Distributed Scheduling for Optical Packet Contention in Mesh Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing Optical Packet Switching (OPS) mesh networks face contention issues when multiple nodes emit optical packets on the same wavelength channel, leading to collisions, which prior art systems attempt to resolve through centralized scheduling that requires network-wide traffic knowledge and centralized decision-making.
Innovation Solution
The implementation of distributed scheduling in OPS mesh networks, where nodes are bidirectionally connected for control information distribution and equipped with delay means to introduce propagation delays, allowing optical packets to be emitted at precisely defined time slots, preventing collisions by ensuring unique physical distances and time slots for each source node relative to the destination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If centralized scheduling is used to avoid optical packet contention, then packet collisions are prevented, but network complexity and control overhead increase due to requiring network-wide traffic knowledge and centralized decision-making
Solution Approach 1:
The centralized scheduling function is segmented and distributed to individual source nodes. Each node independently determines its own transmission time slots based on pre-configured distance information, eliminating the need for a central controller while maintaining contention-free operation. This segmentation reduces control plane complexity while preserving reliability.
Solution Approach 2:
Each source node autonomously determines its transmission schedule by calculating time slots based on its distance to the destination and pre-stored numbering information. Nodes serve themselves by locally computing their transmission parameters without requiring centralized scheduling decisions, thereby reducing network-wide control overhead while ensuring reliable contention avoidance.
2Device complexity
If distributed scheduling is implemented to reduce control complexity, then lightweight control plane is achieved, but coordination between nodes becomes challenging without centralized traffic knowledge
Solution Approach 1:
Time slot numbering information is pre-configured and stored at each source node based on its distance to destinations. This preliminary action provides nodes with the necessary scheduling information in advance, enabling distributed coordination without real-time centralized control. Nodes can independently determine transmission slots using this pre-stored information, achieving both low complexity and effective coordination.
Solution Approach 2:
The system transforms the scheduling problem from requiring real-time traffic knowledge to using static distance-based time slot numbering. By changing the parameter from dynamic traffic levels to static geometric distance information, nodes can independently determine transmission schedules without centralized coordination, reducing control complexity while maintaining synchronization.
3Productivity
If multiple source nodes transmit on the same wavelength channel simultaneously, then network utilization increases, but optical packet collisions occur at intermediate nodes
Solution Approach 1:
The system uses periodic time slot transmission where each source node transmits optical packets at specifically assigned time intervals. This periodic action with staggered time slots based on distance ensures that packets from multiple sources arrive at intermediate nodes at different times, allowing simultaneous transmissions on the same wavelength without collisions while maintaining high network utilization.
Solution Approach 2:
The system resolves the conflict between simultaneous transmissions and collision avoidance by introducing a time dimension. Instead of spatial separation alone, packets are separated in both space and time, with each source node assigned specific time slots based on its distance to the destination. This dimensional approach allows multiple sources to share the same wavelength channel without collisions.
Data Source
Figure 1~2B
Figure 3
Figure 4
AI summary
A method and system of avoiding optical packet contention in a synchronous OPS mesh network operated at defined times slots, and comprising interconnected nodes, of which at least two source nodes emit optical packets on the same wavelength channel toward a single destination node. All nodes in the network share control out of band control information by being bidirectionally connected (401, 411, 421) to a CID (45) element. All nodes in the network comprise delay means configured to introduce a propagation delay of optical packets as a multiple of time slots. Time slots relative to a particular destination are numbered in the network and the numbering information is stored at the nodes. When a source node emits optical packets to a destination, it broadcasts control information to all other nodes, so that a second node emitting optical packets to the same destination may make a distributed decision of optical packet emission, using the delay means, so that no contention of optical packets occurs along the way from the second source to the destination.