Multi-Lane Serial Data Receiver Frame Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-lane digital communication channels, identifying and extracting real data from serial data streams becomes complex due to the potential placement of Start-Of-Frame (SOF) and End-Of-Frame (EOF) characters across multiple lanes, making it difficult to accurately collect and store data without confusion with idle characters.
Innovation Solution
A method and apparatus that convert serial encapsulated data into sideband encapsulated data, using decoders and logic to determine lane-specific SOF and EOF characters, validate character types, and align real data into contiguous groups for proper presentation or storage, employing a parser and alignment circuit to manage character placement relative to SOF and EOF markers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple serial data lanes are used to increase data transfer capacity, then data throughput is improved, but the complexity of identifying and extracting real data from serial streams increases due to SOF and EOF characters appearing on any lane
Solution Approach 1:
The patent divides the multi-lane serial data stream into individual lane segments, each processed independently by dedicated receiver and decoder circuits. The system segments the identification of SOF/EOF characters by lane, then reassembles the real data by tracking character positions across segments. This segmentation allows parallel processing of multiple lanes while maintaining organized control over frame boundaries.
Solution Approach 2:
The patent introduces an intermediary data structure (array or buffer) that temporarily holds received characters from multiple lanes along with their lane identification and position information. This intermediary structure acts as a mediator between the parallel lane inputs and the sequential real data extraction process, allowing the system to track and organize characters before final assembly of the complete data frame.
2Productivity
If SOF and EOF characters are removed and presented as sideband signals, then data processing efficiency is improved, but accurate identification of real data boundaries becomes more difficult across multiple lanes
Solution Approach 1:
The patent performs preliminary identification and marking of SOF and EOF character positions during the initial reception and decoding phase, before the actual real data extraction occurs. By pre-tracking the positions of these boundary characters across multiple lanes and storing this information in the intermediary structure, the system prepares the necessary boundary information in advance, enabling efficient and accurate real data extraction without requiring repeated boundary detection.
Data Source
AI summary
A communication circuit comprises a plurality of receivers to receive the serial data from multiple lanes of a communication channel. The receivers may convert data received from the lanes from a serial to parallel format. Decoders may identify characters recovered from the different lanes, which collectively may define a word of width (i.e., character width) related to the number of lanes. Logic may determine when at least one of a start-of-frame and an end-of-frame character has been received. Parsing circuitry may then determine valid characters of a received word based on their placement relative to a start-of-frame character and/or an end-of-frame character. A controller may control when to present recovered data to at least one of storage registers or an output port, based on the character type identified by the decoder, its placement, an amount of characters parsed, and the number of characters already stored.