FBMC Channel Estimation via Non-Adjacent Pilot Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current channel estimation methods for FBMC/OQAM systems face challenges due to intrinsic interference and high training overheads, especially in mMIMO scenarios with large numbers of users or flat-fading channel conditions.
Innovation Solution
A method involving the assignment of non-adjacent sub-carriers for complex pilot symbol transmission, followed by a frequency domain analysis using a Linear Channel Estimator, which allows for the interleaving of user pilots in time and frequency, reducing the need for guard bands and enabling conventional OFDM channel estimation techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If frequency domain model with Interference Approximation Method is used for channel estimation, then channel estimation accuracy is improved when channel delay spread is small, but significant deterioration occurs when channel delay spread is not adequately small compared to symbol interval
Solution Approach 1:
The patent transforms the channel estimation problem from frequency domain to time domain by changing the mathematical representation parameters. This allows the system to handle varying channel delay spread conditions effectively, as the time domain model with FFT-based estimation can accurately capture channel characteristics regardless of the ratio between symbol interval and delay spread.
Solution Approach 2:
The patent introduces an intermediate transformation step using Fast Fourier Transform (FFT) as a mediator between time domain channel impulse response and frequency domain channel frequency response. This intermediary approach enables accurate channel estimation by leveraging the computational efficiency of FFT while maintaining the accuracy benefits of time domain modeling.
2Measurement precision
If typical channel estimation methods for mMIMO-FBMC systems are used, then channel estimation can be performed, but large training overheads are required particularly for large number of users or flat-fading channel conditions
Solution Approach 1:
The patent segments the channel estimation process into efficient time-domain operations that can be performed independently for different users and subcarriers. By using FFT-based estimation, the system can process channel information in a segmented manner that reduces the overall training overhead, especially in mMIMO scenarios with many users.
Solution Approach 2:
The patent replaces the traditional frequency domain estimation mechanism with a time domain mechanism using FFT transformation. This substitution fundamentally changes how channel estimation is performed, enabling more efficient use of training signals and reducing the time overhead required for channel estimation in multi-user environments.
3Reliability
If FBMC/OQAM modulation is used instead of OFDM, then resilience against challenging channel conditions is improved, but intrinsic interference occurs due to loss of complex orthogonality
Solution Approach 1:
The patent replaces the frequency domain channel estimation mechanism with a time domain mechanism using FFT transformation. This substitution fundamentally changes how channel estimation is performed, enabling more efficient use of training signals and reducing the time overhead required for channel estimation in multi-user environments.
Data Source
Figure 1~2
Figure 3~4b
Figure 5~8
AI summary
A Channel Estimation mechanism in an FBMC or other multicarrier modulation based communication channel offering a plurality of sub-carriers, comprises transmitting pilot symbols based on the Zadoff Chu or other Constant Amplitude Zero Auto-Correlation sequence on non-adjacent (for example alternating) sub-channels. In multi-user scenarios, pilot symbols from each user device are transmitted on the same selected sub-channels, but subjected to a circular shift. These mechanisms obviating the need for Guard bands between each user devices pilot symbols, and the need for the including of a noise correlation matrix.