The course provides an overview of the main techniques used to design and evaluate the performance of telecommunication networks, focusing on data distribution, network management, planning, and design, and quality of service provisioning. Class lectures will introduce techniques to support heterogeneous traffic (voice, video, data) on a single network infrastructure and will discuss how to design computer network system and how to define proper distribution techniques to reach a given target level of QoS and user satisfaction. Particular attention is devoted to algorithms on one hand and to standardization efforts on the other hand.

The student will gain knowledge on algorithms and techniques needed to design, run and manage computer networks while providing a proper level of quality to network users. Besides the knowledge on specific algorithms, the students will gain the ability of selecting the proper algorithms depending on the proper networking framework, which includes the used technology, the user-service goals, the user traffic characterization and control, and the quality of service parameters. Different service and quality of service models will be discussed and analyzed. The students will gain the ability to critically analyze standardization efforts, and to understand the reasons that drove the design choice in standards. A significant ability in discussing algorithm pros and cons is required in the final exam and will be gained during lectures through group discussions. The ability to apply the gained knowledge will be verified through class exercises and discussions, as well as during the oral examination. The oral examination will also help students in improving their communication skills.

Abilities and knowledge that will be acquired:

• Knowledge of the main methodological tools that can be used to design a networking system.
• Knowledge of the main algorithms used to control the network and provide QoS.
• Knowledge of the main techniques to manage a network.
• Knowledge of the most significant architectures for data distribution.
• Ability to understand the fundamental behavior of computer networking systems.
• Ability to select the proper set of algorithms and technologies to reach a specific quality goal.
• Ability to select the proper set of algorithms and tools to properly manage a network.

The course requires basic knowledge of probability theory, signal processing and computer networking.

• Fundamental concepts of computer networks with a QoS perspective (24h)
o Switching techniques properties (circuit, packet with datagram service, packet with virtual circuit service) (4h)
o Topologies and routing techniques (2h)
o Introduction to quality of service techniques and traffic characterization: coding techniques and applications taxonomy (8h)
o Congestion basics (2h)
o Physical layer: SDH and WDM (2h)
o Recap on main architectures and protocols dealing with quality of service (Frame relay, ATM, Ethernet, Internet, wireless LAN) (6h)

• Quality of service: standardization efforts (12h)
o QoS in Frame Relay and ATM (4h)
o Qos in Ethernet (VLANs) and wireless LAN (2h)
o QoS in Internet: Intserv and Diffserv. RSVP. MPLS. RTP and multimedia traffic. (6h)

• Algorithms to support QoS (20h)
o Network planning (2h)
o Connection Admission Control (CAC) and QoS routing (policies, source routing, hierarchical approach) (4h)
o Shaping and policing: leaky bucket, token bucket (2h)
o Scheduling algorithms: FIFO, priority, RR, DRR, WRR, and WFQ (4h)
o Congestion control and fair rate allocation (max-min fairness, buffer monitoring, TCP congestion control) (6h)
o Dropping policies and AQM (Active Queue Management) techniques (2h)

• Network management. (16h)
o SNMP protocol, MIB and ASN.1 (2h)
o Capacity planning (2h)
o Protection and restoration (2h)
o Measurements (4h)
o Running networks: case studies (6h)

• Data distribution systems (8h)
o P2P systems (4h)
o CDN and data center: the SDN approach (4h)

Seminars given by network managers (Telecom Italia, Campus Networks) will be presented as concrete examples of running networks.

Most lectures are given in a traditional fashion, with slide support. Group discussions on some of the presented algorithms are also provided to strengthen the knowledge on practical issues faced when implementing the algorithms described during the lectures and to assess the acquired level of ability. The course hosts a couple of seminars by network operators, who present the latest advances on standardization and the evolution of actual broadband communication networks.

The teaching material is made available in advance by the class teacher on the Didattica web portal.
Useful references (books, scientific papers) are as listed for student convenience.

Oral examination on all the topics covered during lectures. The oral exam lasts roughly 15-30 minutes and is typically based on three-four questions on different topics discussed during lectures. Questions are both theoretical and based on short exercise/examples to check the student ability and knowledge to translate the theoretical concepts in practice. Answers are evaluated considering their correctness, the acquired level of knowledge, the ability to be precise and to put the topic in the proper context, the capacity to clearly communicate the technical content, and the ability to apply the acquired know-how to real networks.