Slice-Based Routing for In-Order Packet Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network architectures, particularly in 5G networks, face challenges in scaling to meet the increased demands of diverse devices and use cases due to their reliance on traditional routing methods that do not account for slice-based routing, leading to out-of-order data delivery and inefficiencies in resource utilization.
Innovation Solution
The implementation of programmable switches with slice-based routing capabilities, utilizing programmable ASICs like the Tofino chip and P4 language, allows for slice ID-based routing decisions, ensuring packets from the same slice are routed through the same link and using modified ECMP and LAG tables to maintain order and optimize resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional routing methods based on destination IP address are used, then routing simplicity is maintained, but slice-based routing requirements cannot be met leading to out-of-order data delivery
Solution Approach 1:
The routing table is segmented into multiple slice-specific routing tables, where each routing table corresponds to a specific network slice identified by a slice ID. This segmentation allows packets belonging to different slices to be routed through different paths while maintaining in-order delivery within each slice, resolving the contradiction between reliability and complexity by organizing routing information in a structured, slice-based manner.
Solution Approach 2:
A slice ID is introduced as an intermediary parameter in the routing decision process. The slice ID acts as a mediator between the packet header and the routing table, enabling the router to select the appropriate slice-specific routing table and ensure packets are forwarded through the correct path for their designated slice, thereby maintaining in-order delivery without requiring complete redesign of the routing architecture.
2Productivity
If link aggregation groups are used to distribute traffic across links, then load balancing is improved, but packets from the same slice may be routed on different links causing out-of-order delivery
Solution Approach 1:
The link aggregation group behavior is made locally adaptive to each network slice. Instead of applying a uniform load balancing policy across all traffic, the system applies slice-specific routing policies that ensure packets from the same slice are always forwarded through the same link, while still utilizing multiple links for different slices to maintain high throughput. This local quality differentiation resolves the contradiction between productivity and reliability.
3Adaptability or versatility
If multiple data centers are distributed across clouds to meet 5G demands, then network scalability is improved, but traditional centralized management cannot handle the increased complexity
Solution Approach 1:
The slice ID-based routing mechanism provides a universal solution that can be applied across multiple data centers distributed across different clouds. The same routing principle and data structure work consistently regardless of the underlying physical infrastructure or cloud provider, enabling scalable deployment across diverse environments while maintaining manageable complexity through a unified approach.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
In a slice-based network, switches can be programmed to perform routing functions based on a slice identifier. The switch can receive a packet and determine a slice identifier for the packet based on packet header information. The switch can use the slice identifier to determine a next hop. Using the slice identifier with a multi-path table, the switch can select an egress interface for sending the packet to the next hop. The multi-path table can ensure that traffic for a slice stays on the same interface link to the next hop, even when a link aggregation group ("LAG") is used for creation of a virtual channel across multiple interfaces or ports.