Interconnexion Reseau - Semestre 8

Annee academique : 2023-2024
ECTS : 2.5
Categorie : Reseaux et Telecommunications

PART A - Presentation Generale du Module

Vue d'ensemble

Ce cours approfondit les concepts et protocoles permettant l'interconnexion de reseaux heterogenes pour former l'Internet global. Il couvre le routage IP, les protocoles de transport TCP et UDP, la qualite de service, les VLANs, la securite reseau, et les architectures modernes. L'objectif est de former des ingenieurs capables de concevoir, deployer et administrer des infrastructures reseau complexes.

Objectifs pedagogiques :

Maitriser les protocoles TCP/IP et le modele en couches
Configurer et administrer routeurs et commutateurs Cisco
Implementer des protocoles de routage dynamique (RIP, OSPF, BGP)
Concevoir des architectures reseau multi-sites
Mettre en oeuvre la qualite de service (QoS)
Segmenter des reseaux avec VLANs
Securiser les infrastructures (ACL, VPN, pare-feu)
Diagnostiquer et resoudre des problemes reseau

Position dans le cursus

Ce module approfondit les bases acquises precedemment :

Reseau (S5) : modele OSI, Ethernet, IP de base, TCP/UDP
Architecture Materielle (S6) : couche physique, signaux
Systemes d'Exploitation (S5) : sockets, communication inter-processus

Il prepare a :

Cloud et Edge Computing (S9) : reseaux virtuels, SDN
IoT (S9) : protocoles specifiques, contraintes reseaux
Cybersecurite : securite reseau avancee
Carriere professionnelle : administration reseau, architecte infrastructure

PART B - Experience Personnelle et Contexte d'Apprentissage

Organisation et ressources

Le module combinait theorie et pratique intensive :

Cours magistraux (15h) :

Modeles OSI et TCP/IP, encapsulation
Adressage IPv4 et IPv6, subnetting
Protocoles de routage (statique, RIP, OSPF, BGP)
TCP et UDP en detail
Qualite de service (QoS)
VLANs et commutation
Securite reseau (ACL, VPN, pare-feu)
NAT et translation d'adresses

Travaux diriges (6h) :

Calculs de subnetting
Conception d'architectures reseau
Analyse de protocoles
Configuration de routage

Travaux pratiques (12h) :

TP1 : Configuration de base routeurs Cisco, routage statique
TP2 : Protocoles de routage dynamique (RIP, OSPF)
TP3 : VLANs, inter-VLAN routing, ACLs

Outils utilises :

Cisco Packet Tracer : simulation de reseaux
GNS3 : emulation equipements reseau
Equipements Cisco : routeurs et switches physiques
Wireshark : capture et analyse de trafic
Cisco IOS : systeme d'exploitation routeurs/switches

Ressources :

Polycope complet (4AE-Interconnexion-Rzo)
TD avec exercices pratiques
3 TP guides avec configurations
Documentation Cisco
RFCs des protocoles (TCP, IP, OSPF, BGP)

Methodologie d'etude

Figure : Modele OSI a 7 couches avec protocoles et encapsulation

Phase 1 : Maitriser les fondamentaux : Reviser modele en couches, adressage IP, principes du routage.

Phase 2 : Approfondir les protocoles : Etudier en detail TCP, UDP, OSPF, BGP avec leurs mecanismes internes.

Phase 3 : Pratiquer la configuration : Utiliser Packet Tracer pour configurer des scenarios reseau de complexite croissante.

Phase 4 : Analyser le trafic : Capturer avec Wireshark pour observer les echanges protocolaires reels.

Phase 5 : Developper le troubleshooting : Diagnostiquer pannes reseau selon methodologie (bottom-up ou top-down du modele OSI).

Difficultes rencontrees

Complexite des protocoles : OSPF et BGP sont des protocoles tres riches avec de nombreux parametres et cas particuliers.

Syntaxe Cisco IOS : La ligne de commande Cisco a sa propre logique et commandes specifiques a memoriser.

Debogage reseau : Identifier la source d'un probleme dans un reseau complexe demande methodologie et patience.

Abstraction vs realite : Les simulations sont simplifiees. Les reseaux reels ont des contraintes physiques, latences, pannes aleatoires.

PART C - Aspects Techniques Detailles

1. Modeles de reference reseau

Modele OSI (7 couches) :

Modele de reference theorique pour l'architecture des reseaux.

Couche	Nom	Role	Exemples
7	Application	Services aux applications	HTTP, FTP, DNS, SMTP
6	Presentation	Format, chiffrement, compression	SSL/TLS, JPEG, MPEG
5	Session	Gestion des sessions	NetBIOS, RPC
4	Transport	Transmission bout en bout	TCP, UDP
3	Reseau	Routage, adressage logique	IP, ICMP, OSPF, BGP
2	Liaison	Transmission sur lien physique, adressage MAC	Ethernet, WiFi, PPP
1	Physique	Transmission bits (signaux electriques, optiques)	Cables, ondes radio

Modele TCP/IP (4 couches) :

Modele pratique utilise sur Internet.

Couche	Equivalent OSI	Protocoles principaux
Application	5-6-7	HTTP, FTP, DNS, SMTP, SSH
Transport	4	TCP, UDP
Internet	3	IP, ICMP, ARP
Acces reseau	1-2	Ethernet, WiFi

Encapsulation :

Chaque couche ajoute son en-tete aux donnees de la couche superieure.

Donnees application → Segment TCP → Paquet IP → Trame Ethernet

Desencapsulation :

Le recepteur retire les en-tetes couche par couche pour extraire les donnees.

2. Adressage IPv4

Format :

Adresse sur 32 bits, notee en 4 octets decimaux separes par des points.

Exemple : 192.168.1.10

Classes historiques (obsoletes) :

Remplacees par CIDR, mais utiles pour comprendre l'evolution.

Classe A : premier bit = 0, reseau sur 8 bits (grands reseaux)
Classe B : premiers bits = 10, reseau sur 16 bits
Classe C : premiers bits = 110, reseau sur 24 bits

CIDR (Classless Inter-Domain Routing) :

Notation flexible : adresse IP / longueur prefixe reseau

Exemple : 192.168.10.0/24 signifie reseau sur 24 bits, 8 bits pour hotes (256 adresses).

Masque de sous-reseau :

Determine quelle partie de l'adresse est le reseau et quelle partie est l'hote.

/24 correspond au masque 255.255.255.0

Adresses speciales :

Adresse reseau : tous les bits hote a 0 (ex : 192.168.1.0)
Adresse broadcast : tous les bits hote a 1 (ex : 192.168.1.255)
Loopback : 127.0.0.1 (interface de bouclage local)
Plages privees (RFC 1918) : non routables sur Internet
- 10.0.0.0/8
- 172.16.0.0/12
- 192.168.0.0/16

Subnetting :

Division d'un reseau en sous-reseaux plus petits pour optimiser l'utilisation des adresses et segmenter le reseau.

Exemple : 192.168.1.0/24 divise en 4 sous-reseaux /26 :

192.168.1.0/26 (adresses .0 a .63)
192.168.1.64/26 (adresses .64 a .127)
192.168.1.128/26 (adresses .128 a .191)
192.168.1.192/26 (adresses .192 a .255)

3. IPv6

Motivation :

Epuisement des adresses IPv4 publiques. IPv6 offre un espace d'adressage quasi illimite.

Format :

Adresse sur 128 bits, notee en 8 groupes de 4 chiffres hexadecimaux separes par deux-points.

Exemple : 2001:0db8:85a3:0000:0000:8a2e:0370:7334

Simplifications d'ecriture :

Omettre les zeros en tete de chaque groupe : 2001:db8:85a3:0:0:8a2e:370:7334
Remplacer une sequence de groupes nuls consecutifs par :: (une seule fois) : 2001:db8:85a3::8a2e:370:7334

Avantages :

Espace d'adressage immense (2 puissance 128 adresses)
Autoconfiguration simplifiee (SLAAC)
IPsec integre natif
En-tete simplifie (meilleure performance)
Plus besoin de NAT

Types d'adresses :

Unicast : adresse unique identifiant une interface
Multicast : groupe d'interfaces (remplace broadcast)
Anycast : groupe d'interfaces, route vers la plus proche

Adresses speciales :

Loopback : ::1
Link-local : fe80::/10 (auto-configurees, communication sur lien local)
Unique local : fc00::/7 (equivalent des privees IPv4)
Global unicast : 2000::/3 (routables sur Internet)

4. Routage - Principes fondamentaux

Role du routeur :

Transferer des paquets IP entre reseaux differents en consultant sa table de routage.

Table de routage :

Contient les routes vers les reseaux de destination.

Chaque entree specifie :

Reseau de destination avec masque
Passerelle (next hop) : adresse IP du prochain routeur ou "directement connecte"
Interface de sortie : par quelle interface envoyer le paquet
Metrique : cout de la route (distance, bande passante, etc.)

Processus de routage :

Recevoir paquet IP
Consulter table de routage pour trouver meilleure route (longest prefix match)
Decrementer TTL (Time To Live)
Recalculer checksum de l'en-tete IP
Transferer paquet sur l'interface de sortie appropriee

Route par defaut (default route) :

Route 0.0.0.0/0 utilisee si aucune route specifique ne correspond. Typiquement vers Internet via passerelle FAI.

Routage statique vs dynamique :

Statique : routes configurees manuellement, adaptees pour petits reseaux ou routes specifiques
Dynamique : protocoles de routage automatisent la decouverte et mise a jour des routes

5. Protocoles de routage dynamique

Classification :

IGP (Interior Gateway Protocol) : au sein d'un systeme autonome (AS)

Distance vector : RIP, EIGRP
Link state : OSPF, IS-IS

EGP (Exterior Gateway Protocol) : entre systemes autonomes

Path vector : BGP

RIP (Routing Information Protocol) :

Protocole de routage distance-vector simple.

Caracteristiques :

Metrique : nombre de sauts (hop count), maximum 15
Algorithme Bellman-Ford
Mise a jour complete de la table toutes les 30 secondes
Convergence lente
Simple mais limite aux petits reseaux

OSPF (Open Shortest Path First) :

Protocole link-state avance.

Principe :

Chaque routeur connait la topologie complete de la zone
Algorithme SPF (Dijkstra) calcule l'arbre des plus courts chemins
Metrique basee sur le cout (fonction de la bande passante)

Fonctionnement :

Decouverte des voisins via paquets Hello
Echange de LSA (Link State Advertisement) decrivant les liens
Construction de la base de donnees topologique (LSDB)
Calcul SPF pour determiner meilleures routes
Installation des routes dans la table de routage

Avantages :

Convergence rapide (mises a jour declenchees, pas periodiques)
Scalabilite via hierarchie (aires)
Supporte VLSM et CIDR
Authentification des mises a jour

Hierarchie OSPF :

Organisation en aires pour reduire trafic et taille des tables.

Aire 0 (backbone) : aire centrale obligatoire
Aires non-backbone : connectees a l'aire 0 via ABR (Area Border Router)

BGP (Border Gateway Protocol) :

Protocole de routage d'Internet entre systemes autonomes.

Caracteristiques :

Path vector : inclut le chemin complet d'AS
Politiques de routage complexes (business, preferences)
eBGP (external) entre AS differents, iBGP (internal) au sein d'un AS
Metrique : pas seulement technique mais aussi politique

Role :

BGP maintient la table de routage globale d'Internet. Chaque AS annonce ses prefixes IP.

6. Protocole TCP (Transmission Control Protocol)

Caracteristiques :

Protocole de transport oriente connexion, fiable.

Oriente connexion : etablissement 3-way handshake avant echange de donnees
Fiable : accuses de reception, retransmission des segments perdus
Ordre preserve : numeros de sequence garantissent l'ordre
Controle de flux : fenetre glissante evite debordement recepteur
Controle de congestion : adapte debit selon etat du reseau

En-tete TCP (principaux champs) :

Ports source et destination (16 bits chacun) : identifient applications
Numero de sequence (32 bits) : position des donnees dans le flux
Numero d'accuse de reception (32 bits) : prochain octet attendu
Flags : SYN (synchronisation), ACK (accuse), FIN (fin), RST (reset), PSH (push), URG (urgent)
Fenetre (16 bits) : taille du buffer recepteur disponible (controle de flux)
Checksum (16 bits) : detection d'erreurs

Etablissement de connexion (3-way handshake) :

Client → Serveur : SYN (synchronisation, numero de sequence initial)
Serveur → Client : SYN+ACK (acquittement + son propre numero de sequence)
Client → Serveur : ACK (acquittement final)

Connexion etablie, echange de donnees peut commencer.

Fermeture de connexion (4-way handshake) :

A → B : FIN (A a fini d'envoyer)
B → A : ACK (acquittement)
B → A : FIN (B a fini d'envoyer)
A → B : ACK (acquittement final)

Ou fermeture brutale avec RST (reset).

Controle de flux (fenetre glissante) :

Le recepteur annonce la taille de son buffer disponible (fenetre).

L'emetteur ne peut pas envoyer plus de donnees que la taille de la fenetre.

Evite debordement du buffer recepteur.

Controle de congestion :

Mecanisme pour adapter le debit selon l'etat du reseau et eviter l'effondrement.

Phases :

Slow Start : augmentation exponentielle de la fenetre de congestion
Congestion Avoidance : augmentation lineaire apres atteinte d'un seuil
Fast Retransmit : retransmission immediate apres 3 ACK dupliques
Fast Recovery : reduction moderee de la fenetre apres perte

Algorithmes : TCP Tahoe, Reno, New Reno, CUBIC (utilise par Linux par defaut).

7. Protocole UDP (User Datagram Protocol)

Caracteristiques :

Protocole de transport simple, non connecte, non fiable.

Sans connexion : pas d'etablissement de connexion
Non fiable : pas d'accuse de reception, pas de retransmission
Pas de controle de flux ni de congestion
Pas de garantie d'ordre
En-tete minimal : seulement 8 octets

En-tete UDP :

Port source (16 bits)
Port destination (16 bits)
Longueur (16 bits)
Checksum (16 bits)

Utilisations :

Applications ou la rapidite prime sur la fiabilite :

VoIP et videoconference : temps reel, perte acceptable
Streaming video/audio : retransmission inutile (deja trop tard)
DNS : requetes courtes, retransmission geree par application
DHCP : configuration reseau initiale
Jeux en ligne : latence critique
SNMP : gestion reseau simple

Avantages :

Faible overhead, latence minimale, adapte aux applications temps reel.

8. Qualite de Service (QoS)

Problematique :

Le reseau Internet est "best-effort" (meilleur effort) : aucune garantie de bande passante, latence, ou fiabilite.

Insuffisant pour applications exigeantes : VoIP (voix), videoconference, streaming.

Parametres de QoS :

Parametre	Description	Importance pour
Bande passante	Debit garanti	Streaming video
Latence	Delai de transit	VoIP, jeux
Gigue (jitter)	Variation de latence	VoIP, video
Perte de paquets	Taux de paquets perdus	Toutes applications

Mecanismes de QoS :

Classification et marquage :

Identifier le type de trafic et le marquer dans l'en-tete pour traitement differencie.

DSCP (Differentiated Services Code Point) : 6 bits dans en-tete IP
CoS (Class of Service) : 3 bits dans tag 802.1Q (VLAN)

Files d'attente (Queuing) :

Gerer priorites et ordonnancement des paquets.

FIFO : First In First Out (pas de priorite)
Priority Queuing : files de differentes priorites (stricte)
WFQ (Weighted Fair Queuing) : partage equitable avec ponderation
CBWFQ (Class-Based WFQ) : WFQ par classe de trafic
LLQ (Low Latency Queuing) : file prioritaire + WFQ pour le reste

Policing et Shaping :

Controler le taux de trafic.

Policing : limiter le debit, rejeter ou remarquer l'excedent
Shaping : lisser le trafic, tamponner l'excedent dans une file

Modeles de QoS :

IntServ (Integrated Services) : reservation de ressources par flux (RSVP), peu deploye (complexite)
DiffServ (Differentiated Services) : classes de service agregees, scalable, largement deploye

9. VLANs (Virtual LANs)

Principe :

Segmenter un reseau physique en plusieurs reseaux logiques isoles.

Avantages :

Isolation des domaines de broadcast : reduit trafic inutile
Securite : separation logique du trafic (finance, RH, invites)
Flexibilite : reorganisation sans recablage physique
Optimisation : grouper utilisateurs par fonction plutot que par localisation

Standard 802.1Q (VLAN Tagging) :

Ajout d'un tag de 4 octets dans la trame Ethernet contenant :

VLAN ID (12 bits) : identifiant du VLAN (4096 VLANs possibles)
Priority (3 bits) : priorite CoS pour QoS

Types de ports :

Type	Description	Utilisation
Access	Appartient a un seul VLAN, pas de tag	Connexion terminaux
Trunk	Transporte plusieurs VLANs, avec tags	Interconnexion switches

Communication inter-VLANs :

Les VLANs sont isoles. Pour communiquer entre eux, il faut un routeur (ou switch layer 3).

Methodes :

Router on a stick : un seul lien physique, sous-interfaces par VLAN
Switch L3 : routage inter-VLAN directement sur le switch

10. Commutation (Switching)

Principe :

Un switch (commutateur) transfere des trames Ethernet selon les adresses MAC.

Table CAM (Content Addressable Memory) :

Associe adresses MAC aux ports du switch.

Apprentissage :

Le switch apprend les adresses MAC en observant l'adresse source des trames recues.

Transfert (Forwarding) :

Adresse destination connue : transferer sur le port associe dans la table CAM
Adresse inconnue : flood (diffuser) sur tous les ports sauf celui de reception
Broadcast ou multicast : flood

Spanning Tree Protocol (STP) :

Protocole pour eviter les boucles dans les topologies redondantes (liens multiples entre switches).

Fonctionnement :

Election d'un root bridge (switch racine)
Calcul du chemin optimal vers le root bridge
Blocage des ports creant des boucles
Reconfiguration automatique en cas de panne

Variantes : RSTP (Rapid STP) pour convergence plus rapide, MSTP (Multiple STP) pour plusieurs VLANs.

11. NAT (Network Address Translation)

Problematique :

Penurie d'adresses IPv4 publiques. Toutes les machines ne peuvent pas avoir une adresse publique.

Principe :

Traduire des adresses IP privees en adresses publiques lors de la traversee d'un routeur.

Types de NAT :

NAT statique :

Mapping fixe 1:1 entre une adresse privee et une adresse publique.

Usage : serveurs accessibles depuis Internet.

NAT dynamique :

Pool d'adresses publiques, allocation dynamique aux machines internes.

PAT (Port Address Translation) ou NAT Overload :

Plusieurs machines internes partagent une seule adresse publique.

Differenciation par numeros de port.

Table NAT associe : IP privee:port prive ↔ IP publique:port public

Usage : la plupart des box Internet (tous les equipements domestiques partagent l'IP publique).

Limitations du NAT :

Casse la connectivite bout-a-bout (end-to-end)
Problemes pour protocoles incluant adresses IP dans donnees (FTP, SIP)
Difficultes pour applications P2P et serveurs entrants
Necessite techniques de traversee (STUN, TURN) pour VoIP

IPv6 et NAT :

IPv6 rend NAT inutile (adresses suffisantes pour tous). Retour a la connectivite bout-a-bout.

12. Securite reseau

Listes de Controle d'Acces (ACL) :

Filtrer le trafic selon criteres definis (adresses IP, ports, protocoles).

Types :

ACL standard : filtre uniquement sur adresse IP source
ACL etendue : filtre sur IP source/destination, ports, protocoles

Application : sur interfaces de routeurs (trafic entrant ou sortant).

Pare-feu (Firewall) :

Equipement ou logiciel filtrant le trafic de maniere plus sophistiquee.

Fonctionnalites :

Filtrage stateful (inspection d'etat des connexions)
Filtrage applicatif (deep packet inspection)
Prevention d'intrusions (IPS)
Journalisation et alertes

VPN (Virtual Private Network) :

Creer un tunnel securise a travers un reseau non securise (Internet).

Types :

Site-to-Site : interconnecter des sites distants d'une entreprise
Remote Access : permettre acces distant securise (teletravail)

Protocoles :

IPsec : chiffrement au niveau IP (tunnel ou transport)
SSL/TLS VPN : base sur HTTPS, plus simple pour remote access

Menaces courantes :

Menace	Description	Contre-mesure
DoS/DDoS	Saturation de ressources	Filtrage, limitation de taux, CDN
Spoofing	Usurpation d'adresse IP	Filtrage ingress/egress, authentification
Man-in-the-Middle	Interception de communication	Chiffrement (TLS, IPsec)
ARP poisoning	Corruption de tables ARP	Dynamic ARP Inspection
VLAN hopping	Acces non autorise a VLANs	Configuration securisee trunk/access

Bonnes pratiques :

Defense en profondeur (multiples couches de securite)
Principe du moindre privilege
Segmentation reseau (VLANs, zones DMZ)
Mises a jour regulieres (firmware, patches)
Surveillance et detection (IDS, SIEM)
Sauvegardes de configurations

PART D - Analyse Reflexive et Perspectives

Competences acquises

Comprehension approfondie TCP/IP : Maitrise des protocoles fondamentaux d'Internet et de leur fonctionnement interne.

Configuration d'equipements reseau : Capacite a configurer routeurs et switches Cisco via ligne de commande (CLI).

Conception d'architectures reseau : Aptitude a concevoir des infrastructures multi-sites avec routage, VLANs, QoS.

Troubleshooting methodique : Competence pour diagnostiquer et resoudre des problemes reseau en suivant le modele en couches.

Securisation des infrastructures : Capacite a implementer ACLs, VPN, segmentation pour securiser un reseau.

Points cles a retenir

1. Le modele en couches est fondamental : Comprendre l'encapsulation et le role de chaque couche permet de diagnostiquer efficacement les problemes.

2. OSPF est le protocole IGP moderne : RIP est obsolete pour reseaux de taille significative. OSPF offre scalabilite et convergence rapide.

3. TCP assure la fiabilite, UDP la rapidite : Choisir le bon protocole de transport selon les besoins de l'application.

4. La QoS est cruciale pour temps reel : VoIP et video necessitent priorite et garanties pour fonctionner correctement.

5. VLANs = segmentation logique efficace : Organiser le reseau en VLANs ameliore securite, performance, et gestion.

Retour d'experience

Aspect formateur : Les TP sur equipements Cisco ont donne une dimension concrete a la theorie. Voir les routes apparaitre dans la table apres configuration d'OSPF est satisfaisant.

Complexite de Cisco IOS : La CLI Cisco a une courbe d'apprentissage. Les modes (user, privileged, configuration) et la hierarchie des commandes demandent pratique.

Importance du troubleshooting : Les configurations ne fonctionnent pas toujours du premier coup. Apprendre a diagnostiquer methodiquement (ping, traceroute, show commands) est essentiel.

Protocoles robustes : TCP et OSPF sont des merveilles d'ingenierie. Leur robustesse face aux pannes et congestions est impressionnante.

Applications pratiques

Pour administrateur reseau :

Deployer et maintenir infrastructures reseau d'entreprise
Configurer routeurs et switches
Implementer redondance et haute disponibilite
Assurer securite et QoS

Pour architecte reseau :

Concevoir architectures multi-sites
Dimensionner bande passante
Choisir equipements et protocoles
Planifier evolution et scalabilite

Pour ingenieur DevOps :

Comprendre reseaux cloud (VPC, subnets, routing tables)
Automatiser configurations reseau (Ansible, Terraform)
Optimiser performances applications distribuees
Implementer Load Balancers

Pour developpeur :

Comprendre latence reseau et son impact sur applications
Choisir protocoles adaptes (TCP vs UDP)
Implementer retry logic et timeouts
Optimiser pour WAN (compression, caching)

Limites et ouvertures

Limites du module :

Focus sur Cisco (marche plus diversifie : Juniper, Arista, etc.)
Peu sur IPv6 (transition inevitable)
Pas de SDN (Software-Defined Networking)
Automatisation peu abordee

Ouvertures vers :

SDN et NFV : virtualisation reseau, OpenFlow, controleurs
Cloud networking : AWS VPC, Azure Virtual Networks
Automation : Ansible, Python avec Netmiko/NAPALM
Monitoring : SNMP, NetFlow, observabilite
5G et reseaux mobiles : architecture, slicing

Evolutions recentes

Software-Defined Networking (SDN) :

Separation du plan de controle (decisions de routage) et du plan de donnees (forwarding).

Controleur centralise programmable via API.

Network Functions Virtualization (NFV) :

Fonctions reseau (firewall, load balancer) en logiciel plutot que materiel dedie.

Flexibilite et reduction des couts.

Intent-Based Networking :

Definir l'intention (politique) plutot que configuration detaillee.

Le systeme traduit automatiquement l'intention en configuration.

Automation et Infrastructure as Code :

Configuration reseau declaree dans des fichiers (Ansible, Terraform).

Versionning, reproductibilite, scalabilite.

Reseaux 5G :

Latence ultra-faible, haut debit, slicing (reseaux virtuels dedies par usage).

Conseils pour reussir

1. Pratiquer regulierement : Utiliser Packet Tracer ou GNS3 pour reproduire scenarios et experimenter configurations.

2. Comprendre, pas memoriser : Comprendre le fonctionnement des protocoles plutot que memoriser syntaxe CLI.

3. Documenter ses configurations : Prendre notes des commandes, architectures deployees. Constituer un referentiel personnel.

4. Utiliser Wireshark : Capturer et analyser trafic reel pour voir protocoles en action.

5. Suivre la methodologie de troubleshooting : Bottom-up (couche 1 → 7) ou top-down selon symptomes. Methodique et systematique.

6. Se certifier : Cisco CCNA est une certification reconnue validant competences reseau de base.

Conclusion

Ce module est essentiel pour tout ingenieur travaillant avec des systemes connectes. Les reseaux sont omnipresents : entreprises, datacenters, cloud, IoT, telecommunications.

Competences transferables :

Vision globale des architectures reseau
Capacite a diagnostiquer et resoudre problemes complexes
Comprehension des compromis (performance, securite, cout)
Maitrise des protocoles universels (TCP/IP)

Pertinence professionnelle : Les competences reseau sont recherchees dans tous les secteurs. L'explosion du cloud, de l'IoT, et des applications distribuees accroit la demande d'ingenieurs reseau qualifies.

Message principal : Internet fonctionne grace a des protocoles robustes et des ingenieurs competents. Comprendre ces fondamentaux permet de concevoir, deployer et maintenir des infrastructures fiables et performantes.

Recommandations :

Continuer la pratique (home lab avec vieux routeurs ou VMs)
Se former sur automatisation (Python pour reseaux)
Explorer SDN (controleurs OpenDaylight, ONOS)
Suivre l'actualite (RFCs, blogs techniques, conferences)
Envisager certifications (CCNA, CCNP, ou equivalents autres vendors)

Liens avec les autres cours :

Reseau - S5 : bases TCP/IP
Cloud et Edge Computing - S9 : reseaux virtuels
Embedded IA for IoT - S9 : connectivite objets
Temps Reel - S8 : contraintes temporelles

Cours suivi en 2023-2024 a l'INSA Toulouse, Departement Genie Electrique et Informatique.

Network Interconnection - Semester 8

Academic Year: 2023-2024
ECTS: 2.5
Category: Networks and Telecommunications

PART A - General Module Overview

Overview

This course delves into the concepts and protocols enabling the interconnection of heterogeneous networks to form the global Internet. It covers IP routing, TCP and UDP transport protocols, quality of service, VLANs, network security, and modern architectures. The goal is to train engineers capable of designing, deploying, and administering complex network infrastructures.

Learning Objectives:

Master TCP/IP protocols and the layered model
Configure and administer Cisco routers and switches
Implement dynamic routing protocols (RIP, OSPF, BGP)
Design multi-site network architectures
Implement Quality of Service (QoS)
Segment networks with VLANs
Secure infrastructures (ACL, VPN, firewall)
Diagnose and resolve network issues

Position in the Curriculum

This module builds on previously acquired foundations:

Networking (S5): OSI model, Ethernet, basic IP, TCP/UDP
Hardware Architecture (S6): physical layer, signals
Operating Systems (S5): sockets, inter-process communication

It prepares for:

Cloud and Edge Computing (S9): virtual networks, SDN
IoT (S9): specific protocols, network constraints
Cybersecurity: advanced network security
Professional career: network administration, infrastructure architect

PART B - Personal Experience and Learning Context

Organization and Resources

The module combined theory with intensive hands-on practice:

Lectures (15h):

OSI and TCP/IP models, encapsulation
IPv4 and IPv6 addressing, subnetting
Routing protocols (static, RIP, OSPF, BGP)
TCP and UDP in detail
Quality of Service (QoS)
VLANs and switching
Network security (ACL, VPN, firewall)
NAT and address translation

Tutorials (6h):

Subnetting calculations
Network architecture design
Protocol analysis
Routing configuration

Lab Work (12h):

Lab 1: Basic Cisco router configuration, static routing
Lab 2: Dynamic routing protocols (RIP, OSPF)
Lab 3: VLANs, inter-VLAN routing, ACLs

Tools Used:

Cisco Packet Tracer: network simulation
GNS3: network equipment emulation
Cisco Equipment: physical routers and switches
Wireshark: traffic capture and analysis
Cisco IOS: router/switch operating system

Resources:

Complete course handout (4AE-Interconnexion-Rzo)
Tutorials with practical exercises
3 guided labs with configurations
Cisco documentation
Protocol RFCs (TCP, IP, OSPF, BGP)

Study Methodology

Figure: 7-layer OSI model with protocols and encapsulation

Phase 1: Master the fundamentals: Review the layered model, IP addressing, and routing principles.

Phase 2: Deepen protocol knowledge: Study TCP, UDP, OSPF, and BGP in detail with their internal mechanisms.

Phase 3: Practice configuration: Use Packet Tracer to configure network scenarios of increasing complexity.

Phase 4: Analyze traffic: Capture with Wireshark to observe real protocol exchanges.

Phase 5: Develop troubleshooting skills: Diagnose network failures using a methodical approach (bottom-up or top-down through the OSI model).

Challenges Encountered

Protocol complexity: OSPF and BGP are feature-rich protocols with many parameters and edge cases.

Cisco IOS syntax: The Cisco command line has its own logic and specific commands to memorize.

Network debugging: Identifying the source of a problem in a complex network requires methodology and patience.

Abstraction vs. reality: Simulations are simplified. Real networks have physical constraints, latency, and random failures.

PART C - Detailed Technical Aspects

1. Network Reference Models

OSI Model (7 layers):

Theoretical reference model for network architecture.

Layer	Name	Role	Examples
7	Application	Application services	HTTP, FTP, DNS, SMTP
6	Presentation	Format, encryption, compression	SSL/TLS, JPEG, MPEG
5	Session	Session management	NetBIOS, RPC
4	Transport	End-to-end transmission	TCP, UDP
3	Network	Routing, logical addressing	IP, ICMP, OSPF, BGP
2	Data Link	Transmission over physical link, MAC addressing	Ethernet, WiFi, PPP
1	Physical	Bit transmission (electrical, optical signals)	Cables, radio waves

TCP/IP Model (4 layers):

Practical model used on the Internet.

Layer	OSI Equivalent	Main Protocols
Application	5-6-7	HTTP, FTP, DNS, SMTP, SSH
Transport	4	TCP, UDP
Internet	3	IP, ICMP, ARP
Network Access	1-2	Ethernet, WiFi

Encapsulation:

Each layer adds its header to the data from the layer above.

Application data → TCP Segment → IP Packet → Ethernet Frame

Decapsulation:

The receiver strips headers layer by layer to extract the data.

2. IPv4 Addressing

Format:

32-bit address, written as 4 decimal octets separated by dots.

Example: 192.168.1.10

Historical classes (obsolete):

Replaced by CIDR, but useful for understanding the evolution.

Class A: first bit = 0, network on 8 bits (large networks)
Class B: first bits = 10, network on 16 bits
Class C: first bits = 110, network on 24 bits

CIDR (Classless Inter-Domain Routing):

Flexible notation: IP address / network prefix length

Example: 192.168.10.0/24 means network on 24 bits, 8 bits for hosts (256 addresses).

Subnet mask:

Determines which part of the address is the network and which part is the host.

/24 corresponds to the mask 255.255.255.0

Special addresses:

Network address: all host bits set to 0 (e.g., 192.168.1.0)
Broadcast address: all host bits set to 1 (e.g., 192.168.1.255)
Loopback: 127.0.0.1 (local loopback interface)
Private ranges (RFC 1918): not routable on the Internet
- 10.0.0.0/8
- 172.16.0.0/12
- 192.168.0.0/16

Subnetting:

Dividing a network into smaller subnetworks to optimize address usage and segment the network.

Example: 192.168.1.0/24 divided into 4 /26 subnets:

192.168.1.0/26 (addresses .0 to .63)
192.168.1.64/26 (addresses .64 to .127)
192.168.1.128/26 (addresses .128 to .191)
192.168.1.192/26 (addresses .192 to .255)

3. IPv6

Motivation:

Exhaustion of public IPv4 addresses. IPv6 offers a virtually unlimited address space.

Format:

128-bit address, written as 8 groups of 4 hexadecimal digits separated by colons.

Example: 2001:0db8:85a3:0000:0000:8a2e:0370:7334

Notation shortcuts:

Omit leading zeros in each group: 2001:db8:85a3:0:0:8a2e:370:7334
Replace a sequence of consecutive zero groups with :: (only once): 2001:db8:85a3::8a2e:370:7334

Advantages:

Immense address space (2 to the power of 128 addresses)
Simplified autoconfiguration (SLAAC)
Native built-in IPsec
Simplified header (better performance)
No need for NAT

Address types:

Unicast: unique address identifying one interface
Multicast: group of interfaces (replaces broadcast)
Anycast: group of interfaces, routed to the nearest one

Special addresses:

Loopback: ::1
Link-local: fe80::/10 (auto-configured, local link communication)
Unique local: fc00::/7 (equivalent of IPv4 private addresses)
Global unicast: 2000::/3 (routable on the Internet)

4. Routing - Fundamental Principles

Role of the router:

Forward IP packets between different networks by consulting its routing table.

Routing table:

Contains routes to destination networks.

Each entry specifies:

Destination network with mask
Gateway (next hop): IP address of the next router or "directly connected"
Outgoing interface: which interface to send the packet through
Metric: route cost (distance, bandwidth, etc.)

Routing process:

Receive IP packet
Consult routing table to find the best route (longest prefix match)
Decrement TTL (Time To Live)
Recalculate IP header checksum
Forward packet on the appropriate outgoing interface

Default route:

Route 0.0.0.0/0 used when no specific route matches. Typically towards the Internet via the ISP gateway.

Static vs. dynamic routing:

Static: manually configured routes, suitable for small networks or specific routes
Dynamic: routing protocols automate route discovery and updates

5. Dynamic Routing Protocols

Classification:

IGP (Interior Gateway Protocol): within an autonomous system (AS)

Distance vector: RIP, EIGRP
Link state: OSPF, IS-IS

EGP (Exterior Gateway Protocol): between autonomous systems

Path vector: BGP

RIP (Routing Information Protocol):

Simple distance-vector routing protocol.

Characteristics:

Metric: hop count, maximum 15
Bellman-Ford algorithm
Full table update every 30 seconds
Slow convergence
Simple but limited to small networks

OSPF (Open Shortest Path First):

Advanced link-state protocol.

Principle:

Each router knows the complete topology of the area
SPF algorithm (Dijkstra) calculates the shortest path tree
Metric based on cost (function of bandwidth)

Operation:

Neighbor discovery via Hello packets
Exchange of LSAs (Link State Advertisements) describing links
Construction of the topological database (LSDB)
SPF calculation to determine best routes
Installation of routes in the routing table

Advantages:

Fast convergence (triggered updates, not periodic)
Scalability through hierarchy (areas)
Supports VLSM and CIDR
Update authentication

OSPF Hierarchy:

Organized into areas to reduce traffic and table size.

Area 0 (backbone): mandatory central area
Non-backbone areas: connected to area 0 via ABR (Area Border Router)

BGP (Border Gateway Protocol):

Internet routing protocol between autonomous systems.

Characteristics:

Path vector: includes the complete AS path
Complex routing policies (business, preferences)
eBGP (external) between different ASes, iBGP (internal) within an AS
Metric: not only technical but also policy-based

Role:

BGP maintains the global Internet routing table. Each AS announces its IP prefixes.

6. TCP (Transmission Control Protocol)

Characteristics:

Connection-oriented, reliable transport protocol.

Connection-oriented: 3-way handshake establishment before data exchange
Reliable: acknowledgments, retransmission of lost segments
Order preserved: sequence numbers guarantee ordering
Flow control: sliding window prevents receiver overflow
Congestion control: adapts throughput based on network state

TCP Header (main fields):

Source and destination ports (16 bits each): identify applications
Sequence number (32 bits): position of data in the stream
Acknowledgment number (32 bits): next expected byte
Flags: SYN (synchronization), ACK (acknowledgment), FIN (finish), RST (reset), PSH (push), URG (urgent)
Window (16 bits): available receiver buffer size (flow control)
Checksum (16 bits): error detection

Connection establishment (3-way handshake):

Client → Server: SYN (synchronization, initial sequence number)
Server → Client: SYN+ACK (acknowledgment + its own sequence number)
Client → Server: ACK (final acknowledgment)

Connection established, data exchange can begin.

Connection teardown (4-way handshake):

A → B: FIN (A has finished sending)
B → A: ACK (acknowledgment)
B → A: FIN (B has finished sending)
A → B: ACK (final acknowledgment)

Or abrupt closure with RST (reset).

Flow control (sliding window):

The receiver advertises the size of its available buffer (window).

The sender cannot send more data than the window size.

Prevents receiver buffer overflow.

Congestion control:

Mechanism to adapt throughput based on network state and prevent collapse.

Phases:

Slow Start: exponential increase of the congestion window
Congestion Avoidance: linear increase after reaching a threshold
Fast Retransmit: immediate retransmission after 3 duplicate ACKs
Fast Recovery: moderate window reduction after loss

Algorithms: TCP Tahoe, Reno, New Reno, CUBIC (used by Linux by default).

7. UDP (User Datagram Protocol)

Characteristics:

Simple, connectionless, unreliable transport protocol.

Connectionless: no connection establishment
Unreliable: no acknowledgment, no retransmission
No flow or congestion control
No ordering guarantee
Minimal header: only 8 bytes

UDP Header:

Source port (16 bits)
Destination port (16 bits)
Length (16 bits)
Checksum (16 bits)

Use cases:

Applications where speed takes priority over reliability:

VoIP and video conferencing: real-time, acceptable loss
Video/audio streaming: retransmission pointless (already too late)
DNS: short queries, retransmission handled by the application
DHCP: initial network configuration
Online gaming: critical latency
SNMP: simple network management

Advantages:

Low overhead, minimal latency, suited for real-time applications.

8. Quality of Service (QoS)

Problem statement:

The Internet is a "best-effort" network: no guarantees of bandwidth, latency, or reliability.

Insufficient for demanding applications: VoIP (voice), video conferencing, streaming.

QoS Parameters:

Parameter	Description	Important for
Bandwidth	Guaranteed throughput	Video streaming
Latency	Transit delay	VoIP, gaming
Jitter	Latency variation	VoIP, video
Packet loss	Rate of lost packets	All applications

QoS Mechanisms:

Classification and marking:

Identify traffic type and mark it in the header for differentiated treatment.

DSCP (Differentiated Services Code Point): 6 bits in the IP header
CoS (Class of Service): 3 bits in the 802.1Q tag (VLAN)

Queuing:

Manage priorities and packet scheduling.

FIFO: First In First Out (no priority)
Priority Queuing: queues with different priorities (strict)
WFQ (Weighted Fair Queuing): fair sharing with weighting
CBWFQ (Class-Based WFQ): WFQ per traffic class
LLQ (Low Latency Queuing): priority queue + WFQ for the rest

Policing and Shaping:

Control traffic rate.

Policing: limit throughput, drop or remark excess
Shaping: smooth traffic, buffer excess in a queue

QoS Models:

IntServ (Integrated Services): per-flow resource reservation (RSVP), rarely deployed (complexity)
DiffServ (Differentiated Services): aggregated service classes, scalable, widely deployed

9. VLANs (Virtual LANs)

Principle:

Segment a physical network into multiple isolated logical networks.

Advantages:

Broadcast domain isolation: reduces unnecessary traffic
Security: logical traffic separation (finance, HR, guests)
Flexibility: reorganization without physical recabling
Optimization: group users by function rather than by location

802.1Q Standard (VLAN Tagging):

Addition of a 4-byte tag in the Ethernet frame containing:

VLAN ID (12 bits): VLAN identifier (4096 VLANs possible)
Priority (3 bits): CoS priority for QoS

Port types:

Type	Description	Usage
Access	Belongs to a single VLAN, no tag	End-device connection
Trunk	Carries multiple VLANs, with tags	Switch interconnection

Inter-VLAN communication:

VLANs are isolated. To communicate between them, a router (or layer 3 switch) is needed.

Methods:

Router on a stick: single physical link, sub-interfaces per VLAN
L3 Switch: inter-VLAN routing directly on the switch

10. Switching

Principle:

A switch forwards Ethernet frames based on MAC addresses.

CAM Table (Content Addressable Memory):

Associates MAC addresses with switch ports.

Learning:

The switch learns MAC addresses by observing the source address of received frames.

Forwarding:

Known destination address: forward on the associated port in the CAM table
Unknown address: flood (broadcast) on all ports except the receiving one
Broadcast or multicast: flood

Spanning Tree Protocol (STP):

Protocol to prevent loops in redundant topologies (multiple links between switches).

Operation:

Election of a root bridge
Calculation of the optimal path to the root bridge
Blocking ports that create loops
Automatic reconfiguration in case of failure

Variants: RSTP (Rapid STP) for faster convergence, MSTP (Multiple STP) for multiple VLANs.

11. NAT (Network Address Translation)

Problem statement:

Shortage of public IPv4 addresses. Not all machines can have a public address.

Principle:

Translate private IP addresses to public addresses when crossing a router.

Types of NAT:

Static NAT:

Fixed 1:1 mapping between a private address and a public address.

Usage: servers accessible from the Internet.

Dynamic NAT:

Pool of public addresses, dynamic allocation to internal machines.

PAT (Port Address Translation) or NAT Overload:

Multiple internal machines share a single public address.

Differentiation by port numbers.

NAT table associates: private IP:private port ↔ public IP:public port

Usage: most home routers (all domestic devices share the public IP).

NAT Limitations:

Breaks end-to-end connectivity
Issues for protocols embedding IP addresses in data (FTP, SIP)
Difficulties for P2P applications and inbound servers
Requires traversal techniques (STUN, TURN) for VoIP

IPv6 and NAT:

IPv6 makes NAT unnecessary (sufficient addresses for all). Return to end-to-end connectivity.

12. Network Security

Access Control Lists (ACL):

Filter traffic based on defined criteria (IP addresses, ports, protocols).

Types:

Standard ACL: filters only on source IP address
Extended ACL: filters on source/destination IP, ports, protocols

Application: on router interfaces (inbound or outbound traffic).

Firewall:

Device or software that filters traffic in a more sophisticated manner.

Features:

Stateful filtering (connection state inspection)
Application filtering (deep packet inspection)
Intrusion prevention (IPS)
Logging and alerts

VPN (Virtual Private Network):

Create a secure tunnel through an unsecured network (Internet).

Types:

Site-to-Site: interconnect remote sites of a company
Remote Access: enable secure remote access (teleworking)

Protocols:

IPsec: IP-level encryption (tunnel or transport mode)
SSL/TLS VPN: HTTPS-based, simpler for remote access

Common threats:

Threat	Description	Countermeasure
DoS/DDoS	Resource saturation	Filtering, rate limiting, CDN
Spoofing	IP address impersonation	Ingress/egress filtering, authentication
Man-in-the-Middle	Communication interception	Encryption (TLS, IPsec)
ARP poisoning	ARP table corruption	Dynamic ARP Inspection
VLAN hopping	Unauthorized VLAN access	Secure trunk/access configuration

Best practices:

Defense in depth (multiple security layers)
Principle of least privilege
Network segmentation (VLANs, DMZ zones)
Regular updates (firmware, patches)
Monitoring and detection (IDS, SIEM)
Configuration backups

PART D - Reflective Analysis and Perspectives

Skills Acquired

In-depth TCP/IP understanding: Mastery of the fundamental Internet protocols and their internal operation.

Network equipment configuration: Ability to configure Cisco routers and switches via the command line interface (CLI).

Network architecture design: Ability to design multi-site infrastructures with routing, VLANs, and QoS.

Methodical troubleshooting: Competence in diagnosing and resolving network issues following the layered model.

Infrastructure security: Ability to implement ACLs, VPN, and segmentation to secure a network.

Key Takeaways

1. The layered model is fundamental: Understanding encapsulation and the role of each layer enables effective problem diagnosis.

2. OSPF is the modern IGP protocol: RIP is obsolete for networks of significant size. OSPF offers scalability and fast convergence.

3. TCP provides reliability, UDP provides speed: Choose the right transport protocol according to the application's needs.

4. QoS is crucial for real-time: VoIP and video require priority and guarantees to function properly.

5. VLANs = effective logical segmentation: Organizing the network into VLANs improves security, performance, and management.

Feedback

Educational aspect: The labs on Cisco equipment gave a concrete dimension to the theory. Seeing routes appear in the table after configuring OSPF is satisfying.

Cisco IOS complexity: The Cisco CLI has a learning curve. The modes (user, privileged, configuration) and command hierarchy require practice.

Importance of troubleshooting: Configurations do not always work on the first try. Learning to diagnose methodically (ping, traceroute, show commands) is essential.

Robust protocols: TCP and OSPF are engineering marvels. Their robustness in the face of failures and congestion is impressive.

Practical Applications

For network administrators:

Deploy and maintain enterprise network infrastructures
Configure routers and switches
Implement redundancy and high availability
Ensure security and QoS

For network architects:

Design multi-site architectures
Size bandwidth
Choose equipment and protocols
Plan evolution and scalability

For DevOps engineers:

Understand cloud networks (VPC, subnets, routing tables)
Automate network configurations (Ansible, Terraform)
Optimize distributed application performance
Implement load balancers

For developers:

Understand network latency and its impact on applications
Choose appropriate protocols (TCP vs UDP)
Implement retry logic and timeouts
Optimize for WAN (compression, caching)

Limitations and Openings

Module limitations:

Focus on Cisco (market is more diverse: Juniper, Arista, etc.)
Little on IPv6 (inevitable transition)
No SDN (Software-Defined Networking)
Automation barely covered

Openings towards:

SDN and NFV: network virtualization, OpenFlow, controllers
Cloud networking: AWS VPC, Azure Virtual Networks
Automation: Ansible, Python with Netmiko/NAPALM
Monitoring: SNMP, NetFlow, observability
5G and mobile networks: architecture, slicing

Recent Developments

Software-Defined Networking (SDN):

Separation of the control plane (routing decisions) from the data plane (forwarding).

Centralized controller programmable via API.

Network Functions Virtualization (NFV):

Network functions (firewall, load balancer) in software rather than dedicated hardware.

Flexibility and cost reduction.

Intent-Based Networking:

Define the intent (policy) rather than detailed configuration.

The system automatically translates intent into configuration.

Automation and Infrastructure as Code:

Network configuration declared in files (Ansible, Terraform).

Versioning, reproducibility, scalability.

5G Networks:

Ultra-low latency, high throughput, slicing (dedicated virtual networks per use case).

Tips for Success

1. Practice regularly: Use Packet Tracer or GNS3 to reproduce scenarios and experiment with configurations.

2. Understand, do not memorize: Understand how protocols work rather than memorizing CLI syntax.

3. Document your configurations: Take notes on commands and deployed architectures. Build a personal reference.

4. Use Wireshark: Capture and analyze real traffic to see protocols in action.

5. Follow the troubleshooting methodology: Bottom-up (layer 1 → 7) or top-down depending on symptoms. Methodical and systematic.

6. Get certified: Cisco CCNA is a recognized certification validating basic networking skills.

Conclusion

This module is essential for any engineer working with connected systems. Networks are ubiquitous: enterprises, data centers, cloud, IoT, telecommunications.

Transferable skills:

Global vision of network architectures
Ability to diagnose and resolve complex problems
Understanding of tradeoffs (performance, security, cost)
Mastery of universal protocols (TCP/IP)

Professional relevance: Networking skills are sought after in all sectors. The explosion of cloud, IoT, and distributed applications increases the demand for qualified network engineers.

Key message: The Internet works thanks to robust protocols and competent engineers. Understanding these fundamentals enables the design, deployment, and maintenance of reliable and high-performance infrastructures.

Recommendations:

Continue practicing (home lab with old routers or VMs)
Train on automation (Python for networks)
Explore SDN (OpenDaylight, ONOS controllers)
Stay current (RFCs, technical blogs, conferences)
Consider certifications (CCNA, CCNP, or equivalents from other vendors)

Links with other courses:

Networking - S5: TCP/IP foundations
Cloud and Edge Computing - S9: virtual networks
Embedded IA for IoT - S9: object connectivity
Real-Time Systems - S8: temporal constraints

Course taken in 2023-2024 at INSA Toulouse, Department of Electrical and Computer Engineering.

← Retour aux Cours 2023-2024

Documents de Cours Course Documents

Cours Complet Interconnexion Reseaux Complete Network Interconnection Course

Cours complet sur les architectures reseaux : routage, protocoles (RIP, OSPF, BGP), VLAN, qualite de service. Complete course on network architectures: routing, protocols (RIP, OSPF, BGP), VLAN, quality of service.

Telecharger Download

Travaux Diriges Tutorials

TD : exercices de configuration routeurs, analyse de trames, deploiement VLAN et depannage reseau. Tutorials: router configuration exercises, frame analysis, VLAN deployment, and network troubleshooting.

Telecharger Download