Communications Network QoS Monitoring via Secondary Learning Process
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing DiffServ network model faces challenges in maintaining continuous Quality of Service (QoS) due to time-variance in learning intervals, leading to disruptions and unmonitored QoS breaches, especially in stateful systems where frequent monitoring is not feasible.
Innovation Solution
A method that involves a primary learning process to assign QoS models to network routes based on historical data, with a secondary learning process that dynamically reallocates communication sessions to ensure QoS compliance by splitting or re-routing traffic to maintain promised service levels, even between learning iterations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a primary learning process is used to assign QoS models to network routes based on historical data, then QoS measures can be offered for the communications network, but QoS monitoring is not continuous and breaches may go undetected between learning iterations
Solution Approach 1:
The system performs preliminary actions by predicting future QoS breaches before they occur. The secondary learning process analyzes current network conditions and predicts whether QoS thresholds will be breached in the near future, allowing proactive session reallocation before actual service degradation happens.
Solution Approach 2:
The invention maintains continuity of QoS monitoring by running a secondary learning process continuously between primary learning iterations. This secondary process constantly predicts QoS breaches and triggers reallocation, ensuring uninterrupted QoS protection even when the primary learning process is not actively executing.
2Measurement precision
If the learning process executes frequently to monitor QoS continuously, then QoS breaches can be detected promptly, but computational resources are consumed and service disruptions may still occur during re-learning
Solution Approach 1:
The monitoring system is segmented into two distinct processes: a primary learning process that executes periodically to relearn QoS models from historical data, and a secondary learning process that runs continuously between primary iterations to predict imminent breaches. This segmentation allows each process to be optimized independently - the primary process handles comprehensive model learning while the secondary process handles rapid breach prediction.
Solution Approach 2:
The secondary learning process performs partial learning - it doesn't relearn all QoS models from scratch like the primary process, but instead focuses specifically on predicting whether breaches will occur based on current trends. This partial action reduces computational overhead while maintaining effective monitoring.
3Adaptability or versatility
If QoS models are updated based on historical data from preceding learning intervals, then the network can adapt to changing conditions, but existing services experience disruptions due to remapping into new classes of service
Solution Approach 1:
The system performs preliminary reallocation of existing sessions to alternative routes before the primary learning process updates QoS models. By predicting breaches in advance and proactively moving sessions, the system prepares for upcoming model changes, minimizing disruptions when remapping occurs.
Solution Approach 2:
The invention provides beforehand cushioning by maintaining multiple route options and pre-establishing alternative paths for existing sessions. When QoS model updates are anticipated, sessions already have backup routes ready, cushioning them against disruptions from remapping into new service classes.
4Device complexity
If a single learning process is used to both learn QoS models and monitor breaches, then the system is simpler to implement, but the learning process takes significant time and QoS remains unmonitored between iterations
Solution Approach 1:
The monitoring and learning functionality is segmented into two specialized processes: a primary learning process that periodically updates QoS models from historical data, and a secondary learning process that continuously predicts breaches between primary iterations. This segmentation resolves the contradiction by providing comprehensive monitoring coverage without requiring a single overly complex process.
Solution Approach 2:
The secondary learning process serves multiple functions simultaneously - it continuously monitors network conditions, predicts QoS breaches, and triggers reallocation decisions. This multi-functionality allows the system to maintain reliable QoS monitoring without adding proportional complexity, as one process handles multiple critical tasks.
Data Source
Figure 1
Figure 2
AI summary
A method of operating a communications network, in which a primary learning process periodically determines which routes communications sessions are assigned to in order that the session can be carried with a requested QoS. A secondary learning process is used, in between instances of the primary learning process to check that the QoS constraints are still satisfied. If not, action is taken, for example using a load balancer to transmit the session over one or more further network routes.