Frequency Offset Estimation Method in OFDM System

Abstract

The present invention discloses a frequency deviation estimation method in an OFDM system, the method comprising: a transmitting end sends a signal x(n) at time T0; a receiving end collects a real-valued signal transmitted at time T1 to obtain a received signal, and a processor is used to obtain: a fast Fourier transform is performed on the analytical signal by the processor to obtain: W(n) is taken as 0; a new function is established by the processor: Y1=Y(1:(N-1)), Y2=Y(2:N), and Y is obtained. New =X * (n‑ε)·X(n‑ε); X1=X(1:(N‑1)), X2=X((2:N)) is obtained by establishing a new function through a processor and outputting it. The present invention realizes rapid measurement of frequency deviation based on time domain received signals, and can be widely used in existing wireless uplink and downlink communication systems to solve the problem of time / frequency asynchronism during communication between terminals.

Description

Technical Field

[0001] The invention relates to a frequency deviation estimation method in an OFDM system, and belongs to the technical field of carrier waves. Background Art

[0002] In the traditional parallel transmission system, the entire signal band is divided into multiple parallel non-orthogonal subcarriers, each subcarrier is modulated separately, and the entire system constitutes frequency division multiplexing. There is no spectrum overlap between two adjacent subcarriers, which is conducive to eliminating subcarrier interference. However, this frequency division multiplexing method reduces the utilization of the spectrum. In order to improve the utilization of the spectrum, the idea of ​​parallel transmission and FDM with spectrum overlap, namely orthogonal frequency division multiplexing OFDM, was proposed in the mid-1960s.

[0003] Multi-carrier technology divides the transmission bandwidth into many narrowband sub-carriers for parallel transmission, which can achieve a higher transmission rate in a limited bandwidth. In traditional frequency division multiplexing technology, the sub-carriers do not overlap with each other and a protection interval is added, resulting in low frequency utilization. In OFDM technology, the sub-carriers are orthogonal. From the perspective of the time domain, each sub-carrier has an integer multiple of periods within an OFDM symbol period, and each adjacent sub-carrier differs by one period. From the perspective of the frequency domain, at the center of each sub-carrier, the values ​​of other sub-carriers are zero, thereby saving precious frequency resources and improving frequency utilization.

[0004] In the OFDM system, the modulated digital signal at the transmitter is converted into an analog signal through D / A, modulated by the RF center frequency, and sent into the wireless channel for transmission. At the receiving end, the RF analog signal is first adjusted, moved from RF to baseband, and then sampled and digitized at a sampling rate F, and finally sent to the OFDM demodulator. However, if a carrier frequency deviation Δf occurs due to unstable crystal oscillators at both ends of the transmitter and receiver, the Doppler effect in a mobile environment, or phase noise introduced into a nonlinear channel, not only will the signal of the corresponding channel fall on other channels when adjusting the signal, resulting in bit errors, but it may also destroy the orthogonality of the subcarriers and cause interference between subcarriers.

[0005] In the prior art, Schmidl, Cox and Paul H. Moose proposed a method for using two OFDM symbols for frequency synchronization to understand the starting points of two repeated OFDM symbols and thus estimate the frequency offset (study and analysis of Carrier frequency offset (CFO) in OFDM). On the one hand, this method requires at least two repeated OFDM symbols, with a large amount of calculation and slow measurement speed. On the other hand, it can only estimate the frequency offset on a bandwidth of ±0.5db, with large errors and low accuracy. Summary of the invention

[0006] The purpose of the present invention is to provide a frequency offset estimation method in an OFDM system. The method and device not only solve the problem of channel confusion and mutual interference between subcarriers caused by frequency offset of OFDM structure signals in the prior art, but also solve the problem of large estimation difference and large correction deviation.

[0007] To achieve the above object, the technical solution adopted by the present invention is: a frequency offset estimation method in an OFDM system, the method comprising:

[0008] The transmitter sends a signal X(n) at time T0;

[0009] The receiving end collects the transmitted real-valued signal at time T1 to obtain the received signal. When the received signal at the receiving end has a frequency offset ε, the processor obtains: τ=T1-T0, w(n) is the sampling value of Gaussian noise;

[0010] The processor performs a fast Fourier transform on the analytical signal to obtain: Take W(n) to be 0;

[0011] The processor creates a new function: Y1 = Y(1:(N-1)), Y2 = Y(2:N), and we get Screen out constant terms Get Y New =X * (n-ε)·X(n-ε); X1=X(1: (N-1)), X2=X((2:N)), we get

[0012] Creating new functions through handlers get And output.

[0013] The further improved scheme in the above technical scheme is as follows:

[0014] 1. In the above solution, the receiving end is a microphone or an antenna.

[0015] 2. In the above scheme, the real-valued signal collected by the receiving end is filtered out through a filter to remove signals of other frequencies except the real-valued signal.

[0016] 3. In the above scheme, the real-valued signal and the imaginary-valued signal transformed from the real-valued signal are respectively passed through a low-pass filter, and then down-sampled by a sampler to output real and imaginary data.

[0017] 4. In the above scheme, a pilot signal is used to replace the real-valued signal.

[0018] 5. In the above solution, the received signal y(λ) is a discrete signal.

[0019] 6. In the above scheme, oversampling is performed by the processor, and the oversampling multiple is M, M = 2, 4, 6, 8, 16..., and the frequency domain sampling points are subdivided to interval;

[0020] Obtained by the processor Take W(k) to be 0;

[0021] Get ε through the processor new =[M·ε] round ,get And output.

[0022] 7. In the above scheme, synchronous output and

[0023] 8. In the above scheme, when oversampling is adopted, the processor only outputs

[0024] Due to the application of the above technical solution, the present invention has the following advantages compared with the prior art:

[0025] 1. The frequency offset estimation method in the OFDM system of the present invention is based on the time domain received signal and establishes a new function. It only needs to obtain one OFDM symbol (one frame) through the observation window to realize the rapid measurement of the frequency offset. It can be widely used in the existing wireless uplink and downlink communication systems to solve the time / frequency asynchronism problem when communicating between terminals. It has fast measurement speed and high collection and correction efficiency.

[0026] 2. The frequency offset estimation method in the OFDM system of the present invention is different from the measurement method proposed by Schmidl, Cox and Paul H. Moose, and can realize frequency offset estimation under full bandwidth, with stronger applicability and higher accuracy.

[0027] 3. The frequency offset estimation method in the OFDM system of the present invention improves the estimation accuracy of the frequency offset by optimizing the sampling method and utilizing oversampling, thereby facilitating improving the correction accuracy and reducing the error. DETAILED DESCRIPTION

[0028] The method and device of the present invention are used to solve the frequency deviation problem of OFDM signal structure and similar signal structures.

[0029] Embodiment 1: A frequency offset estimation method in an OFDM system, the method comprising:

[0030] After the transmitter sends out the signal x(n) at time T0, a microphone is used as the receiver to collect the transmitted real-valued signal at time T1 to obtain an audio signal (received signal), and the audio signal is input into the filter. Here, the filter frequency is set according to the audio signal frequency to filter out clutter for subsequent processing.

[0031] The filtered real-valued signal x(n) is phase-shifted by the processor to obtain a complex-valued analytical signal: τ=T1-T0, w(n) is the sampling value of Gaussian noise.

[0032] The processor further performs a fast Fourier transform (FFT) on the analytical signal, transforming it from the time domain to the frequency domain to facilitate the analysis of the frequency deviation, and obtains: Take W(n) as 0; establish new function Y1=Y(1:(N-1)), Y2=Y(2:N), and get Screen out constant terms Get Y New =X * (n-ε)·X(n-ε); X1=X(1: (N-1)), X2=X(2:N)), we get

[0033] Creating new functions through handlers get And output.

[0034] Embodiment 2: A method for estimating frequency offset in an OFDM system is as follows:

[0035] After the transmitter sends out the signal x(n) at time T0, a microphone is used as the receiver to collect the transmitted real-valued signal at time T1 to obtain an audio signal (received signal), and the audio signal is input into the filter. Here, the filter frequency is set according to the audio signal frequency to filter out clutter for subsequent processing.

[0036] The processor performs phase transformation on the filtered real-valued signal x(n). The real-valued signal and the imaginary-valued signal are respectively passed through a low-pass filter, and then downsampled by the sampler to output the real and imaginary data, and the complex analytical signal is combined: τ=T1-T0, w(n) is the sampling value of Gaussian noise.

[0037] The processor further performs a fast Fourier transform (FFT) on the analytical signal, transforming it from the time domain to the frequency domain to facilitate the analysis of the frequency deviation, and obtains: Take W(n) as 0; establish new function Y1=Y(1:(N-1)), Y2=Y(2:N), and get Screen out constant terms Get Y New =X* (ns)·X(n-ε); X1=X(1: (N-1)), X2=x((2:N)), we get

[0038] Creating new functions through handlers get

[0039] Oversampling is performed through the processor, the oversampling multiple is M, M = 2, 4, 6, 8, 16..., and the frequency domain sampling points are subdivided to Interval, let k = M*n, and further perform fast Fourier transform (FFT) on the analytical signal through the processor, transforming from the time domain to the frequency domain to facilitate the analysis of frequency deviation, and obtain: Take W(k) to be 0.

[0040] The processor uses the same steps as before oversampling to establish a new function based on Y(k) to obtain an estimate of Mε, thereby obtaining ε new =[M·ε] round ,get

[0041] Here, assuming M = 4, ε = 1.31, the normal sampling and The mean square error is 0.0036<0.0961. It can be seen that the oversampling method can improve the accuracy of frequency offset estimation, and the larger M is, the more accurate the estimation is;

[0042] Therefore, take Output as frequency deviation value.

[0043] The pilot signal may be used to replace the real-valued signal to perform measurement and obtain corrected data.

[0044] By adopting the above scheme, a new function is established based on the received signal in the time domain. It only needs to obtain one OFDM symbol (one frame) through the observation window to realize the rapid measurement of frequency deviation. It can be widely used in existing wireless uplink and downlink communication systems to solve the problem of time / frequency asynchrony during communication between terminals. It has fast measurement speed and high collection and correction efficiency.

[0045] In addition, different from the measurement method proposed by Schmidl, Cox and Paul H. Moose, it can realize frequency offset estimation under full bandwidth, which is more applicable and more accurate.

[0046] In addition, by optimizing the sampling method and utilizing oversampling, the estimation accuracy of the frequency offset can be improved, thereby facilitating the improvement of the correction accuracy and the reduction of the error.

[0047] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A frequency offset estimation method in an OFDM system, characterized in that: The method comprises: The transmitter sends a signal x(n) at time T0; The receiving end collects the transmitted real-valued signal at time T1 to obtain the received signal, which is recorded as y(n). Both x(n) and y(n) are time domain baseband signals. When the received signal at the receiving end has a frequency offset ε, the processor obtains: τ=T1-T0, w(n) is the sampling value of Gaussian noise; The processor performs a fast Fourier transform on the analytical signal to obtain: n=1, 2...N, take W(n) as 0; Create a new function through the processor: Y1 = Y(1:(N-1)), Y2 = Y(2:N), and get Screen out constant terms Get Y New =X * (n-ε)·X(n-ε); X1=X(1:(N-1)), X2=X((2:N)), we get Creating new functions through handlers get And output.

2. The frequency offset estimation method in an OFDM system according to claim 1, characterized in that: The receiving end is a microphone or an antenna.

3. The frequency offset estimation method in an OFDM system according to claim 1, characterized in that: The real-valued signal collected by the receiving end is filtered out through a filter to remove signals of other frequencies except the real-valued signal.

4. The frequency offset estimation method in an OFDM system according to claim 1, characterized in that: The real-valued signal and the imaginary-valued signal transformed from the real-valued signal are respectively passed through a low-pass filter, and then down-sampled by a sampler to output real and imaginary data.

5. The frequency offset estimation method in an OFDM system according to claim 1, characterized in that: A pilot signal is used instead of a real-valued signal.

6. The frequency offset estimation method in an OFDM system according to claim 1, characterized in that: The received signal y(n) is a discrete signal.

7. The frequency offset estimation method in an OFDM system according to claim 4, characterized in that: Oversampling is performed through the processor, the oversampling multiple is M, M = 2, 4, 6, 8, 16..., and the frequency domain sampling points are subdivided to interval; Obtained by the processor k=M*n, take W(k) as 0; Get ε through the processor new =[M·ε] round ,get And output.

8. The frequency offset estimation method in an OFDM system according to claim 7, characterized in that: Synchronous output and 9. The frequency offset estimation method in an OFDM system according to claim 7, characterized in that: When oversampling is used, the processor only outputs

Images