Networking Essentials: TCP, IP, Routing, and SDN

1. TCP Congestion Control

Slow Start: TCP initializes congestion window (cwnd) to a small size and doubles it every Round Trip Time (RTT) until a loss occurs (exponential growth) or the slow start threshold is reached.

Congestion Avoidance: After reaching the threshold, cwnd is incremented by one segment for each RTT, growing linearly to avoid congestion.

Fast Recovery: When triple duplicate ACKs are detected, TCP halves the cwnd and reduces the slow start threshold to half of the cwnd size before the loss, then transitions to congestion avoidance.

2. TCP RTT and Timeout

Estimating RTT: Uses Smoothed Round Trip Time (SRTT) which is a weighted average of previous RTT measurements.

Timeout Calculation: Often calculated as Timeout=SRTT+4×RTT Variation, helping to dynamically adjust to network conditions.

3. Network Layer

Data Plane: Physically responsible for forwarding packets based on the routing table, using algorithms implemented in hardware.

Control Plane: Logically separate in SDN, managing how packets are routed by configuring the data plane.

4. Router

Functions as a node that forwards packets between different networks based on their destination IP address, using a routing table.

5. IP (Internet Protocol)

Datagram and Fragmentation: IP treats each packet independently, and large packets are fragmented if they exceed the MTU of any network they traverse.

Addressing and Subnetting: Each device on an IP network is assigned an IP address and a subnet mask, which determines the network portion and the host portion of the address.

IPv4 & IPv6: IPv4 addresses are 32 bits long, whereas IPv6 addresses are 128 bits long, greatly increasing address space and supporting features like autoconfiguration.

6. Forwarding and SDN

Generalized Forwarding: Enables flexible routing decisions based on multiple fields in the IP header, implemented in SDN via programmable flow tables.

SDN: Separates the control logic from the network hardware (control and data planes), allowing centralized management of network behavior.

7. Routing Protocols

Link State: Uses local information to build a complete map of the network topology to route packets.

Distance Vector: Each router shares its view of the network with neighbors without a complete picture of the network structure.

8. OSPF and BGP

OSPF: An interior gateway protocol that uses a link-state routing algorithm, effective within a single AS.

BGP: An exterior gateway protocol used for routing between autonomous systems, crucial for Internet routing.

9. SDN Control Plane

Centralizes network decision-making, allowing easier implementation of policies and improved network responsiveness and flexibility.

10. Link Layer Protocols

Handles framing, addressing, and error detection/correction on direct link connections.

11. Error Detection/Correction

Uses methods like CRC and parity checks to ensure data integrity over transmission.

12. MAC Protocols

Defines how devices uniquely address one another at the data link layer and manage access to the shared medium.

13. Random Access Protocols

• Slotted & Pure ALOHA: Manage shared medium use without a schedule.
• CSMA/CD: Used in traditional Ethernet to detect and handle collisions.

14. LANs

MAC Addressing and ARP: ARP translates IP addresses to physical MAC addresses.

Ethernet Switch: Connects devices within a LAN, using MAC addresses to forward data at the data link layer.

15. Wireless Networking

Involves communication over radio waves, including standards like IEEE 802.11 (Wi-Fi), challenges like interference and attenuation, and security considerations.