Decision Feedback Equalizer Tracking for Severe Multipath Channels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing equalizers face challenges in converging quickly under severe multipath conditions in ATSC DTV signal reception, leading to long convergence times and error propagation issues due to the complex channel impulse responses and time-varying nature of the broadcast television channel.
Innovation Solution
A tracking decision feedback equalizer system is implemented, utilizing two 12-phase trellis decoders with different traceback depths to provide reliable symbol decisions, and a tap weight calculator that updates tap weights based on both initial and updated channel impulse responses, allowing for improved convergence and tracking of channel variations between training sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a long equalizer with many taps is used to handle long channel impulse responses, then the equalizer can compensate for multipath interference, but the convergence time becomes excessively long
Solution Approach 1:
The patent applies preliminary action by using the field sync segment (a known training sequence) to perform initial equalizer adaptation before actual data reception. This preliminary equalization using predetermined symbols allows the equalizer to converge faster when real data arrives, reducing the overall convergence time while maintaining reliable equalization performance throughout the channel impulse response.
2Adaptability or versatility
If blind and decision directed methods are added to adapt equalizers between training sequences, then the equalizer can track channel variations, but convergence may not be achieved under difficult multipath conditions and requires several data fields
Solution Approach 1:
The patent implements feedback by using the output of the trellis decoder and feeding it back to the equalizer for continuous adaptation. The equalizer uses the decoded symbols as reference to update its tap weights between training sequences, enabling the system to track channel variations efficiently without requiring multiple data fields for convergence, thus resolving the contradiction between adaptability and convergence time.
3Reliability
If error propagation control measures are taken in decision feedback equalizer under severe multipath conditions, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent introduces an intermediary element - the trellis decoder - positioned between the equalizer output and the feedback path. This intermediary performs sophisticated error correction decoding that controls error propagation more effectively than simple decision devices, while the patent manages the complexity by integrating this decoder into the overall equalization architecture and using its output for both data recovery and equalizer adaptation.
Data Source
AI summary
An output of an equalizer is decoded to produce first decoded data having a first accuracy. The output of the equalizer is further decoded to produce second decoded data having a second accuracy, and the second accuracy is greater than the first accuracy. Tap weights for the equalizer are calculated using the first and second decoded data. The calculated tap weights are applied to the taps of the equalizer.