8.2 QoS Models Explained
Key Concepts of QoS Models
Quality of Service (QoS) models are strategies used to manage network traffic to ensure that critical applications receive the necessary bandwidth and priority. Key concepts include:
- Best Effort: The default QoS model where all traffic is treated equally.
- Class-Based Queuing (CBQ): Traffic is classified into different classes and queued accordingly.
- Priority Queuing: Traffic is prioritized based on predefined classes.
- Weighted Fair Queuing (WFQ): Traffic is distributed fairly among different flows.
- Differentiated Services (DiffServ): Traffic is marked and treated based on predefined service levels.
- Integrated Services (IntServ): Ensures guaranteed bandwidth and delay for specific applications.
Best Effort
Best Effort is the default QoS model where all network traffic is treated equally without any prioritization. This model is simple but does not provide any guarantees for critical applications.
Example: In a home network, all devices and applications share the available bandwidth equally. If multiple devices are streaming video simultaneously, the quality may degrade due to bandwidth competition.
Class-Based Queuing (CBQ)
Class-Based Queuing (CBQ) classifies network traffic into different classes based on criteria such as source IP, destination IP, or protocol. Each class is then queued separately, allowing for more granular control over traffic management.
Example: In a corporate network, VoIP traffic can be classified into a high-priority class, while email traffic is classified into a lower-priority class. This ensures that VoIP calls are not disrupted by other traffic.
Priority Queuing
Priority Queuing assigns different priorities to network traffic based on predefined classes. Higher-priority traffic is processed before lower-priority traffic, ensuring that critical applications receive the necessary bandwidth.
Example: In a hospital network, critical patient monitoring data can be given the highest priority, ensuring that it is transmitted without delay, even during peak network usage.
Weighted Fair Queuing (WFQ)
Weighted Fair Queuing (WFQ) distributes network traffic fairly among different flows, ensuring that no single flow monopolizes the bandwidth. Each flow is assigned a weight, and traffic is queued based on these weights.
Example: In a university network, student and faculty traffic can be assigned different weights. This ensures that both groups receive a fair share of the bandwidth without one group overwhelming the network.
Differentiated Services (DiffServ)
Differentiated Services (DiffServ) marks network traffic with different service levels based on predefined criteria. Routers and switches then treat traffic based on these marks, ensuring that critical applications receive the necessary service.
Example: In a financial services network, real-time trading data can be marked with a high service level, ensuring that it is transmitted with minimal delay and jitter.
Integrated Services (IntServ)
Integrated Services (IntServ) ensures guaranteed bandwidth and delay for specific applications by reserving resources in the network. This model is more complex but provides strong guarantees for critical applications.
Example: In a military network, secure communication channels can be configured with IntServ to ensure that they receive the necessary bandwidth and delay guarantees, even during network congestion.