Adaptive Transmitter Equalization Using PN11 Link Training
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional equalization techniques for 10GbE over backplanes are sub-optimal due to limited equalization speed and unsuitable criteria, which affect the performance of link partner transmitters in IEEE 802.3ap applications.
Innovation Solution
Adaptive link partner transmitter equalization is achieved by using a local transceiver to adjust equalization parameters based on the reception of a predefined training pattern, such as a PN11 pattern, and determining channel equality through a pseudo random pattern checker, while improving noise and jitter margins by varying voltage thresholds and time offsets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional equalization techniques are used, then equalization can be performed, but equalization speed is limited and performance is sub-optimal
Solution Approach 1:
The system uses the PN11 training pattern transmitted by the link partner to automatically train and adapt the equalization coefficients without requiring external calibration equipment or manual intervention. The receiver self-services by using the incoming training data to optimize its equalization parameters for the specific channel conditions.
Solution Approach 2:
The system implements feedback by continuously monitoring the quality of received training patterns and adjusting the equalization coefficients accordingly. The receiver measures channel quality metrics from the training frames and uses this feedback to iteratively improve equalization performance until optimal convergence is achieved.
2Reliability
If equalization burden is allocated between transmitter and receiver, then equalization can be achieved, but conventional allocation is sub-optimal for IEEE 802.3ap applications
Solution Approach 1:
Instead of the conventional approach where the transmitter adapts to the channel, this system inverts the approach by having the receiver adapt to the transmitter's signal characteristics. The receiver trains on the incoming training patterns and adjusts its equalization coefficients to match the specific link partner's transmission characteristics, which is more effective for backplane applications.
Solution Approach 2:
The system dynamically changes multiple parameters including equalization coefficients, decision thresholds, and timing offsets during the training process. These parameter adjustments are made iteratively based on measured channel quality to optimize the equalization performance for the specific backplane channel conditions.
3Measurement precision
If training frames are exchanged continuously, then equalization parameters can be adjusted, but training time extends the link start-up protocol duration
Solution Approach 1:
The system performs preliminary action by using the mandatory PN11 training pattern that is already transmitted as part of the IEEE 802.3ap standard. This training pattern is sent before data transmission begins, allowing the receiver to pre-adapt its equalization coefficients during the required training period, so that optimal equalization is ready when data transmission starts.
Solution Approach 2:
The system implements dynamic adaptation by continuously adjusting equalization coefficients during the training phase based on real-time measurements of the received training pattern quality. The adaptation process is dynamic and iterative, with coefficients being updated based on current channel conditions until convergence is achieved or training time expires.
Data Source
AI summary
Methods and apparatus are provided for adaptive link partner transmitter equalization. According to one aspect of the invention, a local transceiver adapts one or more equalization parameters of a link partner by receiving a training frame over a channel between the link partner and the local transceiver, wherein the training frame is comprised of a predefined training pattern; adjusting one or more of the equalization parameters of the link partner; and determining whether the equalization of the channel satisfies one or more predefined criteria based on whether the predefined training pattern is properly received by the local transceiver. The predefined training pattern can be a pseudo random pattern, such as a PN11 pattern Noise margins and jitters margins for the channel can optionally be improved.