OTFS Turbo Decision Feedback Equalization for Low-Complexity Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing OTFS receivers suffer from high computational complexity or degraded detection performance due to methods like MAP detection, MMSE symbol detection, SMI, NLMS, and Rake receiver algorithms, which are inadequate for handling severe Doppler effects and multipath delay spreads in wireless communication channels.
Innovation Solution
A Turbo Decision Feedback Equalizer and Decoder (TDFED) system using feedforward and feedback filters in the time domain to equalize OTFS signals, reducing complexity and improving bit error rate performance through Turbo iterations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time-domain MAP detection or MMSE symbol detection is used, then detection performance is improved, but computational complexity increases extremely when multipath length exceeds 20
Solution Approach 1:
The patent segments the equalization process into two distinct domains: time-domain feedforward filtering for initial signal processing and DD-domain processing for final detection. This segmentation allows each domain to handle specific tasks efficiently, avoiding the extreme computational complexity of pure time-domain methods while maintaining detection performance.
Solution Approach 2:
The patent transitions from one-dimensional time-domain processing to two-dimensional delay-Doppler domain processing. By converting the signal to the DD domain using FFT operations and performing equalization in this transformed space, the system achieves better performance-complexity tradeoff by exploiting the structured nature of channel impairments in the DD domain.
2Device complexity
If DD-domain equalizers (SMI, NLMS, or Rake receiver with MRC) are used, then device complexity is reduced, but detection performance degrades
Solution Approach 1:
The patent merges the advantages of both time-domain and DD-domain approaches by combining feedforward filtering in the time domain with equalization in the DD domain. This hybrid structure integrates the strong interference rejection capability of time-domain methods with the computational efficiency and Doppler spread handling of DD-domain methods, achieving both low complexity and high performance.
Solution Approach 2:
The patent introduces an intermediary processing stage that converts between time-domain and DD-domain representations. This intermediary transformation layer enables the system to leverage the strengths of both domains: time-domain filtering for initial conditioning and DD-domain equalization for final detection, thereby overcoming the performance limitations of pure DD-domain methods.
3Measurement precision
If cross-domain Turbo detection algorithm is used, then detection performance is improved, but computational complexity becomes very high
Solution Approach 1:
The patent extracts and eliminates the computationally intensive iterative Turbo detection loop while retaining the beneficial cross-domain structure. By using a single-pass feedforward-filtering-then-DD-domain-equalization architecture without iterative message passing, the system achieves good detection performance without the very high computational complexity associated with iterative algorithms.
Data Source
AI summary
This disclosure pertains to procedures, methods, architectures, apparatus, systems, devices, and computer program products for, and/or directed to wireless communications, and particularly to equalization and decoding of Orthogonal Time Frequency Space (OTFS) signals.