Application Layer Congestion Detection for QoS Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current communication networks, particularly those using packet communication technologies, face challenges in managing bandwidth and latency variability, especially in radio links, which are prone to interference and attenuation, leading to unpredictable transmission capacities and difficulties in ensuring quality of service during congestion.
Innovation Solution
A method for detecting congestion on transmission links is implemented at the application layer, where one or more data senders and receivers measure and compare data rates over defined periods, with a master equipment calculating and sending rate information to slave equipment, allowing for congestion determination and service admission policy adjustments without relying on physical or MAC layer mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If radio link adaptation mechanisms are used at physical or MAC layer, then rapid reaction to bandwidth variations is achieved, but interaction with quality of service mechanisms is prevented due to speed mismatch
Solution Approach 1:
The patent moves congestion detection from the physical/MAC layer to the application layer, creating a new dimensional level of operation. This allows the system to operate at multiple time scales: rapid physical layer adaptations continue unchanged while slower application layer mechanisms handle QoS policy decisions, resolving the speed mismatch conflict
2Adaptability or versatility
If congestion detection is performed at application layer, then service admission control can be adjusted, but reaction time is too slow for rapid bandwidth variations
Solution Approach 1:
The system performs preliminary congestion detection and prediction at the application layer using measured data rates and round trip time variations. By detecting congestion trends before they fully manifest, the system can proactively adjust service admission policies, allowing slower reaction times to be compensated by early detection
Solution Approach 2:
The patent implements dynamic adaptation where the system continuously adjusts its behavior based on current network conditions. The congestion detection mechanism adapts its measurement parameters and the service admission control dynamically adjusts its policies based on detected congestion levels, allowing the system to optimize between reaction speed and accuracy
3Speed
If physical layer radio link adaptation is used, then rapid bandwidth adjustment is achieved, but quality of service management is degraded due to lack of information upload to application layer
Solution Approach 1:
The application layer implements self-service congestion detection by measuring data rates and round trip times directly from application-level traffic. This eliminates the need for information upload from lower layers, as the application layer independently gathers all necessary information to make QoS decisions, preserving the autonomy and speed of physical layer adaptations
Data Source
AI summary
Detection of the congestion on a link based on an exchange of messages between a sender and a receiver, one sender and several receivers or several senders and one receiver. This equipment is connected to the ends of a transmission channel. One of the items of equipment is designated as the master and the others slaves. These exchanges are aimed at enabling the master to calculate and compare the rate of the data send and the rate of the data received. A divergence in the rates calculated reveals a congestion on the link. All these exchanges are made in the application layer and therefore do not involve the underlying layers.