Channel estimator with high noise suppression and low interpolation error for OFDM systems

• US 8,406,319 B2
• Filed: 03/27/2007
• Issued: 03/26/2013
• Est. Priority Date: 03/27/2007
• Status: Active Grant

• Current Assignee: Google Technology Holdings LLC (Alphabet Inc.)
• Original Assignee: Motorola Mobility LLC (Lenovo Group Ltd.)
• Inventors: Sayana, Krishna Kamal, Nollett, Bryan S., Zhuang, Xiangyang
• Primary Examiner(s): Washburn, Daniel
• Assistant Examiner(s): HAILEGIORGIS, FITWI Y
• Application Number:
  US11/691,494
• Publication Number:
  US 20120195397A1
• Time in Patent Office: 2,191 Days
• Field of Search: 375/232, 375/260, 375/285, 375/350, 375/360, 375/340, 370/206, 370/208, 370/210, 370/252, 455/59
• US Class Current: 375/260
• CPC Class Codes: H04L 25/022 of frequency response
• CPC Class Codes: H04L 25/0228 with direct estimation from sounding signals
• CPC Class Codes: H04L 25/025 using least-mean-square [LMS] method
• CPC Class Codes: H04L 27/2647 Arrangements specific to the receiver only equalisation H04L27/01

Abstract

A receiver with DFT based channel estimation having good noise suppression for both high and low signal-to-noise ratios providing advantages over conventional DFT estimators and Minimum-Mean-Square-Error (MMSE) estimators. The use of MMSE may be incorporated for estimation at the band edges providing further improvements. The received signal in the time domain, is transformed to the frequency-domain received signal via an N-point FFT (501). The frequency-domain received signal at the pilot locations is then used to obtain a “noisy” channel estimates at the pilot subcarriers by dividing the known pilot symbols in (503). Uniformly spaced pilots are assumed over a window of usable subcarriers. Padding is applied to the initial channel estimates at both sides of the band to account for unused subcarriers such as guard spacing. Weighting factors during power estimation may take into account various noise characteristics such as combinations of known and time-limited noise power characteristics.

Claims

1. A method of operating a receiver, said method comprising:

transforming a received signal into a frequency domain signal, said received signal having a number of subcarrier channel locations;

determining a time-domain vector from a set of pilot subcarriers of said received signal, wherein determining a time-domain vector, further comprises;

determining a spectral window size, said spectral window size based upon said number of subcarriers;

computing a first channel estimate vector from said set of pilot subcarriers, said pilot subcarriers having a given spacing in the frequency domain;

inserting a first set of padding values, said first set of padding values corresponding to a set of subcarriers at both edges of said window, into said first channel estimate vector;

determining a time-domain vector from said first channel estimate vector having said padding values; and

weighting said time-domain vector according to a set of weighting factors to obtain a weighted time-domain vector;

providing compensation for time domain aliasing in said time-domain vector, wherein providing compensation for time domain aliasing further comprises;

inserting a set of padding zeros into said weighted time-domain vector, starting from a position based on a maximum channel spread and channel characteristics; and

obtaining a second time-domain vector;

transforming said second time-domain vector into a frequency-domain channel estimate vector;

mapping said frequency-domain channel estimate vector to said subcarriers of said spectral window; and

providing compensation for irregularity in said spacing of said pilot subcarriers; and

obtaining a channel estimate vector.

2. The method of claim 1, wherein providing compensation for irregularity in said spacing of said pilot subcarriers, further comprises:

inserting a value into said frequency-domain channel estimate vector at an additional subcarrier location, if a subcarrier spacing is larger than an average spacing of said pilot subcarriers;

or

deleting a value from the frequency-domain channel estimate vector, for a subcarrier spacing being smaller than said average spacing of said pilot subcarriers.

3. The method of claim 1, wherein transforming a received signal into a frequency domain signal, said received signal having a number of subcarrier channel locations, further comprises:

performing an n-point Fast Fourier Transform (FFT) on said received signal.

4. The method of claim 1, further comprising:

computing Minimum Mean Square Error (MMSE) frequency-domain channel estimates for a predetermined number of subcarriers at both edges of said spectral window; and

replacing channel estimates from said frequency-domain channel estimation vector with said MMSE channel estimates.

5. The method of claim 1, wherein determining a spectral window size, said spectral window size based upon said number of subcarriers having a given frequency domain spacing, further comprises:

determining a Fast Fourier Transform (FFT) size and an Inverse Fast Fourier Transform (IFFT) size based upon a number of Orthogonal Frequency Division Multiplexed (OFDM) pilots having a given pilot spacing wherein the FFT size is determine based upon the number of said pilots times said pilot spacing and said IFFT size is determined such that the number of said pilots is less than said number of subcarriers.

6. A method of operating a receiver, said method comprising:

transforming a received signal into a frequency domain signal, said received signal having a number of subcarrier channel locations;

determining a time-domain vector from a set of pilot subcarriers of said received signal, wherein determining a time-domain vector, further comprises;

determining a spectral window size, said spectral window size based upon said number of subcarriers;

computing a first channel estimate vector from said set of pilot subcarriers, said pilot subcarriers having a given spacing in the frequency domain;

inserting a first set of padding values, said first set of padding values corresponding to a set of subcarriers at both edges of said window, into said first channel estimate vector;

determining a time-domain vector from said first channel estimate vector having said padding values; and

weighting said time-domain vector according to a set of weighting factors to obtain a weighted time-domain vector, wherein weighting said time-domain vector according to a set of weighting factors, further comprises;

performing power estimation on each element of said time-domain vector; and

computing said weighting factors based upon a noise and interference characteristic and said power estimation;

providing compensation for time domain aliasing in said time-domain vector; and

obtaining a channel estimate vector.

7. The method of claim 6, wherein computing said weighting factors based upon a noise and interference characteristic, further comprises:

computing said weighting factors based upon additive Gaussian white noise.

8. The method of claim 6, wherein computing said weighting factors based upon a noise and interference characteristic, further comprises:

computing said weighting factors based upon said noise being frequency selective and having a known noise power.

9. The method of claim 6, wherein computing said weighting factors based upon a noise and interference characteristic, further comprises:

computing said weighting factors based upon time-limited interference and noise power;

identifying a window within said time domain vector, where a largest portion of said interference and noise power is concentrated; and

setting weighting factors to zero for each elements of said time domain vector corresponding to said window, that is below a threshold operating signal-to-interference and noise ratio (SINR).

10. The method of claim 6, wherein computing said weighting factors based upon a noise and interference characteristic, further comprises:

computing said weighting factors based upon combining a component for a known noise and interference characteristic with a time-limited component.

11. The method of claim 10, wherein computing said weighting factors based upon combining a component for a known noise characteristic with a time-limited component further comprises:

determining said component for said known noise and interference characteristic;

determining said time-limited component; and

computing an overall weighting factor as a product of said component for said known noise and interference characteristic and said time-limited component.

12. A receiver unit comprising:

a receiver component;

a channel estimation component coupled to said receiver component;

a noise determination and weighting component coupled to said channel estimation component;

a window adjustment component coupled to said channel estimation component, said window adjustment component being configured to determine a spectral window size, said spectral window size based upon said number of subcarriers, said subcarrier further comprising a set of pilots having a given frequency domain spacing;

an aliasing compensation component, coupled to said channel estimation component, said aliasing compensation component being configured to insert a set of padding zeros into a weighted estimation vector, starting from a position based on a maximum channel spread and a channel characteristic;

a pilot spacing compensator component, coupled to said channel estimation component and to said window adjustment component, said pilot spacing compensator configured to insert a set of insertions/omissions corresponding to a set of irregularities in pilot spacing; and

a channel decoding component coupled to said channel estimation component.