Daniel F Liu

US Patent:

20130301697, Nov 14, 2013

Filed:

May 9, 2012

Appl. No.:

13/467394

Inventors:

Daniel N. Liu - Torrance CA, US
Michael P. Fitz - Los Angeles CA, US

International Classification:

H04L 27/01

US Classification:

375233

Abstract:

A blind equalizer apparatus includes a decision-directed (DD) least mean squares (LMS) blind equalizer. A blind equalizer apparatus includes:

US Patent:

20190268026, Aug 29, 2019

Filed:

May 8, 2019

Appl. No.:

16/406291

Inventors:

- Falls Church VA, US
Daniel N. LIU - Torrance CA, US

International Classification:

H03M 13/41
H03M 13/39
H04L 25/03
H03M 13/37

Abstract:

A method includes: generating a trellis; generating one or more predicted symbols using a first non-linear model; computing and saving two or more branch metrics using a priori log-likelihood ratio (LLR) information, a channel observation, and the one or more predicted symbols; if alpha forward recursion has not yet completed, generating alpha forward recursion state metrics using a second non-linear model; if beta backward recursion has not yet completed, generating beta backward recursion state metrics using a third non-linear model; if sigma forward recursion has not yet completed, generating sigma forward recursion state metrics using the branch metrics, the alpha state metrics, and the beta backward recursion state metrics; generating extrinsic information comprising a difference of a posteriori LLR information and the a priori LLR information; computing and feeding back the a priori LLR information; and calculating the a posteriori LLR information.

US Patent:

20170272283, Sep 21, 2017

Filed:

Mar 16, 2016

Appl. No.:

15/071491

Inventors:

- Falls Church VA, US
Daniel N. Liu - Torrance CA, US

International Classification:

H04L 25/03
H03M 13/39
H03M 13/41
H03M 13/27

Abstract:

An apparatus and method for optimizing the performance of satellite communication system receivers by using the Soft-Input Soft-Output (SISO) BCJR (Bahl, Cocke, Jelinek and Raviv) algorithm to detect a transmitted information sequence is disclosed. A Sliding Window technique is used with a plurality of reduced state sequence estimation (RSSE) equalizers to execute the BCJR algorithm in parallel. A serial data stream is converted into a plurality of data blocks using a serial-to-parallel converter. After processing in parallel by the equalizers, the output blocks are converted back to a serial data stream by a parallel-to-serial converter. A path history is determined using maximum likelihood (ML) path history calculation.

US Patent:

20170033809, Feb 2, 2017

Filed:

Jul 27, 2015

Appl. No.:

14/809948

Inventors:

- FALLS CHURCH VA, US
DANIEL N. LIU - TORRANCE CA, US

International Classification:

H04B 1/04

Abstract:

A method for pre-coding a digital signal in a transmitter associated with a digital communications system to remove non-linear distortions. The method converts digital bits to be transmitted to a transmit constellation including symbols defining a plurality of the bits and subtracting a modeled constellation from the transmit constellation to provide a pre-coded constellation. The method performs a modulo operation on the pre-coded constellation to limit the number of symbols that are transmitted at any particular point in time and models the pre-coded constellation after the modulo operation using a Volterra series model, where a plurality of terms in the Volterra series are separately modeled and each modeled term of the constellation is added together to provide the modeled constellation. The method filters the pre-coded constellation to interpolate a symbol sequence in the constellation and shape the constellation and converts the shaped constellation to an analog signal to be transmitted.

US Patent:

20160218752, Jul 28, 2016

Filed:

Jan 26, 2015

Appl. No.:

14/605377

Inventors:

- Falls Church VA, US
Daniel N. LIU - Torrance CA, US

International Classification:

H03M 13/39
H03M 13/37

Abstract:

A method includes: generating a trellis; generating one or more predicted symbols using a first non-linear model; computing and saving two or more branch metrics using a priori log-likelihood ratio (LLR) information, a channel observation, and the one or more predicted symbols; if alpha forward recursion has not yet completed, generating alpha forward recursion state metrics using a second non-linear model; if beta backward recursion has not yet completed, generating beta backward recursion state metrics using a third non-linear model; if sigma forward recursion has not yet completed, generating sigma forward recursion state metrics using the branch metrics, the alpha state metrics, and the beta backward recursion state metrics; generating extrinsic information comprising a difference of a posteriori LLR information and the a priori LLR information; computing and feeding back the a priori LLR information; and calculating the a posteriori LLR information.