Apparatus and methods for maximum likelihood symbol detection in communications systems

An apparatus for detecting transmit symbols at a receiver side in a communications system, comprising:

• a vector to scalar splitter,
• M branches of maximum likelihood estimators, M is integer greater than 1
• a best estimate selector,
• a better estimate comparator, and
• a logic decision circuit,
• wherein the vector to scalar splitter converts a current M-dimensional maximum likelihood symbol vector estimate into M 1-dimensional seed symbols,
• wherein each of the M branches of maximum likelihood estimators is fired up by one of the M 1-dimensional seed symbols from an output of the vector to scalar splitter, respectively,
• wherein the best estimate selector selects a best maximum likelihood estimate from an output of the M branches of maximum likelihood estimators,
• wherein the better estimate comparator compares an output of the best estimate selector with a last best estimate and picks a better estimate
• wherein when a new better estimate is found from the better estimate comparator, the logic decision circuit triggers a start of a next iteration with the estimate being fed into the vector to scalar splitter, otherwise, the logic decision circuit triggers a stop of the apparatus and outputs the estimate as a maximum likelihood estimate for a transmit symbol vector.

A maximum likelihood estimator fired up by an 1-dimensional seed symbol comprising:

• a counter to count a variable i from 1 to M, M is integer greater than 1,
• an augementizer to take an 1-dimensional seed vector as an input and to output an M-dimensional augmented symbol vector,
• an M-dimensional to (i+1)-dimensional vector splitter to take an M-dimensional symbol vector as an input and to output (M−i) (i+1)-dimensional seed symbol vectors,
• an (i+1)-dimensional seed symbol selector to select one out of the (M−i) (i+1)-dimensional seed symbol vectors, (M−i) augmentizer-pivotizer circuits and one extra augmentizer-pivotizer circuit when i=1,
• a best estimate selector to select one out of (M−i) M-dimensional symbol vectors when i>1 and i<M, or to select one out of the M-dimensional symbol vector for i=1,
• a comparator to compare the output of the best estimate selector with a last best estimate and to pick a better estimate.