Kalman filter for channel estimation in OFDM systems

1. A receiver, comprising:

an OFDM demodulator;

a channel corrector connected downstream of the OFDM demodulator; and

a channel estimator connected downstream in the OFDM demodulator and to the channel corrector to generate output to the channel corrector;

wherein the channel estimator comprises a scalar Kalman filter that determines a Least-Square estimated value, H s , at pilot subcarrier locations, s, according to a filter equation

H s k =K gain S n [k],

wherein H s , is received by the filter equation as an input,

wherein a value of H s k is processed for each k th variable,

wherein the filter equation processes a correction value, S n [k], according to a first equation

S n [k]=S+K n ( H s −S ),

wherein the equation filter processes a prediction for the correction value, S, according to a second equation

S=K a S n [k],

wherein filter equation processes a gain, K n , according to a third equation

K n =P/ (1+ P ),

wherein the filter equation processes a minimum MSE according to a fourth equation

P n [k]=P (1− K n ),

wherein the filter equation processes a prediction for a minimum MSE according to a fifth equation

P=K a 2 P n [k]+K b ,and

wherein K a , K b and K gain are constants.

2. A receiver according to claim 1 , wherein K n is a constant.

3. A receiver according to claim 1 , wherein the constants K n , K a , K gain and K b are selected as a function of the modulation mode of the subcarrier.

4. A receiver according to claim 1 , wherein the channel estimator processes scattered pilots whose locations s repeats its pattern every r symbols; and the channel estimator includes r*N s Kalman filters.

5. A receiver according to claim 1 , wherein the filter equation is performed in software.

6. A scalar filter process executed by a receiver, comprising:

demodulating an input signal by an OFDM demodulator to produce a demodulated signal;

producing a channel corrected signal from the demodulated signal by a channel corrector connected downstream of the OFDM demodulator;

generating an output from a channel estimator connected downstream of the OFDM demodulator and connected to the channel corrector; and

supplying the output to the channel corrector;

wherein the channel estimator incorporates a computer readable medium encoded with a computer program for

processing a scalar Kalman filter that determines a Least-Square estimated value, H s , at pilot subcarrier locations, s, according to a filter equation

H s k =K gain S n [k],

receiving H s by the filter equation as an input,

processing a value of H s k for each k th variable,

processing a correction value, S n [k], according to a first equation

S n [k]=S+K n ( H s −S ),

processing a prediction for the correction value, S, according to a second equation

S=K a S n [k],

processing a gain, K n , according to a third equation

K n =P/ (1+ P ),

processing a minimum MSE according to a fourth equation

P n [k]=P (1 −K n ), and

processing a prediction for a minimum MSE according to a fifth equation

P=K a 2 P n [k]+K b ,

wherein K a , K b and K gain are constants.

7. A scalar filter process according to claim 6 , wherein K n is a constant.

8. A scalar filter process according to claim 6 , wherein the constants K n , K a , K gain and K b are selected as a function of the modulation mode of the subcarrier.

9. A scalar filter process according to claim 6 , wherein the channel estimator processes scattered pilots whose locations s repeats its pattern every r symbols; and

wherein the channel estimator includes r*N s Kalman filters.

10. A scalar filter process according to claim 6 , wherein the filter equation is performed in software.