Network Congestion Management: Considerations and Techniques
An Industry Whitepaper
Network congestion occurs when demand for a resource exceeds that resource’s capacity. Congestion management equates to achieving cost-savings while preserving subscriber QoE. These dual objectives are often contradictory, and a balance must be struck.
Many factors must be considered when evaluating solutions, including the regulatory environment. Generally, to be compatible with the spirit of network neutrality, a congestion management solution:
- must be narrowly tailored
- must have a proportional and reasonable effect
- must serve a legitimate technical need
A complete congestion management solution has two functional components:
- A mechanism to manage the impact of congestion
- A mechanism to trigger the management policies
Only by linking an accurate trigger with precise management can a CSP ensure that the goals of congestion management are achieved.
Real-time measurements of access round trip time provide the best possible congestion detection mechanism by informing operators precisely where, when, and for whom congestion is manifesting. Other approaches result in misapplication of management policies, in violation of best practices.
Since congestion is a result of overwhelming demand, it is up to the CSP to define:
- What applications or application types will be prioritized during congestion?
- Will congestion management policies take into account subscriber-related attributes?