Transmitter Linear Equalization for Low Latency High Speed Communication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed wireless communication systems face challenges in mitigating Intersymbol Interference (ISI) and In-phase/Quadrature (I/Q) imbalance, particularly in millimeter wavelength frequency bands, due to complex equalization requirements and processing delays, which existing technologies struggle to address effectively.
Innovation Solution
A communication system employing a combination of variable and fixed linear equalizers with I/Q imbalance compensation, using distinct training sequences in alternating frames for channel estimation and equalization, and feedback mechanisms to update coefficients and compensate for I/Q imbalances, thereby reducing latency and improving spectrum efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If receiver side linear equalization is used to equalize the channel, then the channel equalization is achieved, but the implementation complexity and latency increase significantly due to long impulse response requirements
Solution Approach 1:
The patent inverts the conventional approach by performing linear equalization at the transmitter instead of the receiver. The transmitter applies a linear equalizer with impulse response g(n) to pre-compensate for channel effects, while the receiver uses only a decision feedback equalizer. This inversion transfers the computational burden from the receiver to the transmitter, reducing receiver complexity and latency while maintaining equalization effectiveness.
2Reliability
If receiver side linear equalization is used, then channel equalization is achieved, but noise enhancement effect becomes significant
Solution Approach 1:
By inverting the equalization location from receiver to transmitter, the patent avoids the noise enhancement problem inherent in receiver-side linear equalization. The transmitter operates on the signal before it passes through the channel and noise, so the equalization process does not amplify noise. The receiver-side decision feedback equalizer then operates on a cleaner signal with reduced noise enhancement.
3Reliability
If OFDM with frequency domain equalization is used, then multipath delay spread is handled, but processing delays increase
Solution Approach 1:
The patent inverts the equalization approach from frequency domain (OFDM) to time domain at the transmitter. By applying linear equalization in the time domain before transmission, the system avoids the inherent processing delays of frequency domain equalization while still effectively mitigating multipath delay spread effects.
4Reliability
If OFDM with frequency domain equalization is used, then multipath delay spread is handled, but spectrum efficiency decreases due to guard intervals
Solution Approach 1:
The patent inverts from OFDM frequency domain processing to time domain linear equalization at the transmitter. This approach eliminates the need for guard intervals that waste spectrum resources, as the linear equalizer pre-compensates for multipath effects without requiring the time-gapping that characterizes OFDM systems.
5Reliability
If decision-feedback equalizer is used in high speed systems, then ISI mitigation is achieved, but processing speed is insufficient to satisfy data rate requirements
Solution Approach 1:
The patent inverts the equalization architecture by performing the computationally intensive linear equalization at the transmitter where high-speed processing is feasible, rather than at the receiver where firmware implementation speed becomes a bottleneck. This allows decision-feedback equalization to operate at sufficient speeds for high data rate systems while maintaining effective ISI mitigation.
Data Source
Figure 1A
Figure 1B
Figure 2A~2B
AI summary
A communication system including a transmitter and a receiver is disclosed. The transmitter transmits frames, at least two consecutive frames containing different training sequences. The receiver receives data communicated from the transmitter over a channel. The receiver combines and jointly processes the at least two consecutive frames transmitted by the transmitter to estimate a channel state of the channel.