Centralized Controller for Network Service Placement Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network technologies face challenges in efficiently and flexibly steering traffic through inline services, lacking scalability and openness, and requiring complex configurations to minimize latency and maximize performance.
Innovation Solution
A method using a logically centralized controller to determine service dependencies and compute placement locations that minimize total delay or distance for all subscribers, allowing for iterative service placement based on dependency and delay ratios, ensuring optimal service ordering and reduced latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traffic is steered through multiple inline services using traditional routing methods, then service chaining is achieved, but total delay and latency increase
Solution Approach 1:
The patent pre-computes optimal service placement locations and traffic steering policies before traffic arrives. The controller determines the best path through inline services in advance, caching routing decisions that minimize delay. This eliminates real-time computation overhead and ensures traffic follows pre-optimized paths, reducing total delay while maintaining service chaining.
Solution Approach 2:
The system dynamically adjusts traffic steering decisions based on real-time network conditions, service load, and subscriber requirements. The controller can re-optimize service paths dynamically, moving traffic between different service instances or bypassing services when appropriate. This dynamic adaptation minimizes delay while ensuring service chaining requirements are met under varying conditions.
2Ease of operation
If traditional policy-based routing is used to forward traffic to services, then traffic steering is achieved, but configuration complexity increases
Solution Approach 1:
The patent introduces a centralized controller as an intermediary between network operators and the forwarding infrastructure. The controller abstracts complex routing decisions, translating high-level service chaining requirements into detailed traffic steering rules. Operators simply specify service requirements, and the controller automatically generates and manages the complex routing configurations, eliminating manual complexity while maintaining steering capability.
Solution Approach 2:
The system creates simplified abstract representations of service chains and traffic flows, working with these models rather than directly configuring complex routing tables. The controller maintains virtual copies of service requirements and uses these to generate actual routing decisions, reducing configuration complexity by operating at an abstract level and only translating to detailed configurations when necessary.
3Productivity
If services are placed in distributed locations to improve network performance, then scalability increases, but determining optimal placement becomes more difficult
Solution Approach 1:
The patent implements a feedback mechanism where the controller continuously monitors network performance metrics, traffic patterns, and service execution results. Based on this feedback, the controller iteratively refines service placement decisions, moving services to locations that improve overall network performance. The feedback loop enables the system to automatically discover and adapt to optimal placements without manual intervention, handling the complexity of distributed service placement.
Solution Approach 2:
The system pre-computes service placement strategies based on historical traffic data and network topology analysis. By analyzing traffic patterns in advance, the controller determines optimal service locations before deployment, reducing the complexity of real-time placement decisions. This preliminary analysis enables scalable distributed placement by pre-solving the optimization problem based on aggregate traffic characteristics.
4Adaptability or versatility
If per-subscriber service selection is implemented to increase flexibility, then the number of routing rules increases, but scalability is challenged
Solution Approach 1:
The patent implements universal service chain templates that can be applied to multiple subscribers with similar requirements. Instead of creating unique routing rules for each subscriber, the system defines reusable service chain configurations that automatically apply to groups of subscribers. The controller matches subscriber requirements to appropriate templates, reducing the number of routing rules while maintaining per-subscriber flexibility through template parameterization.
Solution Approach 2:
The system creates abstract templates representing common service chain patterns, which are then instantiated for individual subscribers. Rather than storing separate detailed routing rules for each subscriber, the controller maintains copies of template definitions and generates specific routing decisions by instantiating these templates with subscriber-specific parameters. This copying approach reduces rule complexity while preserving per-subscriber adaptability.
Data Source
AI summary
Embodiments of the present disclosure include a method and apparatus for placing services in a network. The network and a set of services to be placed within the network are defined. A set of requirements for a plurality of subscribers is determined. Each requirement is a sequence of services to traverse for each subscriber. The services are placed such that a total delay or total distance is minimized for all subscribers.