Equalizer Adaptation Using Multi-Phase Error Sampling

The patent combines center equalization and edge equalization approaches by using error signals sampled at multiple phases (including both center and edge phases) to jointly optimize equalizer settings. This merging allows simultaneous improvement of vertical eye margin (from center equalization) and horizontal eye margin (from edge equalization) that cannot be achieved by using either approach alone.

The patent extends the equalization optimization from a single sampling phase to multiple phases by introducing phase diversity. Instead of optimizing equalizer settings based on error signals from only one phase, the system utilizes error signals from multiple phases (center and edge phases), adding a temporal/phase dimension to the optimization process. This enables the system to achieve better performance in both vertical and horizontal eye margins simultaneously.

The patent merges edge equalization with center equalization by incorporating error signals from both edge phases and center phases into a unified equalizer adaptation mechanism. This combination allows the system to benefit from the horizontal eye margin improvement provided by edge equalization while simultaneously maintaining vertical eye margin performance through center equalization, resolving the trade-off between these two metrics.

The patent introduces phase diversity as an additional dimension to the equalization process. By sampling error signals at multiple phases (including edge phases and center phases) and using these diverse samples for equalizer adaptation, the system transcends the limitations of single-phase equalization and achieves improved performance in both horizontal and vertical eye margins simultaneously.

The patent creates a universal equalizer adaptation mechanism that can optimize performance across different channel types (optical channels, backplanes, etc.) using a single multi-phase sampling approach. Instead of requiring different analog circuits for different channel types, the system uses the same multi-phase error sampling and digital signal processing architecture to adapt to various channel characteristics, eliminating the need for channel-specific analog circuitry while maintaining optimal performance.

The patent replaces the need for additional analog circuits with a digital signal processing approach. Instead of using complex analog circuitry to achieve optimal performance across different channel types, the system uses digital error sampling at multiple phases and digital equalizer adaptation algorithms. This substitution of digital processing for analog circuitry simplifies the overall implementation while maintaining adaptability to different channel types.