Sequence Estimation via Reduced Candidate Symbol Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current symbol detection methods, particularly in WCDMA and DS-CDMA systems, face high computational complexity due to the need for extensive channel equalization and sequence estimation, which is burdensome especially when dealing with dispersive channels and extended sequence lengths.
Innovation Solution
An initial demodulation process is employed to reduce the number of possible symbol values, constraining the state spaces in sequence estimation to these reduced values, thereby simplifying the detection process and reducing computational overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If sequence estimation processing is performed with extended sequence lengths to improve detection accuracy, then reliability is improved, but device complexity increases significantly
Solution Approach 1:
The patent segments the sequence estimation process by dividing the full sequence into smaller segments or blocks. Instead of processing the entire extended sequence at once, the receiver processes shorter segments independently, reducing the computational complexity for each segment while maintaining overall detection accuracy through systematic processing of all segments.
Solution Approach 2:
The patent applies partial sequence estimation by processing only a subset of the full sequence length. Rather than performing complete MLSE on extended sequences, the receiver uses reduced complexity estimation over shorter effective sequence lengths, achieving sufficient detection performance without the full computational burden.
2Reliability
If MLSE processing is used to achieve optimal ISI cancellation, then reliability is improved, but computation complexity becomes impractical for extended sequences
Solution Approach 1:
The patent changes the parameter of sequence length from extended to reduced length. By modifying the effective sequence length parameter in the MLSE processing, the receiver achieves sufficient ISI cancellation for practical purposes while reducing the computational complexity to manageable levels, making the process feasible for extended sequences.
Solution Approach 2:
The patent performs preliminary signal processing steps before applying sequence estimation. By pre-processing the received signal to reduce interference and normalize conditions, the subsequent sequence estimation can use shorter effective sequences, reducing computational burden while maintaining cancellation performance.
3Reliability
If bidirectional DFE with arbitration is implemented to improve performance, then reliability is improved, but device complexity increases due to dual-directional processing
Solution Approach 1:
The patent merges the advantages of bidirectional processing with simplified arbitration by combining forward and backward DFE results through a unified decision-making process. Instead of fully implementing separate bidirectional paths with complex arbitration, the receiver integrates both directions' information to make single-direction decisions, reducing complexity while maintaining performance gains.
Data Source
AI summary
Teachings presented herein offer the performance advantages of sequence estimation for received signal symbol detection, while simultaneously providing potentially significant reductions in computational overhead. Initial demodulation of a received signal identifies a reduced number of candidate symbol values for all or a subset of a sequence of symbols represented in a received signal. A sequence estimation process, e.g., an MLSE process, constrains its state spaces to the reduced number of candidate symbols values, rather than considering all possible symbol values.