Back to results
US 2003/0189992 A1Published application
Method and system for channel estimation using iterative estimation and detection
AssigneeNokia Corporation
InventorNikolai Nefedov
Life of this patent
Estimating the term…
Abstract
The invention provides a method and system for a modified Expectation Maximization (EM) algorithm that enables a detector to perform relatively accurate channel estimates on blocks of data that include relatively short training sequences in wireless communication signals. Channel re-estimation is simplified by transferring the estimation calculations from the time domain into a channel state domain. Semi-blind detection methods are employed to combine information obtained from known training sequence and unknown data to improve channel estimates. The modified EM algorithm's pre-calculated low-dimensional matrix inversion reduces a detector's demands on processing resources.
Claims16
A method for joint channel estimation and detection for each transmitted block in at least one wireless signal, comprising:
- (a) receiving each block included in at least one wireless signal, wherein each block includes a training sequence and data;
- (b) employing the training sequence in a received block to generate an initial estimate for each channel tap;
- (c) employing each initial estimate of each channel tap to determine at least one channel state estimate and employing each determined channel state estimate to initialize an iterative process;
- (d) employing the iterative process to determine an update for each channel state estimate and employing each updated channel state estimate to generate an update for each channel tap estimate, wherein generating each updated channel tap estimate is obtained via a reduced observation vector to be generated that includes updated estimates that are generated for all channel states;
- and
- (e) employing the iterative process to generate a result based on a channel model inverse in a channel state domain, wherein the result of the iterative process is employed to detect each symbol in each received block, including:
- (i) recursively performing the actions of sub-paragraphs (c), (d) and (e) until the result of the iterative process is less than a predetermined threshold;
- and
- (ii) when the result of the iterative process is less than the predetermined threshold, employing the result of the iterative process to detect each symbol in the received block for a channel.
6 more elements
A method for iterative channel estimation for each transmitted block in at least one wireless signal, comprising:
- (a) receiving each block included in at least one wireless signal, wherein each block includes a training sequence and data;
- (b) employing the training sequence in a received block to generate an initial estimate for each channel tap;
- (c) employing each initial estimate of each channel tap to determine at least one channel state estimate and employing each determined channel state estimate to initialize an iterative process;
- (d) employing the iterative process to determine an update for each channel state estimate and employing each updated channel state estimate to generate an update for each channel tap estimate, wherein generating each updated channel tap estimate is obtained via a reduced observation vector to be generated that includes updated estimates that are generated for all channel states;
- and
- (e) employing the iterative process to generate a result based on a channel model inverse in a channel state domain, wherein the result of the iterative process is employed to detect each symbol in each received block, including:
- (i) recursively performing the actions of sub-paragraphs (c), (d) and (e) until the result of the iterative process is less than a predetermined threshold;
- and
- (ii) when the result of the iterative process is less than the predetermined threshold, employing the result of the iterative process to detect each symbol in the received block for a channel.
6 more elements
A system for joint channel estimation and detection for each block in at least one received wireless signal, comprising:
- (a) an antennae module for receiving at least one wireless signal;
- and
- (b) a detector module for performing actions related to joint channel estimation and detection of at least one received wireless signal, including:
- (i) receiving each block included in at least one wireless signal, wherein each block includes a training sequence and data;
- (ii) employing the training sequence in a received block to generate an initial estimate for each channel tap;
- (iii) employing each initial estimate of each channel tap to determine at least one channel state estimate and employing each determined channel state estimate to initialize an iterative process;
- (iv) employing the iterative process to determine an update for each channel state estimate and employing each updated channel state to generate an update for each channel tap estimate, wherein generating each updated channel tap estimate is obtained via a reduced observation vector to be generated that includes updated estimates that are generated for all channel states;
- and
- (v) employing the iterative process to generate a result based on a channel model inverse in a channel state domain, wherein the result of the iterative process is employed to detect each symbol in each received block for a channel.
6 more elements
A method for joint channel estimation and detection for each transmitted block in at least one wireless signal, comprising:
- (a) means for receiving each block included in at least one wireless signal, wherein each block includes a training sequence and data;
- (b) means for employing the training sequence in a received block to generate an initial estimate for each channel tap;
- (c) means for employing each initial estimate of each channel tap to determine at least one channel state estimate and employing each determined channel state estimate to initialize an iterative process;
- (d) means for employing the iterative process to determine an update for each channel state estimate and employing each updated channel state to generate an update for each channel tap estimate, wherein generating each updated channel tap estimate is obtained via a reduced observation vector to be generated that includes updated estimates that are generated for all channel states;
- and
- (e) employing the iterative process to generate a result based on a channel model inverse in a channel state domain, wherein the result of the EM process is employed to detect each symbol in each received block for a channel.
3 more elements
Dependencies are inferred from the claim text.