EP1958409A1

EP1958409A1 - Apparatus for estimating time and frequency offset using antenna diversity in ofdm communication system and method thereof

Info

Publication number: EP1958409A1
Application number: EP06824083A
Authority: EP
Inventors: Hee-Wook Kim; Kun-Seok Kang; Do-Seob Ahn; Ho-Jin Lee
Original assignee: Electronics and Telecommunications Research Institute ETRI
Current assignee: Electronics and Telecommunications Research Institute ETRI
Priority date: 2005-12-09
Filing date: 2006-12-11
Publication date: 2008-08-20
Also published as: KR20070060928A; KR100770008B1; WO2007067018A1

Abstract

Provided are an apparatus for estimating a time offset and a frequency offset and a method thereof . The method includes the steps of : a) giving weights to the signals received through the antennas and generating an input signal; b) delaying the input signal by a predetermined index and adding a square value of the delayed signal and a square value of the input signal; c) summating a predetermined number of the output value of the step b) and multiplying the summed value and a predetermined value; d) adding the input signal and a conjugated delayed signal; e) summating a predetermined number of the output value of the step d); f ) detecting the maximum value of the time offset; and g) calculating a radian degree for the absolute value of the output value acquired in the step e) and calculating a frequency offset based on the detected time offset.

Description

APPARATUS FOR ESTIMATING TIME AND FREQUENCY OFFSET USING ANTENNA DIVERSITY IN OFDM COMMUNICATION SYSTEM AND

METHOD THEREOF Description

Technical Field

The present invention relates to an apparatus for estimating a time offset and a frequency offset in an Orthogonal Frequency Division Multiplexing (OFDM) communication system, and a method thereof; and, more particularly, to an apparatus for estimating a time offset and a frequency offset that can estimate a time offset and a frequency offset in an environment having a low signal-to-noise ratio (SNR) and multi-path channels by combining diversities of antennas in an OFDM communication system, and a method thereof.

Background Art

An Orthogonal Frequency Division Multiplexing (OFDM) system includes an Orthogonal Frequency Division Multiple Access (OFDMA) system and an Orthogonal Frequency and Code Division Multiplexing (OFCDM) system.

The OFDMA method which is an effective digital signal transmission method based on limited bandwidth and it is proposed by Chang. Recently, the OFDMA method is applied to a digital signal transmission system such as a digital audio broadcasting or a digital video broadcasting in Europe, an Asymmetric Digital Subscriber Line (ADSL), and a wireless Local Area Network (WLAN).

The OFDMA method has advantages that it is strong to an inter-symbol interference (ISI) discussed for high- efficiency of frequency usage and high-speed communication and that it makes a selective fading of frequency look like a non-selective fading. However, the OFDM access method has a serious disadvantage that inconformity of oscillators between a transmitter and a receiver, and sensitiveness of carrier frequency error by a Doppler frequency variation in the receiver. The carrier frequency error breaks orthogonality between sub carriers in the OFDM system so that the sub carriers are influenced by an inter-carrier interference (ICI) from the other sub carriers. Therefore, a small amount of a carrier frequency error causes bad performance of the OFDM system. The matter of frequency synchronization is important in realizing an OFDM system.

Conventional methods for solving the above problems include a data-dependent method and a blind method.

First, the data-dependent method using two pilot symbols for time synchronization and estimating an integer and a fractional part of a frequency offset to estimate frequency error is faster and more reliable than the blind method. However, data rate and power efficiency of the data-dependent method are decreased due to the use of two pilot symbols.

Meanwhile, since the blind method synchronizes frequencies based on a cyclic prefix (CP) inserted for decreasing the ISI, data rate or power efficiency is not decreased.

In the blind method, the CP and maximum likelihood (ML), which is known as the best estimation method of frequency synchronization, are used to estimate and compensate for the frequency offset. The blind method uses a characteristic that there is only a phase difference between a CP and a copied CP when time is synchronized, and the others except the phase difference are same. However, performance of the blind method presents fine performance at a high signal-to-noise ratio (SNR) of an Additive White Gaussian Noise (AWGN) channel but poor performance in an environment of a low SNR and a multi-path channel.

Disclosure

Technical Problem

It is, therefore, an object of the present invention devised to resolve the above problems is to provide an apparatus for estimating a time offset and a frequency offset that can estimate a time offset and a frequency offset in a communication environment of a low signal-to- noise ratio (SNR) and multi-path channels by combining diversities of antennas in an Orthogonal Frequency Division Multiplexing (OFDM) communication system, and a method thereof.

Other objects and advantages of the present invention will be clearly understood by the following description and embodiments. Also, it is obvious to those skilled in the art that the objects and advantages of the present invention can be realized by the means as claimed and combinations thereof.

Technical Solution In accordance with one aspect of the present invention, there is provided an apparatus for estimating a time offset and a frequency offset based on signals inputted through at least two antennas, including: a signal generating unit for giving weights to each of the signals received through the antennas and generating an input signal; a delay unit for delaying the input signal by a predetermined index; a first adding unit for adding a square value of the delayed signal and a square value of the input signal; a first summating unit for summating a predetermined number of the output value of the first adding unit and multiplying the summed value by a predetermined value, which is decided based on the weights, signal power and noise power, to thereby acquire a first summation value; a second adding unit for adding the input signal and a conjugated signal obtained by changing the sign of an imaginary part of the complex delayed signal; a second summating unit for summating a predetermined number of the output value of the second adding unit to thereby acquire a second summation value; a time offset detecting unit for detecting the maximum value by subtracting an output value of the first summating unit from an absolute value of an output value of the second summating unit; and a frequency offset calculating unit for calculating a radian degree for the absolute value of the output value of the second summating unit and calculating a frequency offset based on the detected time offset acquired in the time offset detecting unit.

In accordance with another aspect of the present invention, there is provided a method for estimating a time offset and a frequency offset based on signals inputted through at least two antennas, including the steps of: a) giving weights to the signals received through the antennas and generating an input signal; b) delaying the input signal by a predetermined index and adding a square value of the delayed signal and a square value of the input signal; c) summating a predetermined number of the output value of the step b) and multiplying the summed value and a predetermined value, which is decided based on the weights, signal power and noise power; d) adding the input signal and a conjugated delayed signal, which is obtained by changing the sign of an imaginary part of the complex delayed signal; e) summating a predetermined number of the output value of the step d) ; f) detecting the maximum value of the time offset by subtracting the output value of the step c) from the absolute value of the the output value of the step e); and g) calculating a radian degree for the absolute value of the output value acquired in the step e) and calculating a frequency offset based on the detected time offset acquired in the step f).

Advantageous Effects The present invention provides excellent performances of estimating a time offset and a frequency offset by using a weight in an environment having a low signal-to-noise ratio (SNR) and multi-path channels in an OFDM communication system.

Description of Drawings

The above and other objects and features of the present invention will become apparent from the following description of the preferred embodiments given in conjunction with the accompanying drawings, in which:

Fig. 1 is a block diagram illustrating a general Orthogonal Frequency Division Multiplexing (OFDM) system;

Fig. 2 is a diagram illustrating a general symbol structure of the OFDM system;

Fig. 3 is a diagram illustrating an OFDM system having two antennas in accordance with an embodiment of the present invention; and

Fig. 4 is a block diagram illustrating an apparatus for estimating a time offset and a frequency offset in accordance with an embodiment of the present invention.

Best Mode for the Invention Other objects and aspects of the invention will become apparent from the following description of the embodiments with reference to the accompanying drawings, which is set forth hereinafter. Also, when it is considered that detailed description on a related art may obscure the points of the present invention unnecessarily in describing the present invention, the description will not be provided herein. Hereinafter, specific embodiments of the present invention will be described with reference to the accompanying drawings.

In the embodiment of the present invention, a basic Orthogonal Frequency Division Multiplexing (OFDM) system using two antennas is described, but it is apparent that the method and apparatus of the present invention can be applied to up and down links of an OFDM-based system using a plurality of antennas such as OFDMA.

Fig. 1 is a block diagram illustrating a general OFDM system. Fig. 1 shows data transmission between a transmitting antenna and a receiving antenna.

As shown in Fig. 1, a transmission signal is transmitted to a receiving block through an inverse discrete Fourier transformer (IDFT) 11 and a parallel-to- serial converter 12, the receiving block receives the transmission signal through a serial-to-parallel converter 13 and a discrete Fourier transformer (DFT) 14 reversely.

In Fig. 1, S₀, Si, S₂ to S_L_i are cyclic prefixes (CP). The OFDM system illustrated in Fig. 1 uses N carriers and L CPs in order to transmit one OFDM symbol. The Maximum likelihood (ML) estimation of the OFDM system will be described referring to Fig. 2.

As shown in Fig. 2, 2N+L samples are observed. Herein, an I sector is a CP in the i^th OFDM symbol and an I' sector is a data sample which is duplicated into the I sector in the i^th OFDM symbol. Since samples are received randomly except I and I' , each received sample is independent from each other. Since I and I' are the same samples, there is a correlation between I and I' . A probability density function of a received signal r is obtained when a time offset (θ) and a frequency offset (ε) are predetermined, and the time offset (θ) and the frequency offset (ε) are estimated by acquiring θ and ε values that maximize the log-likelihood function of the probability density function. Since, the above estimating method of the time offset and the frequency offset, i.e., the maximum likelihood (ML) estimation, is widely known, detailed description on it will be omitted.

Fig. 3 is a diagram illustrating an OFDM system having two antennas to help technical understanding prior to description of a time and frequency offset estimation method in accordance with an embodiment of the present invention.

An OFDM diversity receiver includes two antennas, two decoders for converting analog signals received through each antenna into digital signals and decoding the digital signals into original signals by performing fast Fourier transformation, two equalizers for compensating for distortion of the decoded signals occurring in transmission, and a combiner for combining the distortion-compensated signals acquired in the two equalizers.

In Fig. 3, it is assumed that a first antenna receives an ri signal, and a second antenna receives an r₂ signal. The two antennas are departed from each other by over λ/2 so that the two reception signals are independent from each other. The two signals ri and r₂ are multiplied by two weights pi and p₂ based on an antenna combining method, respectively. In present invention, ML estimation is induced based on the weighted signals, inputted through a plurality of antennas to acquire a diversity gain. When the channel is assumed to be an AWGN channel, the received signal r is expressed as following Eq. 1. _{E χ}

In Eq. 1, ri(k) and r₂(k) are the k^th received signal in the observation duration 2N+L from the first antenna and the second antenna, respectively. An integrated k^th signal r(k) is obtained by multiplying weights and each received signals. Herein, when r(k), samples can be divided into those existing in I and those not existing in I. There is a correlation between I and I', which is apart by N from I, and other signals are transmitted randomly because samples existing in I are duplicated CPs of samples in I' . Correlations between received signals in the observation duration k which is from 1 to 2N+L are represented as following Eq. 2. That is, when k is included in I, autocorrelation characteristics are obtained due to correlation between I and I' . In addition, when k is out of the in I and I' , there is no correlation relationship.

Herein, σ_s ² is an average power of transmission signals; σ_N ² is an average power of noise; and ε is a frequency offset.

It can be seen from Eq. 2 that using at least two antennas acquires an antenna diversity gain based on weights pi and p₂, compared to a conventional method using one antenna.

A log-likelihood function of the time offset (θ) and the frequency offset (ε) for estimating a maximum likelihood (ML) is a logarithm of a probability density function for 2N+L observation signals. The log- likelihood function is expressed as following Eq. 3.

A(θ,ε)=\og(f(r/Θ,ε))

=\og(Uf(r(klr(k+N))U f(r(k))

Herein, r is a received signal vector including received signals acquired in 2N+L observation durations; the received signal vector r is represented as [r(l)... r(2N+L)]^T; and f is the probability density function.

When the received signal vector r has Gaussian distribution, the log-likelihood function of Eq. 3 can be expressed as following Eq. 4. λφ,ε)=\y(θ)\cos(2πε+ly(θ))-p₂Φ(θ) Eq.

Herein, γ(m) denotes a sum of L continuous correlation values between received signals apart by N; Φ(m) is an energy of received signals; and p₂ is a magnitude of a correlation coefficient between r(k) and r(k+N). The γ(m) , Φ(m) and p₂ are expressed as following Eq. 5.

The ML is estimated by acquiring a time offset value and a frequency offset value maximizing the log- likelihood function as shown in the following Eq. 6. m&x.A(6,έ) = maxmaxΛ(^s) = msx.max.(γ(θ)\cos(2πs+^χ(_θy)~pΦ(θ))

= maxΛ(5»^,^) =max(|K^)|-ΛΦ(^)) ^^'

Therefore, the time offset and the frequency offset estimated based on Eq. 6 are expressed as the following Eq. 7.

As mentioned in the description referring to Eq. 2, the antenna diversity gain is obtained from the reception signal r(k). Moreover, referring to Eq. 7, since the correlation coefficient p₂ is larger than that of using one antenna due to the antenna diversity gain, performance of the apparatus for estimating time offset and frequency offset is more excellent than the conventional apparatus.

Fig. 4 is a block diagram illustrating an apparatus for estimating a time offset and a frequency offset in accordance with an embodiment of the present . invention .

The time offset is estimated based on the signals received through two antennas and then the frequency offset is estimated. Herein, pi and p₂ are different based on an antenna combining method. In case of an identical gain combining method, pi and p₂ are equal to 1. In case of a selection combining method, one having a large signal-to-noise ratio (SNR) is 1 and the other is 0. In case of a maximum ratio combining (MRC) method, pi and p₂ determined such that a sum of two received signals is the maximum SNR.

Referring to Fig. 4, the apparatus for estimating the time offset and the frequency offset in accordance with the present invention will be described.

The apparatus for estimating the time offset and the frequency offset includes a signal generator 100, a delay unit 101, a first square value calculator 102, a second square value calculator 103, a first adder 104, a first summator 105, a conjugator 106, a second adder 107, a second summator 108, a subtractor 109, a time offset detector 110 and a frequency offset calculator 111.

The signal generator 100 receives two signals transmitted through two antennas, gives weights to the two signals, and integrates the two weighted signals to thereby generate an input signal. The delay unit 101 delays the input signal with a predetermined index Z . The first square value calculator 102 calculates a first square value of the delayed signal. The second square value calculator 103 calculates a second square value of the input signal. The first adder 104 adds up the first square value and the second square value to thereby generate a first adding value. The first summator 105 summates the predetermined number of the first adding value and multiplies the summated value and a predetermined value which is decided based on the weights, signal power and noise power.

The conjugator 106 conjugates the delayed signal which changes the sign of imaginary part of the complex delayed signal. The second adder 107 adds up the input signal and the delayed signal to thereby generate a second adding value. The second summator 108 summates the predetermined number of the second adding value. The subtractor 109 subtracts the output value from the first summator 105 from the output value from the second summator 108. The time offset detector 110 detects the maximum time offset of the output values from the subtractor 109. The frequency offset calculator 111 calculates a radian degree of the absolute value of the output value from the second summator 108 and calculates a frequency offset based on the detected time offset in the time offset detector 110.

One of the outstanding features of the time offset and the frequency offset estimating apparatus, which is suggested in the present invention, is adding a square value of the delayed input signal and a square value of the input signal to thereby produce an added value, summating a predetermined number of the added value, and estimating the time offset based on a predetermined value acquired based on weights given to the two received signals, signal power and noise power. Hereinafter, the method for estimating the time offset and the frequency offset in the present invention will be described referring to Fig. 4.

First, an input signal is generated by giving weights to signals received through two antennas. Then, the input signal is delayed by a predetermined index Z . A first square value, which is a square value of the delayed signal, is calculated; and a second square value, which is a square value of the input signal, is calculated. Then, the first square value and the second square value are added to thereby generate a first adding value .

Next, the predetermined number of the first adding values are summated, the summated value and a predetermined value p₂/2, which is acquired based on the weights, signal power and noise power, are multiplied to thereby calculate a first summation value.

At the same time, the sign of imaginary part of the complex delayed signal is changed and the input signal and the conjugated delayed signal are added to thereby generate a second adding value. The predetermined number of the second adding values are summated to thereby calculate a second summation value. The absolute value of the second summation value is obtained.

Next, the maximum time offset is acquired by- subtracting the first summation value from the absolute value of the second summation value.

Then, the frequency offset is estimated based on the time offset. That is, radian degree of the absolute value of the second summation value is obtained and the frequency offset is calculated based on the detected time offset.

While the present invention has been described with respect to certain preferred embodiments, it will be apparent to those skilled in the art that various changes and modifications may be made without departing from the scope of the invention as defined in the following claims,

Claims

What is claimed is:

1. An apparatus for estimating a time offset and a frequency offset based on signals inputted through at least two antennas, comprising:

a signal generating means for giving weights to each of the signals received through the antennas and generating an input signal;

a delay means for delaying the input signal by a predetermined index;

a first adding means for adding a square value of the delayed signal and a square value of the input signal ;

a first summating means for summating a predetermined number of the output value of the first adding means and multiplying the summed value by a predetermined value, which is decided based on the weights, signal power and noise power, to thereby acquire a first summation value;

a second adding means for adding the input signal and a conjugated signal obtained by changing the sign of an imaginary part of the complex delayed signal;

a second summating means for summating a predetermined number of the output value of the second adding means to thereby acquire a second summation value; a time offset detecting means for detecting the maximum value by subtracting an output value of the first summating means from an absolute value of an output value of the second summating means; and

a frequency offset calculating means for calculating a radian degree for the absolute value of the output value of the second summating means and calculating a frequency offset based on the detected time offset acquired in the time offset detecting means.

2. The apparatus as recited in the claim 1, wherein the first summating means summates the predetermined number of the output value of the first adding means and multiplies the summing value by a predetermined value p₂, which is decided based on Eq. 8 by using the weights pi and p₂, the signal power σ_s ² and the noise power , is expressed by an equation as:

3. The apparatus as recited in the claim 2, wherein the weights pi and p₂ are equal to 1 when antenna combining method is an identical gain combining method.

4. The apparatus as recited in the claim 2, wherein the weights pi and p₂ are determined such that one having a larger signal-to-noise ratio (SNR) is 1 and the other is 0 when the antenna combining method is a selection combining method.

5. The apparatus as recited in the claim 2, wherein the weights pi and p₂ are determined such that a sum of the two received signals is the maximum SNR when the antenna combining method is a maximum ratio combining method.

6. A method for estimating a time offset and a frequency offset based on signals inputted through at least two antennas, comprising the steps of:

a) giving weights to the signals received through the antennas and generating an input signal;

b) delaying the input signal by a predetermined index and adding a square value of the delayed signal and a square value of the input signal; c) summating a predetermined number of the output value of the step b) and multiplying the summed value and a predetermined value, which is decided based on the weights, signal power and noise power;

d) adding the input signal and a conjugated delayed signal, which is obtained by changing the sign of an imaginary part of the complex delayed signal;

e) summating a predetermined number of the output value of the step d) ;

f) detecting the maximum value of the time offset by subtracting the output value of the step c) from the absolute value of the the output value of the step e); and

g) calculating a radian degree for the absolute value of the output value acquired in the step e) and calculating a frequency offset based on the detected time offset acquired in the step f).

7. The method as recited in the claim 6, in the step c), wherein the predetermined number of the output value of the step b) are summated, and the summing value is multiplied by a predetermined value p₂ which is decided based on the weights pi and p₂, a signal power σ_s ² and a noise power σ_n ² , in the following equation 9:

8. The method as recited in the claim 7, wherein the weights pi and p₂ are equal to 1 when antenna combining method is an identical gain combining method.

9. The method as recited in the claim 7, wherein the weights pi and p₂ are determined such that one having a larger signal-to-noise ratio (SNR) is 1 and the other is 0 when the antenna combining method is a selection combining method.

10. The method as recited in the claim 7, wherein the weights pi and p₂ are determined such that a sum of the two received signals is the maximum SNR when the antenna combining method is a maximum ratio combining method.