Space-time receiver structure for digital communication systems
AssigneeNortel Networks Limited
InventorTharmalingam Ratnarajah
Life of this patent
Estimating the term…
Abstract
A diversity hybrid receiver structure is provided for estimating the sequence of transmitted symbols in a digital communication system. A joint optimizer determines optimal effective channel coefficients for use by an equalizer using a power method, and determines optimal filter coefficients for use by a space-time filter. The space-time filter mitigates co-channel interference and produces an intermediate sequence of signal samples. The equalizer acts on the intermediate sequence of signal samples and corrects for intersymbol interference to estimate the transmitted sequence of symbols. The equalizer uses an efficient block decision-feedback sequence estimation method. An interference detection scheme is also provided.
Claims27
A diversity hybrid receiver in a digital communication system, the receiver including:
- a joint optimization processor including instructions for producing a plurality of optimal filter coefficients and a plurality of optimal effective channel coefficients hopt from a plurality of sequences of received training signal samples and a sequence of known training symbols;
- a space-time filter including means for producing an intermediate sequence of signal samples y from a plurality of received sequences of signal samples and the optimal filter coefficients;
- and
- an equalization processor including instructions for producing an estimated sequence of symbols ŝ, the instructions comprising:
- instructions for forming a matrix of optimal effective channel coefficients H from the optimal effective channel coefficients hopt;
- instructions for determining a lower triangular matrix L from a relationship LLH=HHRee−1H+I where LH is the Hermitian of L, HH is the Hermitian of H, I is an identity matrix, and Ree is a covariance matrix of a disturbance e;
- instructions for calculating a vector z=L−1HHRee−1y;
- and
- instructions for determining an estimated sequence of symbols ŝ belonging to a set of discrete possible symbol values such that the square of the magnitude of a difference vector LHŝ−z is minimized.
A diversity hybrid receiver in a digital communication system, the receiver having M antennae, each antenna receiving one sequence of received training signal samples, the receiver including:
- a space-time filter including means for producing an intermediate sequence of signal samples from a plurality of received sequences of signal samples and a plurality of optimal filter coefficients;
- means for producing L time-taps of each sequence of received training signal samples;
- an equalizer including means for producing an estimated sequence of symbols from the intermediate sequence of signal samples and a plurality of optimal effective channel coefficients hopt;
- and
- a joint optimization processor including instructions for producing the optimal filter coefficients and the optimal effective channel coefficients from the sequences of received training signal samples x and a sequence of known training symbols s, the instructions for producing the optimal filter coefficients comprising:
- instructions for defining a matrix X from the sequences of received training signal samples as X = [ x 1 , k x 1 , k + 1 ⋯ x 1 , k + p ⋮ ⋮ ⋯ ⋮ x M , k x M , k + 1 ⋯ x M , k + p x 1 , k - 1 x 1 , k ⋯ x 1 , k + p - 1 ⋮ ⋮ ⋯ ⋮ x M , k - 1 x M , k ⋯ x M , k + p - 1 ⋮ ⋮ ⋯ ⋮ x 1 , k - L x 1 , k + 1 - L ⋯ x 1 , k + p - L ⋮ ⋮ ⋯ ⋮ x M , k - L x M , k + 1 - L ⋯ x M , k + p - L ] ∈ C M ( L + 1 ) × ( p + 1 )
- where k indicates a sequential position of a received training signal sample in a sequence of received training signal samples and p+1 indicates the number of received training signal samples in each sequence of received training signal samples;
- instructions for defining a matrix S from the sequence of known training symbols as S = [ s k s k + 1 ⋯ s k + p s k - 1 s k ⋯ ⋮ ⋮ ⋮ ⋰ ⋮ s k - v - L ⋯ ⋯ s k - v - L + p ] ∈ C ( v + L + 1 ) × N
- where v+L+1 is the number of a plurality of effective channels considered by the equalizer;
- instructions for initializing hopt to be the first column of a matrix B=(S*P*ST)−1S*ST where P=(I−XH(XXH)−1X), I is an identity matrix, and XH is the Hermitian of X;
- and
- instructions for carrying out iterative steps a predetermined number of times, the iterative steps comprising:
- calculating a vector q=Bhopt;
- and
- calculating a new value for the vector hopt=q/∥q∥.
In a diversity hybrid receiver in a digital communication system, a method of determining a plurality of optimal effective channel coefficients hopt, the receiver including M antennae, each antenna receiving one sequence of received training signal samples, the receiver also including means for producing L time-taps of each sequence of received training signal samples, the method comprising the steps of:
- defining a matrix X from the sequences of received training signal samples as X = [ x 1 , k x 1 , k + 1 ⋯ x 1 , k + p ⋮ ⋮ ⋯ ⋮ x M , k x M , k + 1 ⋯ x M , k + p x 1 , k - 1 x 1 , k ⋯ x 1 , k + p - 1 ⋮ ⋮ ⋯ ⋮ x M , k - 1 x M , k ⋯ x M , k + p - 1 ⋮ ⋮ ⋯ ⋮ x 1 , k - L x 1 , k + 1 - L ⋯ x 1 , k + p - L ⋮ ⋮ ⋯ ⋮ x M , k - L x M , k + 1 - L ⋯ x M , k + p - L ] ∈ C M ( L + 1 ) × ( p + 1 )
- where k indicates a sequential position of a received training signal sample in a sequence of received training signal samples and p+1 indicates the number of received training signal samples in each sequence of received training signal samples;
- defining a matrix S from the sequence of known training symbols as S = [ s k s k + 1 ⋯ s k + p s k - 1 s k ⋯ ⋮ ⋮ ⋮ ⋰ ⋮ s k - v - L ⋯ ⋯ s k - v - L + p ] ∈ C ( v + L + 1 ) × N
- where v+L+1 is the number of a plurality of effective channels considered by an equalizer;
- initializing hopt to be the first column of a matrix B=(S*P*ST)−1S*ST where P=(I−XH(XXH)−1X) I is an identity matrix, and XH is the Hermitian of X;
- and
- carrying out iterative steps a predetermined number of times, the iterative steps comprising:
- calculating a vector q=Bhopt;
- and
- calculating a new value for the vector hopt=q/∥q∥.
For a diversity hybrid receiver in a digital communication system,a processor containing instructions for determining a plurality of optimal effective channel coefficients hopt, the receiver including M antennae, each antenna receiving one sequence of received training signal samples, the receiver also including means for producing L time-taps of each sequence of received training signal samples, the instructions comprising:
- instructions for defining a matrix X from the sequences of received training signal samples as X = [ x 1 , k x 1 , k + 1 ⋯ x 1 , k + p ⋮ ⋮ ⋯ ⋮ x M , k x M , k + 1 ⋯ x M , k + p x 1 , k - 1 x 1 , k ⋯ x 1 , k + p - 1 ⋮ ⋮ ⋯ ⋮ x M , k - 1 x M , k ⋯ x M , k + p - 1 ⋮ ⋮ ⋯ ⋮ x 1 , k - L x 1 , k + 1 - L ⋯ x 1 , k + p - L ⋮ ⋮ ⋯ ⋮ x M , k - L x M , k + 1 - L ⋯ x M , k + p - L ] ∈ C M ( L + 1 ) × ( p + 1 )
- where k indicates a sequential position of a received training signal sample in a sequence of received training signal samples and p+1 indicates the number of received training signal samples in each sequence of received training signal samples;
- instructions for defining a matrix S from the sequence of known training symbols as S = [ s k s k + 1 ⋯ s k + p s k - 1 s k ⋯ ⋮ ⋮ ⋮ ⋰ ⋮ s k - v - L ⋯ ⋯ s k - v - L + p ] ∈ C ( v + L + 1 ) × N
- where v+L+1 is the number of a plurality of effective channels considered by an equalizer;
- instructions for initializing hopt to be the first column of a matrix B=(S*P*ST)−1S*ST where P=(I−XH(XXH)−1X), I is an identity matrix, and XH is the Hermitian of X;
- and
- instructions for carrying out iterative steps a predetermined number of times, the iterative steps comprising:
- calculating a vector q=Bhopt;
- and
- calculating a new value for the vector hopt=q/∥q∥.
In a diversity hybrid receiver in a digital communication system, a method of determining a signal-to-interference-plus-noise ratio, the receiver receiving a plurality of M sequences of received training signal samples, the receiver including means for producing L time-taps of each sequence of received training signal samples, the receiver also including means for producing a plurality of optimal filter coefficients wopt from the plurality of sequences of received training signal samples and a sequence of known training symbols, the method comprising the steps of:
- determining a plurality of orthogonal weights w⊥opt of the optimal filter coefficients wopt;
- defining a matrix X from the sequences of received training signal samples X = [ x 1 , k x 1 , k + 1 ⋯ x 1 , k + p ⋮ ⋮ ⋯ ⋮ x M , k x M , k + 1 ⋯ x M , k + p x 1 , k - 1 x 1 , k ⋯ x 1 , k + p - 1 ⋮ ⋮ ⋯ ⋮ x M , k - 1 x M , k ⋯ x M , k + p - 1 ⋮ ⋮ ⋯ ⋮ x 1 , k - L x 1 , k + 1 - L ⋯ x 1 , k + p - L ⋮ ⋮ ⋯ ⋮ x M , k - L x M , k + 1 - L ⋯ x M , k + p - L ] ∈ C M ( L + 1 ) × ( p + 1 )
- where k indicates a sequential position of a received training signal sample in a sequence of received training signal samples and p+1 indicates the number of received training signal samples in each sequence of received training signal samples;
- determining an interference pulse noise yI from an expression y _ I T = ∑ i = 1 M ( L + 1 ) - 1 ( w _ opt i ⊥ ) T X
- where M(L+1)−1 is the number of orthogonal weights;
- determining an estimated desired signal yD from an expression
- yDT=(wopt)TX;
- and
- determining the signal-to-interference-plus-noise ratio as the ratio of the square of the magnitude of the interference pulse noise to the square of the magnitude of the estimated desired signal.
In a diversity hybrid receiver in a digital communication system, an interference detection processor, the receiver receiving a plurality of M sequences of received training signal samples, the receiver including means for producing L time-taps of each sequence of received training signal samples, the receiver also including means for producing a plurality of optimal filter coefficients wopt from the plurality of sequences of received training signal samples and a sequence of known training symbols, the interference detection processor including instructions for determining a signal-to-interference-plus-noise-ratio, the instructions comprising:
- instructions for determining a plurality of orthogonal weights w⊥opt of the optimal filter coefficients wopt;
- instructions for defining a matrix X from the sequences of received training signal samples as X = [ x 1 , k x 1 , k + 1 ⋯ x 1 , k + p ⋮ ⋮ ⋯ ⋮ x M , k x M , k + 1 ⋯ x M , k + p x 1 , k - 1 x 1 , k ⋯ x 1 , k + p - 1 ⋮ ⋮ ⋯ ⋮ x M , k - 1 x M , k ⋯ x M , k + p - 1 ⋮ ⋮ ⋯ ⋮ x 1 , k - L x 1 , k + 1 - L ⋯ x 1 , k + p - L ⋮ ⋮ ⋯ ⋮ x M , k - L x M , k + 1 - L ⋯ x M , k + p - L ] ∈ C M ( L + 1 ) × ( p + 1 )
- where k indicates a sequential position of a received training signal sample in a sequence of received training signal samples and p+1 indicates the number of received training signal samples in each sequence of received training signal samples;
- instructions for determining an interference pulse noise yI from an expression y _ I T = ∑ i = 1 M ( L + 1 ) - 1 ( w _ opt i ⊥ ) T X
- where M(L+1)−1 is the number of orthogonal weights;
- instructions for determining an estimated desired signal yD from an expression
- yDT=(wopt)TX;
- and
- instructions for determining the signal-to-interference-plus-noise ratio as the ratio of the square of the magnitude of the interference pulse noise to the square of the magnitude of the estimated desired signal.
Dependencies are inferred from the claim text.