the fourth number of subcarrier bands for each symbol time period. Said method. The method of claim 5, The curve joining method in the frequency domain is represented by the following equation (11). The method of claim 1, The first number is three or five, the second number is three, and the third number is 47 in a 50 [km / h] moving speed environment or in a 150 [km / h] moving speed environment. 15, and the fourth number is five. A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimation method of a mobile communication system using a hop-orthogonal frequency division multiple access scheme, Generating a first channel estimate by fast Fourier transforming the received signal and generating the first channel estimate using the reference signal borrowing method and the frequency domain interpolation method; Generating a second channel estimate using the smoothing method in the time domain after generating the first channel estimate; And generating a final channel estimate using the smoothing method in the frequency domain after generating the second channel estimate. The method of claim 8, The generating of the first channel estimate includes fast Fourier transforming a signal received in a first predetermined number of symbol time periods, including a symbol time period to estimate the channel, respectively, for each of the first number of symbol time periods. Detecting reference signals in And generating a first channel estimate by interpolating reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals. The method of claim 9, The generating of the second channel estimate may include generating the second channel estimate by smoothing the first channel estimate in the time domain for each of the subcarriers in a preset third number of symbol time periods. . The method of claim 10, The generating of the final channel estimate may be performed as a final channel estimate of the symbol time interval to be estimated by smoothing the second channel estimate in the frequency domain in units of a fourth predetermined number of subcarrier bands for each symbol time period. The method characterized in that the. The method of claim 11, The generating of the first channel estimate may be performed by using the n-th order Lagrangian interpolation method of Equation 12, based on the reference signals detected in each of the first number of symbol time periods in units of the second number of reference signals. Generating a first channel estimate. In Equation 12, Denotes the first channel estimate, k denotes the index of the subcarrier, Is a LS (Least Square) channel estimate of reference signals borrowed in time intervals other than the current symbol time interval, and k _b represents the number of subcarriers of the symbols including the borrowed reference signals. The method of claim 12, The generating of the second channel estimate may include generating the second channel estimate by using a curve joint method in the time domain for each of the subcarriers in the third number of symbol time periods. Said method. The method of claim 13, The curve joining method in the time domain is represented by the following equation (13). In Equation 13, i represents the symbol index, Denotes the η-th polynomial coefficient of the k-th subcarrier, η = 0, 1, ..., n, wherein Is expressed as in Equation 14 below. In Equation 14, Denotes the number of symbols to which the curved joint method is applied ( ), Denotes the first channel estimate of the i th symbol of the k th subcarrier. The method of claim 14, The generating of the final channel estimate may include generating the final channel estimate by using the curve joint scheme in the frequency domain in the unit of the fourth number of subcarrier bands for each symbol time period. Said method. The method of claim 15, The curve joining method in the frequency domain is represented by the following equation (15). The method of claim 11, The first number is three or five, the second number is three, and the third number is 47 in a 50 [km / h] moving speed environment or in a 150 [km / h] moving speed environment. 15, and the fourth number is five. A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimation apparatus of a mobile communication system using a hop-orthogonal frequency division multiple access scheme, A fast Fourier transformer for fast Fourier transforming a signal received in a first predetermined number of symbol time periods including a symbol time period to estimate the channel and detecting reference signals in each of the first number of symbol time periods; , The first channel estimate is generated by interpolating the reference signals detected in each of the first number of symbol time periods in the frequency domain in units of a second predetermined number of reference signals, and generating the first channel estimate in the third number of symbol time periods. After generating the second channel estimate by smoothing the first channel estimate in the time domain for each subcarrier, the second channel estimate is generated in the frequency domain in units of fourth preset subcarrier bands for each symbol time period. And a channel estimator for smoothing and estimating a final channel estimate of a symbol time period to be estimated. The method of claim 18, The channel estimator generates the first channel estimate by using the n-th order Lagrangian interpolation method of Equation 16, based on the reference signals detected in each of the first number of symbol time periods in units of the second number of reference signals. The device, characterized in that. In Equation 16, Denotes the first channel estimate, k denotes the index of the subcarrier, Is an LS (Least Square) channel estimate of reference signals borrowed in time intervals other than the current symbol time interval, and k _b represents the number of subcarriers of the symbols including the borrowed reference signals. The method of claim 19, And the channel estimator generates the first channel estimate for each of the subcarriers in the third number of symbol time periods as the second channel estimate using a curve joint scheme in a time domain. The method of claim 20, The curve bonding method in the time domain is represented by the following equation (17). In Equation 17, i represents the symbol index, Denotes the η-th polynomial coefficient of the k-th subcarrier, η = 0, 1, ..., n, wherein Is expressed as in Equation 18 below. In Equation 18, Denotes the number of symbols to which the curved joint method is applied ( ), Denotes the first channel estimate of the i th symbol of the k th subcarrier. The method of claim 21, And the channel estimator generates the second channel estimate as the final channel estimate in the frequency domain in units of the fourth number of subcarrier bands for each symbol time period. The method of claim 22, The curve joining method in the frequency domain is characterized by the following equation (19). The method of claim 18, The first number is three or five, the second number is three, and the third number is 47 in a 50 [km / h] moving speed environment or in a 150 [km / h] moving speed environment. And the fourth number is five. A frequency that divides an entire frequency band into a plurality of subcarrier bands, transmits reference signals in the subcarrier bands, and transmits data signals in subcarrier bands other than the subcarrier bands in which the reference signals are transmitted A channel estimation apparatus of a mobile communication system using a hop-orthogonal frequency division multiple access scheme, A fast Fourier transformer which detects reference signals by fast Fourier transforming the received signal; After generating the first channel estimate using the reference signal borrowing method and the frequency domain interpolation method, and generating the second channel estimate using the smoothing method in the time domain, And a channel estimator for estimating the second channel estimate as a final channel estimate using a smoothing scheme. The method of claim 25, The channel estimator detects reference signals in each of the first number of symbol time periods by fast Fourier transforming a signal received in a first predetermined number of symbol time periods including a symbol time period for which the channel is to be estimated. And generating a first channel estimate by interpolating reference signals detected in each of the first number of symbol time periods in a frequency domain in units of a second predetermined number of reference signals. The method of claim 26, Wherein the channel estimator smoothes the first channel estimate in the time domain for each of the subcarriers in a third predetermined number of symbol time periods to generate the second channel estimate. The method of claim 27, The channel estimator smoothes the second channel estimate in the frequency domain in units of fourth preset subcarrier bands for each symbol time period, and estimates the final channel estimate of the symbol time period to be estimated. The device. The method of claim 28, The channel estimator generates the first channel estimate by using the n-th order Lagrangian interpolation method of Equation 20, based on the reference signals detected in each of the first number of symbol time periods in units of the second number of reference signals. The device, characterized in that. In Equation 20, Denotes the first channel estimate, k denotes the index of the subcarrier, Is an LS (Least Square) channel estimate of reference signals borrowed in time intervals other than the current symbol time interval, and k _b represents the number of subcarriers of the symbols including the borrowed reference signals. The method of claim 29, And the channel estimator generates the second channel estimate in the third number of symbol time periods by using a curve joint method in the time domain for each of the subcarriers. The method of claim 30, The apparatus for curve joining in the time domain is expressed by Equation 21 below. In Equation 21, i represents the symbol index, Denotes the η-th polynomial coefficient of the k-th subcarrier, η = 0, 1, ..., n, wherein Is expressed as in Equation 22 below. In Equation 22, Denotes the number of symbols to which the curved joint method is applied ( ), Denotes the first channel estimate of the i th symbol of the k th subcarrier. The method of claim 31, wherein And the channel estimator generates the second channel estimate as the final channel estimate in the frequency domain in units of the fourth number of subcarrier bands for each symbol time period. 33. The method of claim 32, The curve bonding method in the frequency domain is characterized in that expressed by the following equation (23). The method of claim 28, The first number is three or five, the second number is three, and the third number is 47 in a 50 [km / h] moving speed environment or in a 150 [km / h] moving speed environment. And the fourth number is five.