Back to results
US 8,275,075 B2Granted
Low complexity maximum likelihood sequence detection system and methods
AssigneeColorado State University Research Foundation
InventorJie Luo
Life of this patent
Estimating the term…
Abstract
A system and method implement low complexity maximum likelihood sequence detection. A decision feedback algorithm computes x(M+D+L−1). Optimality examination is performed for x(M), and state values and values of Markov states along paths from states in x(M) to xl(M+L) are computed.
Claims4
A method for low complexity maximum likelihood sequence detection, comprising:
- computing, with a decision feedback detector, x(M+D+L−1) using a decision feedback algorithm, wherein x is a signal sequence, M is a running index of a window center, D is a predetermined constant, and L is a memory parameter;
- performing, with an optimality examiner, optimality examination for x(M);
- and
- computing state values and values of Markov states along paths from states in x(M) to x(M+L).
1 more element
A system for low complexity maximum likelihood sequence detection, comprising:
- a detector for capturing an input signal;
- an optimality examiner for processing the input signal to determine whether it is optimal, wherein the optimality examiner has a complexity that is polynomial with respect to a memory parameter of the system;
- and
- an advanced decoder for further analyzing the input signal if it is not determined optimal.
1 more element
A method for low complexity maximum likelihood sequence detection, comprising the steps of:
- (i) initializing M=0, wherein M is a running index of a window center;
- (ii) using a decision feedback detector to obtain a decision sequence {{circumflex over (x)}(n)|n≦D+L−2}, wherein {circumflex over (x)} is a signal sequence, D is a predetermined constant, and L is a memory parameter;
- (iii) computing {{circumflex over (x)}(M+L−1)} using the decision feedback detector;
- (iv) using an optimality examination method with an optimality examiner to check whether {circumflex over (x)}(M)=xML(M), based upon {{circumflex over (x)}(n)|M−D−L+1<n<M+D+L−1};
- (v) putting {circumflex over (x)}(M) in a search list for time M if {circumflex over (x)}(M)=xML(M), and putting possible input symbols in the search list for time M if {circumflex over (x)}(M)≠xML(M);
- (vi) finding maximum likelihood (ML) sequence using a revised Viterbi algorithm to search symbols in the search list;
- (vii) incrementing M;
- and
- (viii) repeating the steps of (iii) to (vii), using the optimality examination method, putting 1 to three for M.
6 more elements
Dependencies are inferred from the claim text.