Name: Sampad Kar

Student Code: BWU/BTA/22/225

Course Name: Computer Networks
Course Code: PCC-CSM502
Group: D
Session: 2024-25
Quality of Service (QoS)
Improvement Techniques
in Computer Networks
 01 Introduction to QoS
Table of
Contents
02 QoS Challenges

03 Traffic Management Strategies

04 Traffic Shaping and Policing

05 Congestion Management Techniques

06 Future Trends in QoS

 01
Introduction to QoS
Definition and Importance

Importance in Applications Goals of QoS

QoS is critical for applications The primary goals of QoS include

requiring high reliability and optimizing latency, bandwidth,
performance, such as Voice over IP jitter, and packet loss. By
(VoIP), video conferencing, and addressing these metrics,
online gaming. These applications networks can ensure that user
demand minimal delays and experiences are enhanced,
consistent throughput to function facilitating the smooth operation of
correctly, underscoring the applications and reducing the
necessity for effective QoS chances of service disruption.
strategies in network
management.
Key QoS Parameters
Bandwidth represents the Packet loss occurs when data
maximum rate of data transfer packets fail to reach their
across a network path. destination. Even small
Sufficient bandwidth is percentages of packet loss can
essential to accommodate critically impact the quality of

01 03
various applications services like VoIP and
simultaneously without causing streaming, leading to choppy
significant slowdowns or audio or video and overall
interruptions in service quality. unsatisfactory performance.

Latency Bandwidth Jitter Packet Loss

Latency refers to the time taken Jitter is the variance in packet

for data to travel from the

02
arrival times, affecting the
source to the destination. High
latency can lead to noticeable
smoothness of data streams.
High jitter can lead to uneven
04
delays, especially in real-time performance in real-time
communications, which may applications, causing
result in frustration for users interruptions or a degraded user
relying on timely data experience in video calls and
exchange. online gaming.
 02
QoS Challenges
Overview of Challenges
Network Congestion

Network congestion arises when demand for bandwidth

surpasses its availability, leading to increased latency and
packet loss. Effective management strategies are
necessary to mitigate the effects of congestion on service
performance.
Traffic Prioritization
Differentiating between traffic types is crucial in QoS
management. High-priority traffic needs guaranteed
resources to function effectively, while low-priority traffic
can tolerate delays without affecting user experience.

Variable Network Conditions

Network performance can fluctuate due to various factors
such as time of day, usage spikes, or environmental
conditions. These unpredictable changes necessitate
adaptive QoS strategies to maintain service quality.
 03
Traffic Management Strategies
Traffic Classification and Marking

Example of Voice Packet

Marking

Marking Techniques
For instance, voice packets

Classification of Traffic are often marked with a high

Marking techniques, such as priority to ensure minimal
the Differentiated Services latency and jitter. This
Traffic classification involves
Code Point (DSCP), assign classification allows the
identifying different types of
priority levels to packets. network to prioritize voice
data packets and organizing
Proper marking ensures that communication, enhancing
them based on their
high-priority packets receive call quality even during high
characteristics (e.g., voice,
expedited processing, traffic periods.
video, and data packets) to
essential for maintaining
apply specific QoS strategies
quality in sensitive
tailored to their needs.
applications.
Queueing and Scheduling Techniques
Priority Queueing Strategies
Priority queueing allows higher-priority packets to be
FIFO and Its Limitations processed before lower-priority ones. This approach is

First In First Out (FIFO) is a simple queueing crucial for ensuring that time-sensitive data, such as voice

technique where packets are processed in the and video, is delivered swiftly.

order they arrive. However, FIFO struggles to

Weighted Fair Queueing (WFQ)
provide the desired QoS, especially when managing
Weighted Fair Queueing (WFQ) offers a balanced approach
mixed traffic types with varying priorities.
by fairly allocating bandwidth to all traffic while ensuring
that higher-priority traffic is prioritized. This technique
helps achieve both fairness and responsiveness
simultaneously.

Custom Queueing Techniques

Custom queueing techniques allow network administrators
to configure multiple queues for different traffic types. This
flexibility enhances QoS by enabling specific management
approaches tailored to diverse application needs.
 04
Traffic Shaping and Policing
Traffic Shaping Methods

Overview of Traffic Shaping Techniques for Rate-

Limiting

Traffic shaping involves Techniques such as token bucket

controlling the flow of data and leaky bucket algorithms are
packets injected into the network. commonly used for rate-limiting.
By managing the rate of outgoing They help ensure that traffic does
traffic, networks can maintain not exceed set thresholds,
performance levels and reduce maintaining network stability and
the risk of congestion. efficiency.
 05
Congestion Management Techniques
PREVENTIVE Strategies

Random Early Detection (RED)

Random Early Detection (RED)

proactively drops packets before a
queue becomes full. By signaling Explicit Congestion Notification
impending congestion, RED allows (ECN)
senders to decrease transmission rates, Explicit Congestion Notification (ECN)
thereby preventing significant delays. works by marking packets to indicate
congestion without dropping them. This
mechanism allows the sender to adjust
transmission rates preemptively,
enhancing overall performance under
high load.
Reactive Strategies

Adaptation of Transmission
Rates
Adapting transmission rates based on
TCP Congestion Control real-time feedback from the network is
crucial. This reactive approach ensures
TCP congestion control techniques that network traffic stays within
monitor network conditions and manageable limits, ultimately preserving
dynamically adjust the transmission rate. QoS for all users.
By reacting to packet loss and round-trip
times, TCP can maintain optimal
throughput while preventing network
overload.
 06
Future Trends in QoS
QoS Architectures

Multiprotocol Label
1 Switching (MPLS)
Differentiated Services
Integrated Services (DiffServ) MPLS enhances traffic
(IntServ) forwarding efficiency through
Differentiated Services
the use of QoS-based labels. By
Integrated Services (IntServ) is (DiffServ) utilizes per-hop
tagging packets with priority
designed to guarantee network behaviors to classify traffic
information, MPLS can route
resources for specific flows. While into different categories,
traffic dynamically and ensure
it provides high-quality service allowing for scaled QoS
that high-priority data streams
delivery, IntServ requires implementation. This
are delivered swiftly.
complex management and can architecture strikes a balance
be challenging to scale. between performance and