Receiver, transceiver system and associated receiving method
The receiver includes a sampler designed to provide one or more samples (y n ) per received symbol, the symbols belonging to a predefined alphabet; an equalizer designed to compute, for each received symbol, an estimate (z n ) of this symbol based on a linear combination (y′ n ) of the samples (y n ) for this symbol; and a decision module designed to determine the symbol of the alphabet closest to the estimate (z n ) as detected symbol. The alphabet exhibits a decentering such that the transmitted symbols have a non-zero predefined expectation, and the equalizer is designed to add a non-zero scalar component (θ) to the linear combination (y′ n ) in order to compensate at least partially for the decentering.
1 . A digital data receiver comprising:
a sampler configured to sample a received signal (r(t)) that has propagated in a propagation channel, N symbols being encoded in the received signal (r(t)), N being an integer, in order to provide a set of blocks of samples for the N symbols, the N symbols being selected from a predefined alphabet (Ω) having K symbols, K being an integer, the sampler providing N blocks of M samples, M being at least equal to 1;
an equalizer configured to compute, for each symbol (s n ) among the N symbols, an estimate (z n ) of said symbol (s n ) based on the samples provided by the sampler, the equalizer comprising a sample combiner configured to receive said samples provided by the sampler, and to provide a linear combination of said samples, said estimate (z n ) of said symbol (s n ) being determined from said linear combination; and
a decision module configured to determine detected symbols, a detected symbol being determined for each estimate (z n ) computed for a symbol (s n ), a detected symbol (ŝ n ) being a symbol selected from said K symbols in the predefined alphabet (Ω) that is the closest to the estimate (z n ) computed for each symbol (s n );
wherein the predefined alphabet (Ω) exhibits a decentering such that each of said N symbols have a non-zero predefined expectation ( [s n ]), and the equalizer is configured to add a non-zero scalar component (θ) to said linear combination in order to compensate at least partially for the decentering;
the equalizer being configured to iteratively update the non-zero scalar component (θ) from the detected symbols previously determined by the decision module.
2 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute the non-zero scalar component (θ) from coefficients of the linear combination of samples and the predefined expectation ( [s n ]).
3 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute the non-zero scalar component (θ) based on a channel matrix (H) representative of the propagation channel.
4 . The digital data receiver as claimed in claim 3 , wherein said channel matrix (H) is a block Toeplitz matrix.
5 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute said non-zero scalar component (θ) so as to minimize a mean squared error between the symbols (s n ) and the estimates (z n ).
6 . The digital data receiver as claimed in claim 1 , wherein the equalizer comprises a feedforward filter and the equalizer is designed to apply the feedforward filter to the samples in order to provide said linear combination.
7 . The digital data receiver as claimed in claim 6 , wherein the equalizer is configured to compute the non-zero-scalar component (θ) based on the feedforward filter.
8 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute, for each symbol (s n ), the estimate (z n ) of said symbol (s n ) independently of previously detected symbols.
9 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute, for each symbol (s n ), the estimate (z n ) of said symbol (s n ) based on a difference (d) between the linear combination of the samples for said symbol (s n ) and an other linear combination of previously detected symbols (ŝ n−1 , . . . , ŝ n−Δ ), the non-zero scalar component (θ) being added to said difference (d).
10 . The digital data receiver as claimed in claim 9 , wherein the equalizer further comprises a backward filter, and the equalizer is configured to apply said backward filter to the previously detected symbols (ŝ n−1 , . . . , ŝ n−Δ ) so as to provide said other linear combination of previously detected symbols (ŝ n−1 , . . . , ŝ n−Δ ).
11 . The digital data receiver as claimed in claim 10 , wherein the equalizer is configured to compute the non-zero scalar component (θ) based on the backward filter.
12 . The digital data receiver as claimed in claim 1 , wherein the equalizer is configured to compute the non-zero scalar component (θ) using the following recurrence relation:
Θ
n
+
1
=
Θ
n
-
μ
2
e
n
where e n denotes the error signal defined as the difference between the estimate z n and a symbol s n−Δ transmitted with a delay Δ for a symbol period T n ; and μ 2 denotes an adaptation step, for the adaptation of the non-zero scalar component θ.
13 . A digital communication system, comprising a transmitter configured to transmit symbols selected from the predefined alphabet (Ω) with the non-zero expectation ( [s n ]), and the digital data receiver as claimed in claim 1 .
14 . A method for receiving digital data, comprising:
a step of sampling, by a sampler, a received signal (r(t)) that has propagated in a propagation channel, N symbols being encoded in this received signal (r(t)) N being an integer, in order to provide a set of blocks of samples for the N symbols, the N symbols being selected from a predefined alphabet (Ω) having K symbols, K being an integer, the sampling step providing N blocks of M samples, M being at least equal to 1;
a step of performing a linear equalization, by an equalizer, comprising computing, for each symbol (s n ), an estimate (z n ) of said symbol (s n ) from the samples provided by the sampling step, the linear equalization step comprising performing a linear combination of the samples provided by the sampling step, said estimate (z n ) of said symbol being determined from the linear combination; and
a step of determining detected symbols, a detected symbol (ŝ n ) being determined for each estimate (z n ) computed for a symbol (s n ), a detected symbol (ŝ n ) being a symbol selected from said K symbols in the predefined alphabet (Ω) that is the closest to the estimate (z n );
wherein the predefined alphabet (Ω) exhibits a decentering such that each of the N symbols (s n ) have a non-zero predefined expectation ( [s n ]), and the linear equalization comprises adding a non-zero scalar component (θ) to the linear combination in order to compensate at least partially for the decentering, the non-zero scalar component (θ) being iteratively updated from the previously determined detected symbols.
15 . A non-transitory computer readable storage medium having stored thereon instructions that, when executed by a computer, implement the method as claimed in claim 14 .