Hypernetwork Kalman filter for channel estimation and tracking
AssigneeQUALCOMM Incorporated
InventorsKumar Pratik, Arash Behboodi, Joseph Binamira Soriaga, Max Welling
Life of this patent
Estimating the term…
Abstract
A processor-implemented method is presented. The method includes receiving an input sequence comprising a group of channel dynamics observations for a wireless communication channel. Each channel dynamics observation may correspond to a timing of a group of timings. The method also includes determining, via a recurrent neural network (RNN), a residual at each of the group of timings based on the group of channel dynamics observations. The method further includes updating Kalman filter (KF) parameters based on the residual and estimating, via the KF, a channel state based on the updated KF parameters.
Claims28
A processor-implemented method by a communication device, comprising:
- receiving an input sequence comprising a plurality of channel dynamics observations for a wireless communication channel, each channel dynamics observation corresponding to a timing of a plurality of timings;
- determining, via a recurrent neural network (RNN), a residual at each of the plurality of timings based on the plurality of channel dynamics observations;
- updating Kalman filter (KF) parameters based on the residual;
- and
- estimating, via the KF, a channel state based on the updated KF parameters.
An apparatus at a communication device, comprising:
- a processor;
- and
- a memory coupled with the processor and storing instructions operable, when executed by the processor, to cause the apparatus:
- to receive an input sequence comprising a plurality of channel dynamics observations for a wireless communication channel, each channel dynamics observation corresponding to a timing of a plurality of timings;
- to determine, via a recurrent neural network (RNN), a residual at each of the plurality of timings based on the plurality of channel dynamics observations;
- to update Kalman filter (KF) parameters based on the residual;
- and
- to estimate, via the KF, a channel state based on the updated KF parameters.
A non-transitory computer-readable medium having program code recorded thereon at a communication device, the program code executed by a processor and comprising:
- program code to receive an input sequence comprising a plurality of channel dynamics observations for a wireless communication channel, each channel dynamics observation corresponding to a timing of a plurality of timings;
- program code to determine, via a recurrent neural network (RNN), a residual at each of the plurality of timings based on the plurality of channel dynamics observations;
- program code to update Kalman filter (KF) parameters based on the residual;
- and
- program code to estimate, via the KF, a channel state based on the updated KF parameters.
An apparatus at a communication device, comprising:
- means for receiving an input sequence comprising a plurality of channel dynamics observations for a wireless communication channel, each channel dynamics observation corresponding to a timing of a plurality of timings;
- means for determining, via a recurrent neural network (RNN), a residual at each of the plurality of timings based on the plurality of channel dynamics observations;
- means for updating Kalman filter (KF) parameters based on the residual;
- and
- means for estimating, via the KF, a channel state based on the updated KF parameters.
Dependencies are inferred from the claim text.