US 2014/0233684 A1 · Sparse Channel Detection, Estimation, And Feedback

Abstract

Most wireless channels are sparse, so sparse channel-based methods can be used for channel estimation and feedback with much better estimation accuracy and much lower feedback overhead. However, certain wireless channels can be non-sparse, for which sparse channel-based methods may cause degraded estimation quality and increased feedback overhead. Means of detecting the channel sparseness are described that provide simple and effective channel sparseness indicators and safeguard against the mismatch between non-sparse channels and sparse channel-based methods. Various fallback options can be used under non-sparse channels such that estimation degradation and feedback overhead are both minimized. Fake multipath removal in continuous time-domain parameter extraction, a sparse channel-based method, is also described that further improves estimation quality and reduces feedback overhead.

Claims20

Dependent claims

A method for channel estimation, comprising:

(a) computing a detection variable or a plurality of detection variables from a received signal,
(b) detecting the sparseness of a channel from which said received signal is received based on said detection variable or said plurality of said detection variables, and
(c) estimating said channel with a sparse channel-based method if said channel is detected to be a sparse channel, or estimating said channel with a fallback option if said channel is detected to be a non-sparse channel.

A method for extracting time-domain parameters of a channel, comprising:

(a) generating a set of channel observations from a pilot signal in a received signal,
(b) forming a matching error between said channel observations and a time-domain channel model, said time-domain channel model being defined by a set of time-domain parameters, said time-domain parameters comprising the number of multipaths, the complex amplitudes and the propagation delays of said multipaths.
(c) detecting one or more said multipaths, and determining said complex amplitude(s) and said propagation delay(s) of the detected multipath(s) by minimizing said matching error recursively,
(d) performing fake multipath detection on each said detected multipath, and removing said detected multipath if said detected multipath is detected to be a fake multipath,
(e) repeating steps (c) and (d) until said fake multipath is detected and removed, or until a predetermined number of recursions has been reached, or until a predetermined number of multipaths have been detected, and
(f) refining said complex amplitudes of all said detected multipaths that are not fake multipaths by minimizing said matching error given said propagation delays of said detected multipaths,
whereby the number of said detected multipaths, the refined complex amplitudes of said detected multipaths, and said propagation delays of said detected multipaths are extracted time-domain parameters.

4 more elements

Dependencies are inferred from the claim text.

Sparse Channel Detection, Estimation, And Feedback

Life of this patent

Sparse Channel Detection, Estimation, And Feedback

Life of this patent

Abstract

Claims20