Optical Network Grant Scheduler for Latency Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In optical networks, particularly passive optical networks (PONs), the current scheduling methods for upstream traffic from optical network terminals (ONTs) to optical line terminals (OLTs) are inefficient, leading to inadequate time slot allocation and increased latency due to reactive scheduling and high processing power requirements, especially in systems that do not support status reporting.
Innovation Solution
Implementing a grant scheduler that pre-computes grant cycles pending (GCPs) based on reported or monitored upstream data, allowing for more efficient allocation of time slots and reducing processing power needs, while also proactively allocating additional grant cycles based on downstream bandwidth usage to improve scheduling responsiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the OLT determines the upstream grant map after each scheduling round using conventional methods, then the grant map can accommodate different levels of service and past delivery of service, but the processing power and storage capacity requirements become significantly high due to the limited time available (approximately 125 μs in GPON)
Solution Approach 1:
The patent pre-calculates and stores grant cycle pending (GCP) values for each ONT before the scheduling round ends. These GCP values represent the number of grant cycles each ONT would need based on their current data amounts. By having these values ready in advance, the OLT reduces the computational burden during the actual grant map generation phase, eliminating the need for complex real-time calculations while maintaining accurate service level accommodation.
2Measurement precision
If the OLT monitors upstream traffic from ONTs that do not support status reporting, then the OLT can determine data amounts, but the scheduling becomes reactive rather than proactive, causing inadequate time slot allocation and increased latency
Solution Approach 1:
The patent implements a feedback mechanism where the OLT monitors the actual upstream traffic transmitted by ONTs and compares it with the granted time slots. Based on this feedback, the OLT adjusts future grant allocations to be more proactive. The system learns from past scheduling outcomes and modifies subsequent grants to prevent both over-allocation and under-allocation, transitioning from reactive to proactive scheduling while maintaining measurement precision.
3Productivity
If the OLT allocates time slots based on real-time status reports from all ONTs, then the scheduling can be optimized for current network conditions, but the scheduling round duration increases, reducing overall network throughput
Solution Approach 1:
The patent extracts the GCP calculation step from the main scheduling round process. Instead of calculating grant cycles for all ONTs during the scheduling round, the system pre-extracts and stores GCP values in advance. This separation allows the scheduling round to focus only on generating the grant map from pre-computed values, significantly reducing the scheduling round duration and increasing network throughput while still optimizing for current network conditions.
4Device complexity
If the OLT grants each ONT a set amount of time for upstream transmission without status reporting, then the system can operate with simpler ONTs, but the OLT cannot proactively schedule adequate time slots, hampering network operation
Solution Approach 1:
The patent enables ONTs to self-report their data amounts to the OLT, even if they cannot provide detailed status reports. Each ONT independently determines and communicates its upstream data volume, allowing the OLT to make informed scheduling decisions without requiring complex ONT functionality. This self-service approach maintains simple ONT design while ensuring reliable network operation through accurate data-driven scheduling.
Data Source
AI summary
In general, techniques are described for performing grant scheduling in optical networks. An optical line terminal (OLT) comprising a control unit may implement the techniques. The control unit determines an amount of upstream data associated with a category of service that is waiting at a first one of a plurality of ONTs to be transmitted upstream to the OLT and computes a number of GCPs for each of the ONTs based on a determined amount of data associated with the category of service that is waiting to be transmitted upstream to the OLT for each of the ONTs. After computing the number of GCPs, the control unit then grants time slots to the one or more of the ONTs based on the number of GCPs computed for each of the ONTs, wherein the time slots comprise time slots for upstream communication form the ONTs to the OLT.