Centralized Data Plane Flow Control for Mobile Network Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile communications networks face challenges in preventing overload of data plane applications, leading to network capacity consumption and 'blind load' when rejecting data flows, as current mechanisms rely on back pressure and rejection of flows, which consume additional capacity and require sources to find alternative nodes.
Innovation Solution
A central control entity that detects data plane applications, determines their load through flow information reports from forwarding elements, and distributes data plane flows to balance the load across multiple nodes of the same type, avoiding the need for initial rejection and reducing network capacity consumption by routing flows directly to less loaded applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If data plane applications use back pressure mechanism to handle overload, then application capacity management is simplified, but network capacity is consumed and blind load is generated
Solution Approach 1:
The central control entity performs preliminary load assessment by detecting flow information reports from forwarding elements before routing data flows. This advance detection allows the system to identify less loaded applications in advance, preventing the need for back pressure rejection mechanisms and avoiding blind load generation.
Solution Approach 2:
The system implements a feedback mechanism where forwarding elements continuously report flow information to the central control entity, which then uses this information to make intelligent routing decisions. This feedback loop enables dynamic load balancing without requiring applications to implement complex back pressure management.
2Reliability
If data flows are rejected by overloaded applications, then application overload is prevented, but additional network capacity is consumed for rejection routing
Solution Approach 1:
The central control entity performs preliminary load assessment by detecting flow information reports from forwarding elements before routing data flows. This advance detection allows the system to identify less loaded applications in advance, preventing the need for back pressure rejection mechanisms and avoiding blind load generation.
Solution Approach 2:
The system takes preliminary anti-action by proactively routing flows away from overloaded applications before rejection becomes necessary. The central control entity uses detected flow information to predict overload conditions and redirects flows to less loaded applications, preventing the harmful rejection process entirely.
3Adaptability or versatility
If all data flow sources implement mechanisms to find alternative instances, then data flow routing flexibility is improved, but device complexity increases
Solution Approach 1:
The central control entity acts as an intermediary between data flow sources and applications. Instead of requiring flow sources to implement complex mechanisms for finding alternative instances, the central control entity detects flow information, assesses application loads, and makes routing decisions, thereby simplifying the overall system architecture while maintaining routing flexibility.
Solution Approach 2:
The central control entity provides a universal routing service that handles load assessment and flow distribution for all applications in the network. This multi-functional entity consolidates the complexity of finding alternative instances into a single centralized component, allowing individual flow sources to remain simple while the system as a whole maintains high adaptability.
Data Source
AI summary
The invention relates to a central control entity (100) configured to control a data plane flow of a stream of data packages in an access network part of a mobile communications network, the central control entity comprising:—a detecting unit (140), configured to detect a presence of data plane applications (300) linked to forwarding elements (200) of the access network part, configured to detect information about a type of data plane application that is to be applied to said data plane flow, and configured to detect flow information reports from forwarding elements (200) which are directly linked to one network node containing the determined type of data plane application, each flow information report from one of the forwarding elements (200) containing information about a data packet volume transmitted from said one forwarding element to said one data plane application,—a flow control unit (110) configured to determine a path of the data plane flow through the forwarding elements (200) of the access network part to one of the network nodes containing the determined type of data plane application (300), wherein the flow control unit (110) is configured to determine a load of said one data plane application provided on said one network node, taking into account the flow information reports from the forwarding elements (200) which are directly linked to said one data plane application, wherein the flow control unit (110) determines the path of the data plane flow, taking into account the determined load of said one data plane application.