Apparatus and method for carrier frequency offset estimation and correction in a wireless communication system
View Patent ↗A method is disclosed for correcting carrier frequency offset (CFO) in a received data packet that includes one or more M data streams. The data packet has a payload portion preceded by a preamble portion, the payload portion having a plurality of data symbols (L) each of which include a plurality of data tones (K). The method comprises, making a per-stream preamble CFO estimate for each data stream by correlating repeated preamble portions that precede the data packet. The per-stream preamble CFO estimates are weighted and averaged to obtain a single preamble CFO estimate. Any CFO in each received data stream is then corrected according to the preamble CFO estimate. Additionally, for an l th data symbol in a data packet, a per-tone CFO estimate is derived and successively averaged over frequency (tones), time (symbols) and space (data steams) to derive data CFO estimate p data . Any CFO in each received data stream may be corrected according to the data CFO estimate, and/or the preamble CFO estimate.
1. A method for correcting carrier frequency offset (CFO) in a communication system receiving a plurality of data packets, each data packet including multiple data streams (M), each data packet having a payload portion including a number (L) of data symbols, each data symbol having a number (K) of tones, the method comprising:
for an l th symbol in a data packet:
(a) generating a per-tone CFO estimate (p tone (k)) for each of the K data tones in a data symbol based on (i) a received signal; (ii) a channel estimate; and (iii) a data estimate or pilot information;
(b) assigning a weight to each p tone (k);
(c) generating a per-symbol CFO estimate (p sym (l)) for the l th symbol by averaging the K weighted per-tone CFO estimates;
(d) assigning a weight to each p sym (l) for the L symbols in the data packet;
(e) generating a per-stream CEO estimate (p stream (m)), by averaging the L weighted per-symbol CEO estimates;
(f) assigning a weight to each p stream (m);
(g) generating a single data CEO estimate (p data ) by averaging the M weighted per-stream CEO estimates;
(h) generating a packet CEO estimate (p packet ) using p data and/or a preamble CEO estimate (p preamble );
(i) correcting the CEO in each data stream according to p packet ;
(j) repeating steps (a) through (i) for each data symbol in the data packet.
2. The method of claim 1 , wherein the K tones in each data symbol include data tones for carrying data, and pilot tones having a known data constellation, and wherein step (b) further comprises selectively assigning a first weight to a pilot tone estimate and a second weight to a data tone estimate.
3. The method of claim 2 , wherein said first pilot tone weight is higher than said second data tone weight.
4. The method of claim 1 , wherein step (d) further comprises selectively assigning a higher weight to a more recently calculated p sym (l).
5. The method of claim 1 , wherein step (f) further comprises selectively assigning weights to p stream (m) based on signal quality, wherein higher weights are assigned to the p stream (m) associated with a higher quality signal stream.
6. The method of claim 1 , wherein the data packet undergoes a fast Fourier transform (FFT), and wherein the estimate of step (a) is made after said FFT while the correction of step (i) is made prior to said FFT.
7. The method of claim 1 , wherein each data packet is preceded by a preamble having repeated preamble portions, and wherein the repeated preamble portions are correlated to determine p preamble .