Nonlinear inversion
View Patent ↗A system and method are disclosed for processing a signal propagated through a nonlinear channel. The method includes modeling the channel characteristics to produce a linearized channel model, deriving an inverse linearized channel model from the linearized channel model, and filtering the signal using the inverse linearized channel model.
1. A method for processing a signal propagated through a nonlinear channel, including:
receiving the signal;
adaptively determining an inverse linearized channel model, wherein the inverse linearized channel model has a linearized transfer function that includes a plurality of linear operators and nonlinear operators, wherein the nonlinear operators include minimum and maximum operators; and
filtering the signal using the inverse linearized channel model.
2. A method for processing a signal propagated through a nonlinear channel as recited in claim 1 , wherein adaptively determining an inverse linearized channel model includes using a least mean squared algorithm.
3. A method for processing a signal propagated through a nonlinear channel as recited in claim 1 , wherein adaptively determining an inverse linearized channel model includes determining coefficients of the plurality of linear operators.
4. A method for processing a signal propagated through a nonlinear channel as recited in claim 1 , wherein the plurality of linear operators are each first order polynomials.
5. A method for processing a signal propagated through a nonlinear channel as recited in claim 1 , wherein adaptively determining an inverse linearized channel model includes determining coefficients of the plurality of nonlinear operators.
6. A receiver for processing a signal from a transmitter, wherein the signal propagated through a nonlinear channel, comprising:
a nonlinear filter implementing an inverse linearized channel model that is used to filter a received input signal, wherein the inverse linearized channel model is adaptively determined, and wherein the inverse linearized channel model has a linearized transfer function that includes a plurality of linear operators and nonlinear operators, wherein the nonlinear operators include minimum and maximum operators.
7. A receiver for processing a signal propagated through a nonlinear channel as recited in claim 6 , wherein adaptively determining an inverse linearized channel model includes using a least mean squared algorithm.
8. A receiver for processing a signal propagated through a nonlinear channel as recited in claim 6 , wherein adaptively determining an inverse linearized channel model includes determining coefficients of the plurality of linear operators.
9. A receiver for processing a signal propagated through a nonlinear channel as recited in claim 6 , wherein the plurality of linear operators are each first order polynomials.
10. A receiver for processing a signal propagated through a nonlinear channel as recited in claim 6 , wherein adaptively determining an inverse linearized channel model includes determining coefficients of the plurality of nonlinear operators.