SERDES Wake Lock Using Retained Calibration Parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SERDES link technologies in vehicles, such as GMSL, face challenges in rapidly re-establishing link lock and starting video data transmission due to lengthy calibration processes, which can miss critical safety data during emergencies.
Innovation Solution
Implement periodic calibration of SERDES parameters and adaptive equalization settings, storing them in retention memory, allowing for quick wake-up and link lock using saved parameters to minimize calibration time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the SERDES performs full calibration process upon wake-up, then the link lock reliability is improved, but the wakeup time increases significantly
Solution Approach 1:
The patent performs calibration and adaptive equalization (AEQ) operations before the SERDES enters sleep mode, storing the calibrated parameters in retention memory. Upon wake-up, the device retrieves these pre-calibrated parameters instead of performing full calibration again, thus maintaining link lock reliability while significantly reducing wakeup time from 70-100 ms to approximately 5 ms
Solution Approach 2:
The patent changes the operational state of calibration parameters by storing them in retention memory during active mode and retrieving them during sleep/wake cycles. This parameter state change allows the system to transition from a full calibration process to a parameter retrieval process, resolving the contradiction between reliability and wakeup time
2Use of energy by moving object
If the SERDES enters sleep mode to save power, then energy consumption is reduced, but the calibration time increases when waking up
Solution Approach 1:
The system performs calibration and AEQ operations in advance before entering sleep mode, storing the results in retention memory. This preliminary action ensures that when the SERDES wakes up from sleep mode, it can quickly retrieve the stored parameters without performing time-consuming calibration, thus maintaining low power consumption while minimizing calibration time upon wake-up
Solution Approach 2:
The retention memory automatically preserves the calibration parameters without requiring continuous power, allowing the SERDES to maintain calibrated state information through sleep cycles. This self-service mechanism enables the system to retain calibration data passively, reducing both power consumption and wake-up calibration time
Data Source
AI summary
Examples of technology disclosed herein determine, by a serializer and a deserializer, one or more of calibration parameters and adaptive equalization (AEQ) parameters to perform calibration and AEQ for a communication link between the serializer and the deserializer. Such examples store the one or more parameters into a retention memory, and then put at least one of the serializer and deserializer into a sleep/low power mode. Upon receiving a wakeup command at each of the serializer and the deserializer put into the sleep mode, the serializer and the deserializer perform link lock of the communication link using the stored parameters.