Budapest University of Technology and Economics, Faculty of Electrical Engineering and Informatics

    Belépés
    címtáras azonosítással
    vissza a tantárgylistához   nyomtatható verzió    

    Network and Traffic Management

    A tantárgy neve magyarul / Name of the subject in Hungarian: Hálózat- és forgalommenedzsment

    Last updated: 2024. március 5.

    Budapest University of Technology and Economics
    Faculty of Electrical Engineering and Informatics     		BSc
    Course ID Semester Assessment Credit Tantárgyfélév
    VIHIAC11 5 2/2/0/v 5  
    3. Course coordinator and department Dr. Pekár Adrián,
    4. Instructors Dr. Adrián Pekár, associate professor, Department of Networked Systems and Services
    5. Required knowledge Communication Networks
    6. Pre-requisites
    Ajánlott:

    System Modeling

    Management of Information Systems

    7. Objectives, learning outcomes and obtained knowledge The goal of the course is to deepen the knowledge acquired in the field of Communication Networks started in the previous semester by providing students with a comprehensive overview of modern networks and their services, with particular emphasis on practicality in technical implementation, feedback, and application of generated information. The principles of measuring and monitoring networks and their services, protocols and standards related to them, performance metrics, as well as the management of traditional and virtualized infrastructures, will be presented. Furthermore, advanced technologies such as control and data plane programmability and their role in optimizing networks and their services will also be introduced.

    The course of Network and Traffic Management results in the establishment of fundamental knowledge that aids in the mastery of further related network and IT subjects, including system and network performance optimization, internet security, and cloud-based technologies.


    A student who successfully completes the course:

    •    (K3) Is capable of communicating, arguing, and thinking creatively about network management.
    •    (K3) Understands and comprehends all necessary components for a complete understanding of the increasingly important field of measuring and managing networks and their services — ranging from the role played in various layers of architecture to various applications.
    •    (K3) Familiar with the full spectrum of traffic classification based on machine learning, from packet capture to data stream processing and model training processes to data analysis.
    •    (K3) Understands and comprehends the main implementations of data processing, their advantages and disadvantages, and the frequently encountered problems associated with them, aiding in making system design and implementation decisions.
    •    (K2) Possesses comprehensive basic knowledge of various aspects of measurement and management, including topology, routing and routing guidelines, performance, faults, traffic, and applications.
    •    (K3) Knows and can use the tools necessary for operating and managing networks.
    •    (K2) Understands and comprehends the differences between traditional and virtualized infrastructure management.
    •    (K2) Understands the differences in control and data plane programmability.

    8. Synopsis 1.    Overview of basic concepts and terms used in measuring networks and their traffic, introduction to network measurement infrastructures, and updating TCP/IP stack knowledge.
    2.    Overview of the pragmatics of internet measurement. Where and how measurements can be performed, how time is measured and why it is important, existing data sources, and measurement at different levels. Principles of capturing and analyzing IP packets, overview of frequently used applications and tools.
    3.    Introduction to unified standards used for supervising traffic passing through networks (SNMP, sFlow, Netflow, and IPFIX protocols). Overview of practical techniques and methods used.
    4.    Principles of Deep Packet Inspection (DPI) operation and description of its role in examining and managing network traffic. Methods of DPI application in developing ground-truth.
    5.    Description and applicability of measurement methods for popular everyday applications (web, P2P, DNS, games). Overview of practical problems associated with this.
    6.    Description of methods for uncovering and correcting connectivity, performance, security, and other network-related issues. Overview of frequently encountered related problems.
    7.    Anatomy of cyberattacks targeting computer networks and infrastructures. Overview of related concepts and terms. Presentation of methods for detecting and mitigating typical simpler attacks.
    8.    Introduction to the field of programmable data planes, explanation of the roles of SDN, NFV, and P4 paradigms in network and service management.
    9.    Overview of heavy-hitter data stream detection methods, explanation of related concepts, practical application problems, and the technique's stance.
    10.    Description of supervised machine learning applications in network classification and management. Overview of related techniques and methodologies.
    11.    Description of methods for unsupervised machine learning application in attack detection.
    12.    Description of the role of artificial intelligence in predictive maintenance. Introduction to network telemetry, explanation of basic terms and concepts. Overview of practical applications.
    13.    Current topics in network infrastructure operation and management.
    14.    Current topics in network traffic measurement, feedback, and management.


    1.    Practical application of active and passive measurement tools.
    2.    Practical handling of IP packets using tshark, tcpdump, capinfos, and editcap tools.
    3.    Configuration of SNMP, NetFlow, and IPFIX, and their practical application.
    4.    Protocol classification based on IP packet headers and payload using the nDPI library.
    5.    Practical use of Wireshark for application-level packet analysis.
    6.    Network troubleshooting and resolution in practice.
    7.    Analysis of collated, botnet, and DoS attacks in a laboratory environment.
    8.    Practical application of SDN, NFV, and P4.
    9.    Machine learning-driven detection of heavy-hitter data flows.
    10.    Implementation of time series analysis using stream processing paradigm, data processing with Elasticsearch search engine, and information visualization with Kibana interface.
    11.    Application classification with supervised machine learning 1 (overview of the framework, exploratory analysis of various publicly available datasets suitable for classification with a graphical user interface, creation of custom datasets, preparation, and labeling of datasets).
    12.    Application classification with supervised machine learning 2 (simple application of linear regression, decision tree, and KNN).
    13.    Anomaly detection with unsupervised machine learning 1 (exploratory analysis of various publicly available datasets suitable for anomaly detection, creation of custom datasets, preparation of datasets).
    14.    Anomaly detection with unsupervised machine learning 2 (simplified application of cluster analysis and K-Means clustering).

    9. Method of instruction The syllabus includes 14 lectures and 12 labs.
    •    During the lectures, theoretical aspects are presented, and a high-level overview of the materials necessary for the practical exercises is provided. The topics are explained only to the extent necessary to ensure that students understand the main objectives of network and service management elements and mechanisms, emphasizing practicality in technical implementation, and feedback and application of generated information.
    •    The majority of the labs (7 sessions) are guided labs, during which students, prepared from the knowledge acquired in the preparatory lectures, familiarize themselves with the basic tools of network and service management with the help of instructors. Based on the knowledge and skills acquired in this way, students independently solve simpler and more complex tasks in a smaller portion of the labs (5 sessions).
    •    Depending on the schedule of the semester, an additional 2 guided labs are held for partial review and integration. The scheduling of these labs depends on ensuring that the lecture always precedes the lab based on the lecture material, even in the case of missed lectures/lab sessions due to midweek holidays or breaks.
    •    Guided laboratory sessions build intensively on the topics covered in the corresponding lectures, so students must prepare from the material of these lectures for the lab sessions. The sessions start with a quick review to ensure students' preparedness.
    •    For the independent problem-solving laboratory exercises, students must prepare based on the lectures and guided lab sessions for the given thematic block. The knowledge related to independent problem-solving, and the results of the tasks are assessed through partial performance evaluation (mini exams).
    10. Assessment
    Participation in at least 70% of the laboratory sessions (guided and independent problem-solving combined) is mandatory.
    •    In the case of 14 laboratory sessions, participation of at least 10 sessions is required. If a laboratory session is canceled due to a midweek holiday or break during the semester, for 13 laboratory sessions, at least 10 sessions are mandatory, and for 12 laboratory sessions, at least 9 sessions are mandatory.
    At least 3 out of 5 independent problem-solving labs must be successfully completed.
    •    Out of the 5 mini exams (partial performance evaluations) conducted during the independent problem-solving labs, at least 3 must be successfully completed.
    •    The criterion for successful completion is achieving at least a 40% score in each case.

    A written exam is mandatory. The subsequent oral exam is optional.

    Based on the midterm work, as well as the written and oral exams, a total of 100 points can be obtained.
    •    A maximum of 30 points can be obtained based on the midterm work: each of the three most successful mini exams is worth a maximum of 10 points (in addition, a maximum of 7 bonus points can be earned from quizzes).
    •    A maximum of 30 points can be obtained in the written exam (10 points maximum for the test, 20 points maximum for practical tasks).
    •    A maximum of 40 points can be obtained in the oral exam.
    To pass the exam successfully, the total score of the midterm work, written exam, and oral exam (without the midterm bonuses) must reach 40 points. (Taking the oral exam is not mandatory; the student can accept the grade determined based on the scores obtained in the written exam and midterm work.)

    11. Recaps Participation in laboratory sessions cannot be fulfilled through makeup sessions.

    Independent problem-solving labs cannot be made up.

    Mini exams cannot be made up or retaken.
    12. Consultations In accordance with the instructors, supervisors, and laboratory coordinators of the course.
    13. References, textbooks and resources
    Lecture materials and resources for the practical exercises are available on the VIK Moodle platform.

    Recommended literature:
    •    Crovella, Mark, and Balachander Krishnamurthy. Internet measurement: infrastructure, traffic and applications. John Wiley & Sons, Inc., 2006.
    •    R. Hofstede et al., "Flow Monitoring Explained: From Packet Capture to Data Analysis With NetFlow and IPFIX," in IEEE Communications Surveys & Tutorials, vol. 16, no. 4, pp. 2037-2064, Fourthquarter 2014, doi: 10.1109/COMST.2014.2321898.

    14. Required learning hours and assignment
    Kontakt óra56
    Félévközi készülés előadásokra14
    Félévközi készülés gyakorlatokra
    		30
    Felkészülés zárthelyire-
    Házi feladat elkészítése-
    Kijelölt írásos tananyag elsajátítása-
    Vizsgafelkészülés50
    Összesen150
    15. Syllabus prepared by Dr. Adrián Pekár, associate professor, Department of Networked Systems and Services
    IMSc program IMSc students will be assigned separate homework assignments, the successful completion of which will earn IMSc points.

    On exams, extra tasks at an advanced level can be completed for IMSc points.

    At a mutually agreed-upon time, an additional, voluntarily chosen, advanced-level session will be provided, where we will discuss timely research and development issues closely related to the subject's topics and their solutions. The aim is to motivate interested (primarily, but not exclusively, IMSc) students to continue their education beyond the MSc level within the framework of PhD training.
    IMSc score Each student can earn up to 20 IMSc points according to the following:
    •    Successfully completed optional homework assignments: maximum of 15 IMSc points.
    •    Successfully completed extra tasks on exams: maximum of 5 IMSc points. These extra tasks will only be evaluated if the student has earned the highest grade on all other tasks in the given assessment.
    IMSc points can be obtained by students not participating in the IMSc program according to the above criteria.