System and method for automatic recovery and covariance adjustment in linear filters
View Patent ↗A communications device includes a time/frequency error measurement circuit that receives a communications signal and measures its timing and frequency errors. A Kalman filter receives the communications signal from the time/frequency error measurement circuit and processes the signal using a multi-level state error covariance matrix P for controlling the Kalman gain in the Kalman filter. An on-line monitoring circuit is operative with the Kalman filter for monitoring the actual state errors in time and frequency and controlling the state error covariance matrix P based on a measured error threshold.
1. A communications device, comprising:
a time/frequency error measurement circuit that receives a communications signal and measures its timing and frequency errors within the communications signal;
a Kalman filter that receives the communications signal from the time/frequency error measurement circuit and processes the communications signal using a multi-level state error covariance matrix P for controlling any Kalman gain in the Kalman filter; and
an online error monitoring circuit operative with the Kalman filter and configured to monitor any actual state errors in time and frequency and control the multi-level state error covariance matrix P based on a measured error threshold, wherein levels within the multi-level state error covariance P are defined initially as true initial errors.
2. The communications device according to claim 1 , wherein the state error covariance matrix P is reset from a theoretical P to an actual P to open up bandwidth to an actual state error covariance matrix P when one of at least excessive time and frequency errors occur.
3. The communications device according to claim 1 , wherein said multi-level state error covariance matrix P comprises a value determined from a look-up table comprising a plurality of downlink tracking states having different synchronization pulse types.
4. The communications device according to claim 3 , wherein said online error measuring circuit is operative for receiving and passing residuals at a maximum limit for each of a plurality of track states.
5. The communications device according to claim 3 , wherein said look-up table comprises a plurality of steady state time tracking errors and steady state frequency tracking errors.
6. The communications device according to claim 5 , wherein said multi-level state error covariance matrix P comprises a steady state time or frequency tracking error value determined from said look-up table.
7. The communications device according to claim 1 , wherein said Kalman filter is operative for producing minimum mean square error (MMSE) estimates of timing and frequency errors.
8. The communications device according to claim 1 , and further comprising a time/frequency tracker operative to acquire and track time and frequency variations in synchronization signals conveyed over communications links.
9. The communications device according to claim 8 , and further comprising a receiver clock, wherein said time/frequency tracker is operative to synchronize the receiver clock with a clock signal embedded within the communications signal.
10. The communications device according to claim 9 , and further comprising a front-end demodulator in which said receiver clock is incorporated.
11. A communications system, comprising:
a plurality of signal sources that transmit communications signals operating at respectively different data rates over respective ones of a plurality of communications links;
a receiver that receives the plurality of communications signals, and comprising a
a front end demodulator having a receiver clock that is used in the recovery of data from communications signals received from the signal sources;
a time/frequency error measurement circuit that receives the communications signal and measures any timing and frequency errors with the communications signals;
a Kalman filter that receives communications signals from the time/frequency error measurement circuit and processes the communications signals using a multi-level state error covariance matrix P for controlling any Kalman gain within the Kalman filter;
an online error monitoring circuit operative with the Kalman filter configured to monitor actual state errors in time and frequency and control state error covariance matrix P based on a measured error threshold; and
a time/frequency tracker operative to acquire and track time and frequency variations in synchronization signals conveyed over the communications links and operative to synchronize the receiver clock with a clock signal embedded within the communications signal, wherein levels within the multi-level state error covariance P are defined initially as true initial errors.
12. The communications system according to claim 11 , wherein the state error covariance matrix P is reset from a theoretical P to an actual P to open up bandwidth to an actual state error covariance matrix P when one of at least excessive time and frequency errors occur.
13. The communications system according to claim 11 , wherein said multi-level state error covariance matrix P comprises a value determined from a look-up table comprising a plurality of downlink tracking states having different synchronization pulse types.
14. The communications system according to claim 13 , wherein said online error measuring circuit is operative for receiving and passing residuals at a maximum limit for each of a plurality of track states.
15. The communications system according to claim 13 , wherein said look-up table comprises a plurality of steady state time tracking errors and steady state frequency tracking errors.
16. The communications system according to claim 11 , wherein said Kalman filter is operative for producing minimum mean square error (MMSE) estimates of timing and frequency errors.
17. A method of communicating, comprising:
receiving a communications signal and measuring any timing and frequency errors within a time/frequency error measurement circuit;
processing data from the time/frequency error measurement circuit using a multi-level state error covariance matrix P for controlling any Kalman gain in a Kalman filter;
monitoring actual state errors in time and frequency and controlling the multi-level state error covariance matrix P based on a measured error threshold; and
resetting the state error covariance matrix P from a theoretical P to an actual to open up bandwidth to an actual state error covariance matrix P when one of at least excessive time and frequency errors occur.
18. The method according to claim 17 , which further comprises receiving and passing residuals at a maximum limit for each of a plurality of track states.
19. The method according to claim 17 , which further comprises generating a multi-level state error covariance matrix P from a value determined from a look-up table comprising a plurality of downlink tracking states having different synchronization pulse types.