SIMULATION AND NETWORK PERFORMANCE

The course introduces the main analysis and modeling techniques of computer networks in order to analyze and evaluate their performance.
In particular, the main topics of the course include:
- design principles of computer networks, protocols and services, with attention to the quality of the service;
- knowledge of the main formalisms for modeling computer systems, with particular attention to discrete simulation models

Expected outcomes from studying the course include the development of the following knowledge and skills:
- design of computer networks with special attention to quality of service, in particular performance;
- ability to develop a simple simulator for evaluating the performance of a network protocol or network architecture
- design of simulation experiments and scientific interpretation of the results
- identification of performance analyzes and related patterns of a communication system

The course requires knowledge of the following prerequisites:
- programming in a procedural or object-oriented language
- knowledge of the basic tools of descriptive statistics (averages, dispersion indices)
- continuous and discrete random variables: moments, maximum likelihood estimators

1. Introduction
Modeling. System models and methods for perfomance analysis and evaluation.
Analytical and simulation models and methods. Measurement.
Performance evaluation of computer systems. Computer systems performance evaluation.
Computer networks performance evaluation and analysis modeling.
Performance indices.
Types of simulation. Examples from different application domains.
Discrete event simulation.
Network simulations. Network protocols.
Network performance evaluation.

2. Fundamentals
Random number generators.
Introduction to the generation of random variables. Methods and examples.
Discrete and continuous random variables. Common probability distributions.
Monte Carlo simulation.
Introduction to basic queueing models. Markov chains.
Basic models for performance analysis, single queue, e.g. M/M/1, M/M/1/K, M/G/1, ...

3. Discrete-Event Simulation
Type of simulation. Trace-driven simulation.
Discrete-event simulation (DES). DES concepts and event scheduling.
DES examples. Simulation debugging.
Simulation verification, validation and testing.
Planning of simulasion experimets.

4. Simulation Input/Output Analysis
Stochastic properties of simulations. Transient and steady-state analysis.
Confidence intervals.
Workload characterization. Goodness of fit tests.
Analysis of the results of simulation experiments.
Results analysis techniques. Batch analysis. Correlation analysis.

5. Network Simulations
Network modeling. Topology models. Protocol models. User models.
Traffic models. Mobility models.
Examples: IEEE 802.11, TCP, VANETs.
Parallel and distributed simulation.

Material provided by the teacher
It includes: reading notes, specific scientific papers on introduction to discrete-event simulation.

Some references
- J. Banks, J.S. Carson, B. Nelson, D. Nicol, Discrete-Event System Simulation, Prentice Hall/Pearson , any ed.
- A. Law, Simulation Modeling and Analysis, McgrawHill, any ed.
- Sheldon N. Ross, Stochastic processes, J. Wiley, any ed.
- William, J. Stewart, Probability, Markov Chains, Queues, and Simulation: The Mathematical Basis of Performance Modeling, Princeton University Press, 2009

The exam can be passed in two ways:
- a lab project and a discussion of course topics
- a written intermediate test and a discussion of the thesis

The laboratory project is aimed at verifying the student's ability to apply theoretical knowledge in real case studies.

The written test consists of open questions on the topics of the course and exercises.
The exercises aim to verify the acquisition of mathematical theoretical models and the ability to apply them in problematic situations.
The open questions aim to verify the theoretical knowledge acquired by the student.

As regards the gradation of the grade (how the grades will be assigned):
1. scores in the 18-22 range will be awarded in the case of answers to the written and oral test that indicate sufficient knowledge and understanding of the topics and methods addressed in the course; limited ability to develop autonomous solutions and application of the methods studied; limited ability in explaining the methods, solutions and topics of the course.
2. scores in the 23-26 range will be awarded in the case of answers to the written and oral test that indicate a fair knowledge and understanding of the topics and methods addressed in the course; discrete ability to develop autonomous solutions and application of the methods and topics studied; discrete ability in explaining methods and solutions and ability to create connections between topics studied.
3. scores in the 27-30 range will be awarded in the case of answers to the written and oral test that indicate good or excellent knowledge and understanding of the topics and methods addressed in the course; good or excellent ability to develop autonomous solutions and apply the methods and topics studied; good or excellent ability to explain methods and solutions and ability to create connections between topics studied.
4. praise will be awarded only in the presence of demonstration of excellent knowledge, ability to understand with reference to the program, excellent ability to connect topics and illustrate with excellent communication skills.

Lectures and seminars.
Exercises proposed and solved in class.
Projects of simulation models applied to examples of computer networks.
In-depth topics.
Tutorials.

Italian

oral