Computer Networks & Data Communication Essentials

Explore how application layer protocols power popular networking apps like Google search, Gmail, Facebook, Skype, YouTube, and TikTok, designed to run on end systems that communicate over the internet.

Explore client-server architecture where servers in data centers use permanent IPs. Clients connect intermittently with dynamic IP addresses and communicate via application-layer protocols such as SMTP, HTTP, and FTP.

Describe peer-to-peer architecture where end devices act as peers, connect intermittently with changing IP addresses, and manage service requests without an always-on server. BitTorrent exemplifies this self-scaling, direct sharing model.

Examine how end devices use application layer protocols for services: smtp for email, http for web, ftp for file transfer, sip for ip telephony, and rtp for streaming.

Different application layer protocols demand distinct transport services by data loss, throughput, and time sensitivity; ftp and smtp require no data loss, while rtp needs low delay for audio/video throughput.

See how processes on separate hosts communicate via sockets, using IP addresses and ports, with TCP or UDP transport; include client-server and peer-to-peer models and protocols like HTTP and SMTP.

Compare tcp and udp transport layer protocols, noting that tcp is connection-oriented with a handshake, reliability, and flow and congestion control, while udp is connectionless and lightweight.

Explore how application layer protocols rely on tcp or udp at transport, with ftp, smtp, and http using tcp for reliability and real time protocol using udp for media.

Explore the http protocol, a client–server web model used by browsers to request pages; servers respond with html and referenced objects via urls that include host names and paths.

A client browser requests web objects from a server using HTTP, the server responds, and the browser renders the page; HTTP runs over TCP (port 80) and is stateless.

Explain non-persistent HTTP, introduced in HTTP 1.0, where each object uses a separate TCP connection, and persistent HTTP, in HTTP 1.1, where multiple objects download over a single connection.

The lecture explains the non-persistent http protocol by tracing a web page with eleven objects: one master html file and ten jpeg images, downloaded via tcp connections to port 80.

Explain how non-persistent HTTP downloads require two RTT plus the object transmission time, including TCP connection setup, HTTP request and response, and the object transfer.

Persistent http halves the response time versus non-persistent http by reusing a single tcp connection for object downloads, avoiding two rtts per object and enabling parallel transfers.

Explore the general ascii format of an http request message, with the request line, headers like host and user-agent, and the get method’s empty body for retrieving a resource.

Explore http request methods: post sends form data in the body, get uses the url for queries, head returns only headers, and put or delete manage server objects.

Explore the general format of an http response, including the status line, headers (date, server, content-length, content-type, last-modified), the blank line, and the response body.

Explain http response status codes such as 200 ok, 301 moved permanently with a location header, and 400, 404, and 505 not supported by the server.

Use cookies to maintain the client and server state across multi-step http transactions, such as shopping cart, since http is stateless. Resume the transaction after a crash to continue.

Explore how cookies power shopping carts, restore session state for email composition, enable personalized product recommendations, and control user authorization by tracking browsing patterns.

Demonstrates how cookies maintain client and server state across visits by assigning a cookie ID, storing it in the browser, and using it to personalize recommendations and pre-fill carts.

Explore how the browser loads the master html file, fetches an ad from the third-party addicts.com server without cookies, and Dexcom serves a random ad.

Explore how third party cookies track a user's browsing behavior across sites to display targeted ads, using first party and third party cookies, referrers, and ad servers.

Understand how cookies raise privacy concerns by tracking user behavior, including first party and third party cookies, and how browsers like Firefox, Safari, and Google plan to disable them.

The web cache, a proxy between the client and origin server, stores recently requested objects and serves them quickly, fetching from the origin on a miss and respecting cache-control headers.

Learn how conditional Get s update web caches by using the If-Modified-Since header; when unmodified, servers return 304 with no body, and when modified, 200 with the new object.

Explore the file transfer protocol in its client-server model, enabling uploading and downloading between a local client and a remote host, via FTP interfaces like FileZilla and RFC 959.

The ftp protocol uses separate control and data connections, with control on port 21 for authentication and directory browsing, and a data connection on port 20 for file transfers.

The ftp protocol uses a command-response model over the control connection, with ascii commands like user, pass, retr, and stor for authentication, listing, and transfers, with status codes and phrases.

Examine the simple mail transfer protocol and its role in delivery. Identify the three components: user agents, mail servers, and smtp protocol, and their roles in a client-server email system.

Alice sends an email to Bob via a mail server using SMTP over TCP, delivering to Bob’s mailbox, where Bob retrieves it with IMAP or HTTP.

Explore the email message format and the SMTP protocol, including RFC 5321 and RFC 282, detailing the header and body structure, blank line separation, and ASCII content.

Describe the smtp protocol per RFC 5321, an application-layer protocol using tcp reliability. It exchanges commands and responses over port 25 with a three-phase sequence: handshake, transfer messages, and closure.

Explore an SMTP client–server interaction over TCP, detailing the handshake, the mail from and rcpt to commands, the data transfer, and the closure with 221, 250, and 354.

Explore mail access protocols such as pop, imap, and http, and learn how smtp transfers emails between servers and how users manage their mailboxes.

Learn how the domain name system translates domain names to IP addresses and back, acting as the internet’s phonebook. Explore the DNS hierarchy and its application layer protocol for queries.

Explain how the DNS resolves a domain name through root, TLD, and authoritative servers, using mail.amazon.com as an example to convert to an IP address.

Discover how root DNS servers, top-level domain servers, and authoritative servers enable domain name resolution, with ICANN oversight, and organization-specific DNS for domains like yahoo.com and amazon.com.

learn how a local dns server queries root, .edu top-level, and university authoritative servers in an iterative process to resolve a domain name to its ip address.

Illustrates how a recursive dns query flows from the local server to the root, then to the edu top-level domain and University of Massachusetts authoritative server, highlighting the root's load.

Learn how a local DNS server caches domain name to IP mappings to speed responses. Explore TTL expiry, and how caching reduces root DNS server load.

BitTorrent enables peer to peer file sharing by distributing large files among peers instead of a client server model, using torrent metadata and trackers to coordinate chunks.

Explore how BitTorrent uses a tit-for-tat upload strategy, choked and unchoked peers, and optimistic unchoking to maximize file download speed.

Explore how the transport layer enables end-to-end communication between processes on different hosts by breaking messages into segments, through lower layers, and reassembling at destination, implemented only in end hosts.

The transport layer uses the network layer's services to deliver host-to-host communication, using sockets, segmentation, and headers (TCP/UDP) to form IP packets routed over the internet.

Understand the two principal transport layer protocols, tcp and udp, and learn tcp’s reliability, order, congestion and flow control, along with udp’s connectionless, best-effort delivery.

The transport layer performs demultiplexing and multiplexing, routing segments from p3 to p1 and p4 to p2 using headers, then multiplexing messages from p1 and p2.

The lecture shows how transport layer de-multiplexing uses the destination IP and port to deliver a segment to the correct process via the socket address.

Shows UDP demultiplexing: packets with the same destination IP and port go to the same socket, regardless of source ports, and replies reverse source and destination ports.

Tcp demultiplexes incoming segments to distinct sockets based on source ip or source port when destination ip and port are identical.

Explain how UDP is a connectionless protocol with no handshaking, an eight-byte header, no acknowledgments, and out-of-order delivery, used for speed and real-time audio or video with no congestion control.

Explore UDP use cases for online gaming, streaming, and voice over IP, highlighting low latency and tolerance for packet loss, and note DNS resolves to IPs and DHCP assigns IPs.

Explain the UDP segment format, detailing the header fields—source port, destination port, length, and checksum—and the payload that carries application data. Observe that each header field is 16 bits.

Compute the UDP checksum by summing 16-bit words of the segment (excluding the checksum field) and taking the one's complement. On the receiver, recompute and compare to detect errors.

Explore the tcp overview: a point-to-point, connection-oriented protocol that establishes a connection via handshaking, enabling reliable, in-order delivery and bidirectional exchange, with flow and congestion control, and maximum segment size.

Explore how tcp uses sequence numbers and acknowledgments to ensure reliable, in-order data delivery by segmenting a file into 1000-byte blocks starting at zero and advancing with each ack.

Demonstrate bidirectional data transfer over TCP using a telnet scenario where client A and server B exchange segments, acknowledgments, and sequence numbers to echo typed characters.

The tcp layer provides reliable delivery over ip by using a timeout-based retransmission and, with three duplicate acknowledgments, quickly retransmits the missing segment before the timer expires.

Examine TCP retransmission scenarios with premature timeouts on back-to-back segments (seq 92 and 100) and how cumulative acknowledgments and lost acks influence resends.

Explore tcp segment structure, detailing header fields, source and destination ports, sequence and acknowledgement numbers, checksum, receive window, and flags, plus maximum segment size negotiated during handshake via options.

Explore TCP flow control, where the receiving buffer's receive window guides the transmitter to limit unacknowledged data and prevent overflow.

Explore how tcp uses end-to-end congestion control without network feedback, adjusting the congestion window based on segment losses and ack delays to keep unacknowledged data within cw.

During slow start, begin with a congestion window of 1 MSS and double it after each round trip, reaching 2, 4, and 8 MSS, until a segment loss stops growth.

explains tcp tahoe congestion control, showing slow start doubling cwnd to 16 mss, loss resets cwnd to one mss and ssthresh to 8, then congestion avoidance with linear growth.

Explore the fast recovery phase in tcp Reno, contrasting with tcp Tahoe, where three duplicate acks halve cwnd and then increment it by one mss per round trip.

Describe how a TCP connection forms via a three-way handshake, exchanging syn and ack messages with sequence numbers to agree on connection parameters and data readiness.

Close a tcp connection from both sides by sending a message with the thin bit set, using sequence and acknowledgement numbers, and finish in three messages with simultaneous thin exchange.

The network layer handles end-to-end IP packet routing from source host to destination host, encapsulates transport layer segments into IP datagrams with headers, and routes via routers using IP addresses.

Explore how routers perform routing and forwarding, using a local forwarding table to forward IP packets to the correct interface based on destination addresses, guided by a routing algorithm.

Aggregate IP addresses into ranges to keep routing tables manageable and forward packets using the longest prefix matching to the most specific interface.

Explain how IP addresses are assigned in data networks. See how hosts connect via Ethernet switches or Wi-Fi base stations to a router in a local area network.

Identify interfaces as the physical connections between hosts and the data network, including wired and wireless options, with each interface using a 32-bit IPv4 address in dotted decimal notation.

Explore how subnets form isolated networks, identify a /24 prefix as the subnet address, and explain when a router is needed for inter-subnet communication.

Explore a data network example with six subnets, including 223.1.1.0/24 and a point-to-point link subnet 223.1.9.0/24.

ISPs obtain a /20 address block from ICANN, allocate it to eight organizations, and extend to a /23 to provide two to nine IP addresses per organization.

Explore how hierarchical subnet addressing enables route aggregation and efficient routing, using ISPs, routing tables, and longest prefix match to forward packets to the correct organizations.

Explore classless inter-domain routing (CIDR) and subnetting, including how increasing the subnet part subdivides blocks, enables route aggregation, and improves IP address efficiency over classful addressing.

Learn how hosts obtain IP addresses either by hard coding or via dynamic DHCP from an ISP, enabling plug-and-play connectivity and IP reuse for mobile users.

Explains why network address translation is needed to address IPv4 address exhaustion, noting that 32-bit IPv4 addresses yield about 4 billion possibilities while 15.14 billion IoT devices connected in 2023.

Explain how private IP addresses in local networks map to a single public IP via NAT, letting the router translate internal hosts to external destinations and return traffic.

Learn how a nat scenario translates a private host's ip address and port to a public ip address and port at the gateway router, using a translation table.

Network address translation uses private IPs and a single public IP, reducing public IP usage. It enables easy local IP changes and, by hiding private addresses, improves security.

Explore NAT traversal problem where a private IP client connects to a server behind a gateway via a public IP. Skype relay bridges two clients for a bidirectional connection.

Learn how routers forward packets via routing tables and choose paths with routing algorithms, and compare static routing for small networks to dynamic routing for scalable, auto-updated tables.

Explore how the internet is built from autonomous systems interconnected by edge routers, with interior gateway protocols inside each AS and BGP as the inter-AS routing protocol.

present interior gateway routing protocols, distance vector and link state, with rip and ospf examples, and show how hello messages, neighbor link-state, Bellman-Ford refinement, and hop-count metrics shape routing tables.

Builds link state information by exchanging hello messages, floods it to all routers to form a network topology, and uses the Gastre algorithm to determine shortest paths for routing tables.

Explain why inter autonomous system routing protocols, also called exterior gateway protocols, are needed to build routing tables and propagate reachability between autonomous systems.

an inter-autonomous system routing example shows how reachability to subnet x is learned in as1 and inserted into the forwarding table to send packets via the least-cost path to as3.

Describe how inter autonomous system routing uses hot potato routing to send a packet via closest border gateway router, choosing the least cost link to autonomous system three or two.

Explore the Border Gateway Protocol (BGP) for inter autonomous system routing on the internet. Learn how EBGP and IBGP exchange reachability advertisements over semi-permanent TCP connections to shape routing table.

Learn how ICMP provides error reporting and diagnostic messages within IP packets. Understand the two main ICMP types: error reporting and query messages used by administrators.

Understand how time exceeded ICMP signals routing loops as TTL drops at routers, discarding the packet and returning an ICMP error with the router’s name and IP.

Traceroute serves as a diagnostic tool to trace the route of an IP packet from the source host to the destination host using ICMP time exceeded messages and TTL values.

Learn how the ping tool uses the ICMP echo and echo reply messages to test connectivity between hosts, measure round-trip times, and verify reachability with examples like pinging google.com.

Explains how IPv6 addresses solve IPv4 scarcity by offering 2^128 addresses, formatted as eight 16-bit hexadecimal groups separated by colons, with leading zeros dropped and zeros compressed using double colon.

Explore the role of the link layer in moving frames between adjacent nodes, from source to destination via routers, while the network layer handles end-to-end IP routing.

Explain why the link layer matters and why different link layer protocols fit different physical links, such as wifi for wireless and ethernet for wired, with reliability emphasized for wireless.

Differentiate router from switch: the router is a layer three device routing ip packets, while the switch is a layer two device forwarding frames within local network using Mac addresses.

Provide framing by adding a header to IP packets to form frames. Manage link access for shared media using MAC addresses and ensure reliability on wireless links.

Explore how the link layer provides flow control to prevent receiver overload, and how error detection and correction in frame headers support reliable delivery and reduce retransmissions to save bandwidth.

Understand how a network interface card implements the link layer inside a host, adding headers to IP packets and using MAC addresses to enable adjacent host communication.

Identify the two main link access protocols: point-to-point protocols for dedicated links and multiple access protocols for shared media, such as Ethernet, Wi-Fi, satellite, and mobile networks.

This lecture explains channelization protocols, detailing how bandwidth is divided into frequency channels (FDMA), time slots (TDMA), or codes (CDMA) to allocate dedicated channels.

Demonstrates how frequency division multiple access partitions bandwidth into bands and channels and assigns channels to mobile stations, while time division multiple access uses time slots for uplink and downlink.

Explore random access protocols used in local area networks and wireless LANs, where no station dominates others and collisions occur, including Aloha, Slotted Aloha, CSMA, and CSMA/CD.

Explore the Aloha protocol, a random access method for shared media, detailing collision handling, acknowledgments, retransmissions, and the equal-length frame assumption in wireless LANs.

Explore the pros and cons of the pure Aloha protocol, including its lack of synchronization and easy implementation, alongside its high collision probability and low efficiency of about 18%.

The slotted aloha protocol uses synchronized time slots and transmits only at slot starts. It reduces collision windows to 36.8% efficiency, but channel sensing can further improve performance.

The CSMA/CD protocol reduces collisions on a shared wire LAN by listening before transmitting and deferring when the channel is busy; collision detection then stops transmission to minimize wasted time.

Explain how the Ethernet csma/cd algorithm coordinates access to a shared wired LAN by sensing the channel, handling collisions with a jam signal, and using binary exponential backoff to retry.

Explore how controlled access (taking turns) protocols blend channel partitioning and random access methods to achieve efficiency across low and high network loads.

Explore polling-based centralized access protocols where a central controller polls stations in a round-robin sequence to share the channel efficiently. Note overhead, latency, and single-point-failure risks; Bluetooth uses polling.

Explore token passing controlled access protocols, where only the token-holding station may transmit data, and understand overhead, delays, single point of failure risks, and recovery procedures.

Compare IP addresses and MAC addresses: IPs are 32-bit for routing at the network layer; MACs are 48-bit, burned in the NIC ROM, for forwarding at the link layer.

Explain how a MAC address remains unique by combining the identifier assigned to manufacturers with a network interface card specific identifier, and show its portability across LANs, unlike IP addresses.

Explain how arp maps ip addresses to mac addresses on a local area network, building arp tables with time to live values that refresh every 20 minutes.

Demonstrate how the arp protocol resolves a host's mac address from its IP address by sending an arp query, receiving an arp reply, updating the arp table with IP–MAC mappings.

Explain how an IP packet routes from one subnet to another through a gateway router, including DHCP or manual IP assignment and ARP to obtain MAC addresses.

Explore Ethernet, a link layer protocol and wired LAN technology, evolving from 10 Mbps to 400 Gbps. Learn how modern switched topologies using layer two switches prevent collisions.

Explore how Ethernet frames carry IP payloads between hosts, starting with the preamble for synchronization, then destination and source MAC addresses, the protocol type, and cyclic redundancy check bits.

Ethernet operates as a connectionless, unreliable protocol that transmits frames without handshaking or acknowledgments; reliability comes from higher layers as TCP, with unslotted carrier sense multiple access with collision detection.

Understand how an Ethernet switch buffers and forwards frames by destination MAC address, handles broadcasts, and provides plug-and-play, self-learning, full-duplex links with no collisions.

Learn how switches self-learn forwarding tables by observing source MAC addresses in frames, mapping each MAC to its connected interface, and using a time to live for entries.

Explain the switch working algorithm with two interconnected switches and three computers on sides, showing how a frame's source Mac address is learned and how destination Mac address guides forwarding.

An example shows how a switch self-learns its forwarding table and forwards frames by flooding unknown destinations, then updates entries to unicast to the correct interface.

Explore how interconnected switches learn forwarding tables through broadcast frames. That enables path-specific forwarding between computers A and G as entries populate across S1, S3, and S4.

Virtual LANs reduce broadcast traffic across switches, improving network efficiency while enhancing security and privacy by isolating departmental traffic and enabling logical connectivity despite physical moves.

Explore how to connect VLANs across multiple switches and why simple one-to-one port connections across buildings are not scalable for multiple virtual LANs.

Explore VLAN trunking by declaring trunk ports and establishing a trunk link that carries electrical engineering and computer science VLANs, using 802.1Q frames with VLAN ID and priority.

Explain how IEEE 802.3 defines the link and physical layers of Ethernet, maintaining a common link layer while media-dependent physical layers yield speeds from 10 Mbps to 400 Gbps.

Explore common ethernet implementations: 10base5 thick coaxial (802.3) at 10 Mbps over 500 m, 10base2 thin coaxial (185 m) with BNC, 10baseT unshielded twisted pair, and 10baseF optical fiber.

Explain the physical layer of the ten base T standard, highlighting twisted pair media, unshielded and shielded cables, and the hub-based broadcast LAN with RJ-45 connections.

Explain the physical layer of ten base f, where optical fiber carries light through a core and cladding with refractive indices, linking each computer to a hub via st connectors.

Compare fast ethernet and gigabit ethernet implementations, detailing 100base-tx, 100base-fx, 100base-t4 with cat5 utp or cat3 utp copper, and 1000base-sx, lx, c, and t fiber or copper cabling.

Explore how a firewall, combining hardware and software, isolates a private network from the public internet, enforces a local security policy to permit authorized traffic and block unauthorized traffic.

Protect networks by using firewalls to prevent denial of service attacks and ensure only authorized users and traffic access the network by filtering bogus TCP SYN messages.

Explore stateless packet filtering firewalls that inspect every packet at the access router and block UDP traffic and telnet by IP addresses and ports.

Explore how firewall rules enforce a network security policy through packet filtering, blocking port 80 web traffic, and preventing smurf ddos, trace root command, and unauthorized ICMP and UDP access.

Explore how access control lists in the firewall implement stateless packet filtering by permitting inside-to-outside HTTP traffic on port 80 and DNS on port 53, while blocking other inbound traffic.

Stateless packet filtering makes firewall decisions per packet, allowing bogus tcp messages that meet acl rules and enabling denial of service from public hosts to a private server.

Explore how stateful packet filtering firewalls track active TCP connections using a connection table, validate passing packets, and drop invalid ones, with inactivity timeouts removing idle entries.

Implement stateful packet filtering inside a firewall's access control list by adding a connection-check column, and inspect incoming TCP and UDP traffic against the connection table to drop non-matching packets.

Explain how application gateway firewalls, paired with a packet filter, operate at the application layer to relay authenticated Telnet sessions between private networks and the public internet.