公开(公告)号：US11038719B2

公开(公告)日：2021-06-15

申请号：US16399835

申请日：2019-04-30

Applicant: QUALCOMM Incorporated

Inventor： Parisa Babaheidarian ,  Vijayvaradharaj Tirucherai Muralidharan ,  Murali Menon

IPC: H04L25/02 ,  H04L5/00 ,  H04L5/16 ,  H04L7/04

Abstract: Channel estimation performance depends on the amount of averaging performed by a channel impulse response coherent filter. For half-duplex UEs, which use a single phase locked loop (PLL) for both downlink transmissions and uplink transmissions, averaging may not be performed across downlink subframes before and after uplink subframes if the PLL's phase changes and locks to a random initial value when switching transmission directions. Techniques disclosed herein facilitate estimating the PLL's random initial phase and enable correcting the phase of symbols accordingly. By correcting the phase of the symbols, it is possible to average across symbols before and after a frequency re-tune and/or a transmission direction switch based on the phase correction. This may serve to improve the accuracy of channel estimation. Further techniques disclosed herein may improve the accuracy of Doppler estimations by enabling the inclusion of symbols before and after a frequency re-tuning when performing the Doppler estimation.

公开(公告)号：US12165029B2

公开(公告)日：2024-12-10

申请号：US17119912

申请日：2020-12-11

Applicant: QUALCOMM Incorporated

Inventor： Parisa Babaheidarian

IPC: G06N3/04 ,  G06F17/16 ,  G06N3/08

Abstract: A method performs eigen decomposition with an artificial deep neural network. The deep neural network receives an input covariance matrix. The deep neural network has a number of convolutional layers and also a number of pooling layers. The deep neural network predicts dominant eigen information of the input covariance matrix, after applying the convolutional layers and the pooling layers to the input covariance matrix. The input covariance matrix may be a real-valued covariance matrix or a complex-valued covariance matrix having a concatenated pair of matrices, including a first matrix of real components and a second matrix of imaginary components. The dominant eigen information may be absolute values of a pair of dominant eigen values and sign information of the pair of dominant eigen values, and/or absolute values of a pair of dominant eigen vectors and sign information of the pair of dominant eigen vectors.