Neural Network Secure Data Transport via Label Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current packet networks, particularly those using MPLS, are insecure due to reliance on IP addresses for routing, which can be easily identified by hackers, and suffer from inefficiencies such as slow routing processes, congestion issues, and lack of reliability in dynamic environments like ad-hoc or mobile networks.
Innovation Solution
A Data Neural Network (DNN) architecture that operates at Layer 4, utilizing label-switched paths to establish secure, reliable, and efficient connections by employing hunting packets to dynamically discover and optimize paths based on real-time network conditions, including bandwidth, latency, and policy constraints, without relying on external routing computers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If IP addresses are used for routing in MPLS networks, then packet delivery is achieved, but network security deteriorates because addresses can be easily identified by hackers
Solution Approach 1:
The patent extracts the routing identification function from traditional IP addresses by implementing a two-stage process: first, a route discovery packet carries the destination address to establish a path; second, subsequent data packets use only labels for forwarding without carrying the original destination address. This separates the address usage (only in route discovery) from the data transmission, eliminating address exposure during normal communication.
Solution Approach 2:
The patent introduces labels as intermediary elements that mediate between the destination address and the actual data packets. The label acts as a substitute that carries routing information during path establishment but is stripped or replaced before data transmission, preventing hackers from directly identifying destination addresses while maintaining routing functionality.
2Productivity
If traditional IP routing is used, then packet delivery is achieved, but routing efficiency deteriorates due to slow routing processes and congestion
Solution Approach 1:
The patent implements preliminary route discovery before actual data transmission. A route discovery packet is sent in advance to establish the optimal path and obtain labeling information. This preliminary action allows subsequent data packets to be forwarded immediately using pre-computed labels without undergoing time-consuming routing decisions, significantly reducing transmission delay.
Solution Approach 2:
The patent introduces dynamic label switching that adapts to real-time network conditions. The labeling information is computed based on current network state and can be updated as conditions change, allowing the system to dynamically optimize routing paths rather than relying on static IP routing tables, thereby improving efficiency in dynamic environments.
3Productivity
If MPLS label switching is used, then packet flow efficiency is improved, but network adaptability deteriorates in dynamic environments like ad-hoc or mobile networks
Solution Approach 1:
The patent implements feedback mechanisms where route discovery packets carry information about current network conditions and receive responses that inform labeling decisions. The system continuously monitors network state and adjusts labeling information accordingly, allowing MPLS to adapt to dynamic environments while maintaining efficient label-based forwarding. The feedback loop ensures that labels remain valid even as network conditions change.
Data Source
AI summary
This application discloses a neural network that also functions as a secure packet data network using an MPLS-type label switching technology. The neural network uses its intelligence to build and manage label switched paths (LSPs) to securely transport user packets and solve complex mathematical problems. This architecture is well suited to interconnect large numbers of processors or computers into a secure neural network exhibiting advanced intelligence which can be used for complex activities such as managing the power grid. However, the methods taught here can be applied to other data networks including ad-hoc, mobile, Information Centric, Content Centric, Sensor, and traditional IP packet networks, cell or frame-switched networks, time-slot networks and the like.