Cloud Application Recognition for Thin CPE Traffic Steering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Thin customer premises equipment (CPE) devices lack the capability to perform resource-intensive application recognition, hindering their ability to leverage advanced networking functionalities such as application steering and firewalling, which are crucial for secure and optimized network operations in software-defined wide area networks (SD-WANs).
Innovation Solution
Offloading application recognition functions to a server in the core network or network edge, which identifies applications and provides mappings to CPE devices, enabling them to implement application-aware functionalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If application recognition is performed on thin CPE devices, then application-aware functionality can be implemented, but the resource-intensive processing exceeds the capabilities of thin CPE devices
Solution Approach 1:
The patent extracts the resource-intensive application recognition function from the thin CPE device and relocates it to a server in the core network or network edge. The CPE device only needs to forward packets for inspection, while the server performs deep packet inspection and application identification, returning results to the CPE for policy enforcement.
Solution Approach 2:
The patent introduces a server as an intermediary between the thin CPE device and the applications. This intermediary performs the resource-intensive application recognition tasks that the CPE cannot handle, enabling the CPE to provide application-aware services without having the processing capability itself.
2Adaptability or versatility
If application recognition is offloaded to the cloud, then thin CPE devices can leverage advanced networking functionalities, but network latency and dependency on cloud connectivity increase
Solution Approach 1:
The server performs application recognition in advance by inspecting packets as they pass through the network, generating application identifiers and mappings before the CPE device needs to make routing decisions. This preliminary action reduces latency at the CPE device, as the heavy processing is already completed upstream.
3Measurement precision
If deep packet inspection is performed to identify applications, then accurate application recognition is achieved, but processing resources and computational power are excessively consumed
Solution Approach 1:
The patent extracts the deep packet inspection function from the CPE device and places it on a server with superior processing capabilities. This allows accurate application recognition through resource-intensive deep inspection without burdening the thin CPE device with the computational load.
Data Source
AI summary
An example method for performing application recognition in the cloud to support application-aware functionality on customer premises equipment includes detecting a data flow associated with a customer premises equipment device, identifying an application associated with the data flow, generating a mapping that maps the application to an identifier, and providing the mapping to the customer premises equipment device.


