Turbo Equalization Loop With Volterra Interference Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current turbo-equalization methods using Volterra series for correcting interference in satellite communication systems face high complexity due to the need to model both linear and non-linear interference, especially when combined with techniques like Faster-Than-Nyquist signaling, which results in suboptimal interference cancellation and increased computational burden.
Innovation Solution
A method that performs turbo-equalization in the frequency domain, using discrete Fourier transforms to separate and correct non-linear interference with Volterra series, while treating the linear interference in the frequency domain, thereby reducing the complexity of Volterra series modeling and simplifying coefficient updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Volterra series are used to model both linear and non-linear interference in the time domain, then interference correction is performed, but device complexity increases significantly
Solution Approach 1:
The patent segments the interference correction process into two distinct parts: linear interference correction using frequency domain equalization (FD-EQ) and non-linear interference correction using Volterra series. By separating these functions and applying them in different domains, the complexity of the overall system is reduced while maintaining effective interference correction.
Solution Approach 2:
The patent introduces an intermediary approach by using frequency domain equalization as a preliminary step before applying Volterra series correction. This intermediary FD-EQ stage removes the linear component of interference, allowing the subsequent Volterra series to focus solely on non-linear interference, thereby reducing computational complexity.
2Reliability
If Volterra series correct both linear and non-linear interference in the time domain, then comprehensive interference cancellation is achieved, but computational burden increases
Solution Approach 1:
The computational task is segmented into two parts: linear interference handling in the frequency domain and non-linear interference handling in the time domain. This segmentation reduces the computational burden on the Volterra series by eliminating the need to process linear interference components.
Solution Approach 2:
The patent extracts and removes the linear interference component from the signal before applying Volterra series correction. By taking out the linear part first through frequency domain equalization, the remaining computational task for the Volterra series is significantly reduced, focusing only on non-linear interference.
3Measurement precision
If Volterra coefficients are updated to follow linear interference variations due to mobility or phase noise, then tracking accuracy is improved, but complexity of the update process increases
Solution Approach 1:
The patent segments the coefficient update requirements by associating linear interference tracking with frequency domain equalization coefficients and non-linear interference tracking with Volterra series coefficients. This segmentation allows each coefficient set to be updated independently based on their respective characteristics, reducing overall update complexity.
Solution Approach 2:
The patent applies dynamic update strategies differentiated by interference type: frequency domain equalization coefficients are updated to track rapidly varying linear interference caused by mobility and phase noise, while Volterra series coefficients are updated more slowly to track the relatively stable non-linear interference characteristics.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The present invention concerns a method and device for demodulating received symbols using a turbo equalization loop for a single carrier system. A receiver: - performs a discrete Fourier transform on the received symbols, - equalizes the discrete Fourier transform received symbols in order to obtain equalized symbols, - executes an inverse discrete Fourier transform on the equalized symbols in order to obtain equalized symbols in the time domain, - executes a first subtracting step by subtracting, from equalized symbols in the time domain, non-linear interference, - de-maps the output of the first subtracting, - decodes the output of the de-mapping, - executes a second subtracting by subtracting, from the output of decoding, the input of the decoding, and the output of second subtracting is provided to the equalization, to the non-linear interference generation, the non-linear interference being obtained by a Volterra series, the coefficients of which are determined from the equalized symbols in the time domain.