Maximum likelihood sequence estimation apparatus
AssigneeNEC Corporation
InventorTakayuki Shibata
Life of this patent
Estimating the term…
Abstract
A maximum likelihood sequence estimation (MLSE) apparatus which can estimate a maximum likelihood sequence (MLS) correctly, even when an optimum range of an impulse response is changed, includes a range setting device for selecting an optimum range in the impulses detected by an impulse response detector. The MLSE apparatus estimates the MLS by using the impulses in the optimum range as the tap coefficients which are renewed by adaptive estimation in an adaptive control device. The range setting device renews the optimum range, every time the tap coefficients are renewed. The adaptive control device sets up the impulses in the optimum range as the tap coefficients in MLSE means. The MLSE apparatus estimates MLS on the basis of the impulses in the optimum range.
Claims6
A maximum likelihood sequence estimation (MLSE) apparatus for estimating a maximum likelihood sequence, which comprises:
- a first and second memory for storing received signal;
- an impulse response detection means for detecting impulses in training signal included in said received signal;
- a range setting means for selecting an optimum range among said impulses and for renewing a timing signal (TS) on the basis of said optimum range and for outputting the renewed timing signal (TSRN);
- an adaptive control means for controlling said first and second memory and for outputting “j” tap coefficients;
- and
- a MLSE means for estimating said maximum likelihood sequence on the basis of said “j” tap coefficients,
- wherein:
- said first memory outputs, to said MLSE means, said received signal by using a first timing which is designated by said TSRN;
- said second memory outputs, to said adaptive control means, said received signal by using a second timing which is defined by (said first timing—t), namely said first timing subtracted by “t”, where “t” is equal to (j+u), where “j” is a tap length of said MLSE means and “u” is a length of a forward range in said adaptive control means;
- said adaptive control means renews said impulses on the basis of said received signal outputted from said second memory, said maximum likelihood sequence outputted from said MLSE means, and said impulses in said optimum range outputted from said range setting means.
A maximum likelihood sequence estimation method for estimating a maximum likelihood sequence on the basis of adaptively selected optimum range among impulses in a received training signal, which comprises the steps of:
- for selecting said optimum range at time “n” consisting of“j” impulses wp, wp+1, . . . , wp+j−1 which maximize (wp2+wp+12 . . . wp+j−12), among “N”impulses w0, w1, . . . , wN−1, where said “j” impulses are obtained by correlation of “N” pseudo noise codes for training signal and said received training signal, where the indices subscripts of“w” are replaced by said indices of“w” subtracted by “N”, when said indices of“w” is greater than or equal to “N”;
- estimating said maximum likelihood sequence at time “n”, by convoluting “j” candidate symbols with said “j” impulses;
- for selecting impulses wp−u, wp−u+1, . . . , wp−u+N−1 at time “(n+1)” on the basis of said “j” impulses wp, wp+1, . . . , wp+j−1 in said optimum range, where “u” is a length of a forward range which consists of impulses wp+j, wp+j+1, . . . , wN−1;
- selecting said optimum range at time “(n+1)” consisting of “j” impulses wq, wq+1, . . . , wq+j−1 which maximize (wq2+wq+12+ . . . +wq+j−12), among “N” impulses W0, w1, . . . , wN−1, where said “j” impulses are obtained by correlation of “N” pseudo noise codes for training signal and said received training signal, where the indices subscripts of“w” are replaced by said indices of “w” subtracted by “N”, when said indices of“w” is greater than or equal to “N”;
- and
- for estimating said maximum likelihood sequence at time “(n+1)”, by convoluting “j” candidate symbols with said “j” impulses.
Dependencies are inferred from the claim text.