Systems, apparatus, and methods for channel estimation and data detection
View Patent ↗Systems, apparatus, and methods for channel estimation and data detection. In one exemplary embodiment, the data is obtained in a two-phase transmission structure that alternates known data with unknown data according to a regular or otherwise pre-determined time interval. Then, the receiver iteratively updates a postulated channel, and provides a predicted channel to the next time slot. Conceptually, the exemplary techniques iteratively improve its postulates for channel condition and data over multiple time slots. More directly, instead of linear detection and decoding of a pilot for channel estimation in each time slot, the exemplary techniques described herein use postulated channel conditions to attempt data detection and use the recovered data from data detection (unknown data) to re-postulate the channel conditions, etc. until channel conditions are stable and/or the next time slot is ready for processing.
1 . A method, comprising:
obtaining a first known data and a first unknown data from a first time slot;
postulating a first channel condition;
recovering a first recovered data from the first unknown data based on the first channel condition;
estimating a second channel condition based on the first recovered data;
obtaining a second unknown data from a second time slot; and
recovering a second recovered data from the second unknown data based on the second channel condition.
2 . The method of claim 1 , further comprising:
calculating a noise variance for the second channel condition based on a first mean-field variational Bayes approximation of a first probability distribution of the first unknown data and the first recovered data; and
updating the noise variance for a third channel condition based on a second mean-field variational Bayes approximation of a second probability distribution of the second unknown data and the second recovered data.
3 . The method of claim 1 , further comprising:
calculating a noise covariance matrix for the second channel condition based on a first variational Bayes approximation of a first probability distribution of the first unknown data and the first recovered data; and
updating the noise covariance matrix for a third channel condition based on a second variational Bayes approximation of a second probability distribution of the second unknown data and the second recovered data.
4 . The method of claim 1 , further comprising:
estimating a third channel condition based on the second recovered data; and
determining whether the second channel condition and the third channel condition converge.
5 . The method of claim 1 , where the first known data and the first unknown data are obtained according to a two-phase multiple-input multiple-output time slot; and
where the first known data comprises a pilot signal and the first unknown data comprises first data traffic.
6 . The method of claim 5 , where a second known data and the second unknown data are obtained according to the two-phase multiple-input multiple-output time slot; and
where the second known data comprises the pilot signal and the second unknown data comprises second data traffic.
7 . The method of claim 5 , j the second unknown data is obtained without an other pilot signal.
8 . An apparatus, comprising:
a modem front end;
a processor; and
a non-transitory computer-readable medium comprising instructions which when executed by the processor, causes the processor to:
obtain unknown data via a channel in a first time slot;
postulate a first condition of the channel in the first time slot;
estimate a variational density of the channel or the unknown data;
calculate a fading coefficient based on the variational density; and
predict a second condition of the channel in a second time slot based on the fading coefficient.
9 . The apparatus of claim 8 , where the modem front end comprises a multiple-input multiple-output modem front end and the channel comprises a two-phase structure comprising a known pilot and the unknown data.
10 . The apparatus of claim 8 , where the first condition comprises a noise variance based on a mean-field variational Bayes approximation of the channel or the unknown data.
11 . The apparatus of claim 8 , where the first condition comprises a noise covariance matrix based on a variational Bayes approximation of the channel.
12 . The apparatus of claim 8 , where the fading coefficient comprises a path loss fading coefficient and a time varying fading coefficient.
13 . The apparatus of claim 8 , where the instructions further cause the processor to:
estimate a new variational density of the channel based on the second condition;
calculate a new fading coefficient based on the new variational density; and
predict a third condition of the channel in a third time slot based on the new fading coefficient.
14 . The apparatus of claim 13 , where the instructions further cause the processor to determine whether the variational density and the new variational density converge.
15 . A method for channel estimation and data detection, comprising:
obtaining unknown data via a channel in a first slot;
postulating a first condition of the channel in the first slot;
recovering data from the unknown data and the first condition;
estimating a fading coefficient based on the recovered data; and
predicting a subsequent condition of the channel in a subsequent slot.
16 . The method of claim 15 , where the first slot comprises a known signal and the subsequent slot does not comprise the known signal.
17 . The method of claim 15 , where the first condition comprises a first noise variance postulated from a mean-field variational Bayes approximation of the channel in the first slot, and the subsequent condition comprises a subsequent noise variance postulated from the mean-field variational Bayes approximation of the channel in the subsequent slot.
18 . The method of claim 15 , where the first condition comprises a first noise covariance matrix postulated from a variational Bayes approximation of the channel in the first slot, and the subsequent condition comprises a subsequent noise covariance matrix postulated from the variational Bayes approximation of the channel in the subsequent slot.
19 . The method of claim 15 , where predicting the subsequent condition of the channel occurs in a next slot.
20 . The method of claim 15 , where predicting the subsequent condition of the channel in the subsequent slot iteratively occurs over multiple slots.