Switching Network Equalizer Training for Fast Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In point-to-point communication networks, achieving faster convergence of equalizer coefficients is challenging due to time-varying channel conditions and the need for optimized equalization settings for each channel.
Innovation Solution
The method involves a training phase where known training signals are transmitted through each channel to determine and store transmit-side equalizer coefficients. During the communication phase, these coefficients are used for pre-equalization based on the activated channel, and adaptive equalization is applied at the receiver to account for time-varying channel conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If equalization coefficients are determined through traditional adaptive equalization methods, then the equalizer can adapt to channel conditions, but the convergence speed is slow which increases latency
Solution Approach 1:
The patent applies preliminary action by pre-determining equalization coefficients during a training phase before actual data transmission. The transmitter sends known training sequences through each channel, and the receiver calculates and stores optimal equalization coefficients for each channel in advance. When switching between channels, the pre-calculated coefficients are immediately available, eliminating the need for slow adaptive convergence during actual communication. This resolves the contradiction by preparing equalization parameters beforehand, achieving fast switching without latency penalties.
2Reliability
If channel-specific equalization coefficients are optimized for each channel, then communication quality improves, but the system complexity increases due to managing multiple coefficient sets
Solution Approach 1:
The patent applies segmentation by dividing the equalization process into channel-specific segments. Each channel has its own dedicated set of equalization coefficients stored in the receiver's memory, allowing independent optimization for each channel's characteristics. The system manages multiple coefficient sets by organizing them in an associative array or lookup table structure, where each channel identifier maps to its corresponding coefficient set. This segmentation approach maintains high communication quality for each channel while managing complexity through structured organization rather than monolithic processing.
Solution Approach 2:
The receiver performs self-service by autonomously determining and storing equalization coefficients for each channel during the training phase. The receiver uses the known training sequences transmitted through each channel to calculate optimal coefficients without requiring continuous transmitter involvement. Once stored, the receiver independently selects and applies the appropriate coefficient set based on the active channel, reducing the burden on the transmitter and simplifying overall system coordination while maintaining high communication quality.
3Adaptability or versatility
If traditional adaptive equalization is used in point-to-point networks, then the system can handle time-varying channel conditions, but the convergence time is too long for latency-sensitive applications
Solution Approach 1:
The patent applies periodic action by implementing periodic training sequences at the transmitter to refresh equalization coefficients. Rather than relying solely on continuous adaptive algorithms, the system periodically re-transmits known training sequences through each channel to update and refresh the stored coefficient sets. This periodic refresh ensures that the pre-calculated coefficients remain accurate despite time-varying channel conditions, while avoiding the continuous computational overhead and slow convergence of traditional adaptive equalization. The periodic training approach maintains adaptability with minimal time loss.
Data Source
AI summary
A switching network for effecting point-to-point communication between nodes has a time-varying switching configuration, which causes successive activation and deactivation of multiple channels of the switching network, a first of the channels connecting, when activated, a transmitter node and a first receiver node, and a second of the channels connecting, when activated, the transmitter node and a second receiver node. In a training phase, a method comprises: transmitting from the transmitter node via each channel a known training signal, to cause each receiver node to receive a distorted training signal, using the first distorted training signal and knowledge of the first known training signal to determine respective one or more transmit-side equalizer (EQ) coefficients for each channel, and storing, in memory accessible to the transmitter node, the first transmit-side EQ coefficients, in association with each channel, for use in conducting scheduled communications over the switching network in a communications phase.