Turbo Decoding Feedback for Multiuser Interference Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current communication systems, such as IEEE 802.16e, face interference issues when users share the same tile due to coinciding pilot sequences, leading to degraded performance and limited separation capabilities using conventional antenna technology.
Innovation Solution
A method employing a turbo decoder to generate whole word code bits with a confidence level exceeding a threshold, using recovered data symbols as pilots to iteratively reduce interference and limit error propagation by directly generating all code bits, rather than re-encoding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If users share the same time/frequency resources in SDMA systems, then spectrum efficiency is improved, but interference between users increases when pilot sequences coincide
Solution Approach 1:
The patent changes the parameter of pilot sequence design by using recovered data symbols with high confidence levels as pseudo-pilots, rather than relying solely on traditional pilot symbols. This parameter change allows the system to maintain spectrum efficiency while reducing interference through iterative interference cancellation based on confidently decoded symbols.
Solution Approach 2:
The patent implements a feedback mechanism where decoded data symbols with high confidence levels are fed back into the system as pseudo-pilots for subsequent interference cancellation iterations. This feedback loop enables continuous refinement of interference suppression, allowing the system to maintain high spectrum efficiency while progressively reducing user interference.
2Device complexity
If conventional antenna technology is used for user separation, then device complexity is reduced, but separation capability is limited when pilot sequences coincide
Solution Approach 1:
The patent introduces recovered data symbols with high confidence levels as intermediary elements (pseudo-pilots) that mediate between the received signal and the decoding process. These pseudo-pilots enable enhanced user separation capability without requiring complex antenna technology, as they provide additional reference information for interference cancellation.
Solution Approach 2:
The patent adds a temporal dimension to the separation capability by using iteratively recovered symbols from multiple decoding passes. Instead of relying solely on spatial separation through antenna technology, the system exploits the time dimension through iterative decoding and interference cancellation, thereby improving separation capability without increasing antenna complexity.
3Measurement precision
If re-encoding is used to generate code bits for interference cancellation, then accuracy is improved, but error propagation increases
Solution Approach 1:
The patent performs preliminary decoding to identify and select code bits with high confidence levels before using them for interference cancellation. This preliminary action ensures that only reliable code bits are used, maintaining accuracy while preventing error propagation from low-confidence decoded bits.
Solution Approach 2:
The patent changes the selection criterion for code bits from using all decoded bits to selectively using only those bits exceeding a confidence threshold. This parameter change in the selection process maintains code bit accuracy for interference cancellation while minimizing error propagation by excluding unreliable bits.
Data Source
AI summary
Disclosed is a method and communication device for suppressing interference. The method comprises performing, with a turbo decoder (314), at least one turbo decoding attempt (1106) on a received signal (1104). The turbo decoding attempt generates at least one whole word code bit therefrom (1108). The whole word code bit (1108) corresponds to a group of bits comprising a transmitted symbol. The method determines if the whole word code bit (1108) has a confidence level exceeding a given threshold (1110). If the whole word code bit (1108) does have a confidence level exceeding the given threshold, the whole code word bit is selected for use in data symbol recovery (1114).