CCN-w6

Department Overview

Institution:

Department of Computer Science and Engineering, University of KURDISTAN Hewler

Subject:

Computer Communication and Networks

Week:

6

Network Layer Design Issues

Key Design Aspects:

  1. Store-and-Forward Packet Switching:

    • Involves receiving the entire packet before forwarding it, allowing for error checking and reducing the likelihood of corrupted data being sent.

  2. Services Provided to the Transport Layer:

    • Ensures reliable communication by establishing clear protocols for data transfer, thereby maintaining data integrity and order.

  3. Implementation of Connections:

    • Connectionless Service (Datagram Subnets):

      • Ideal for applications where speed is preferred over reliability.

    • Connection-Oriented Service (Virtual-Circuit):

      • Provides a reliable path ensuring all packets follow the same route, useful for real-time applications such as video conferencing.

  4. Comparison between Datagram Subnets and Virtual-Circuit:

    • Illustrates the trade-offs between flexibility and reliability in data transmission methods.

Store-and-Forward Packet Switching:

Process:

  1. Host transmits packet to the nearest router.

  2. Packet is temporarily stored until fully received to verify integrity.

  3. Each link verifies the checksum before forwarding to the next router, ensuring reliable transmission until the packet reaches its destination.

Connectionless Service Implementation

Characteristics:

  • Packets are referred to as datagrams, and the subnet is recognized as a datagram subnet.

  • Each router maintains a dynamically updated table to determine the best route for reaching any node within the network. This adaptability enhances network resilience and efficiency.

Connection-Oriented Service Implementation

Virtual-Circuit Characteristics:

  • A dedicated path must be established from the source to destination before packet transmission begins, ensuring that all packets follow a predetermined route.

  • The established path is often referred to as Virtual Circuit (VC) within a virtual-circuit subnet, allowing for enhanced reliability during transmission.

Comparison of Subnet Types

Datagram Subnet:

  • Advantages:

    • No circuit setup needed, allowing for faster transmissions.

    • Each packet contains the full source and destination address, enabling independent routing.

    • Robustness against router failures, as only packets specifically routed through failing routers are lost.

  • Disadvantages:

    • More challenging to manage Quality of Service (QoS), as there is no guaranteed bandwidth or latency in packet delivery.

Virtual-Circuit Subnet:

  • Advantages:

    • Requires a circuit setup but offers a reliable method of transmission with allocated resources.

    • Uses shorter VC numbers in packets, reducing data overhead.

  • Disadvantages:

    • Requires routers to hold state information about each VC, adding complexity to router management.

    • Complete termination of VCs occurs during router failures, potentially impacting real-time data streams.

Services to the Transport Layer

Key Objectives:

  • Emphasizes independence from underlying router technology, providing flexibility for changes in hardware and software.

  • Shields the transport layer from the complexities of router numbers, types, and topologies, promoting simplified network management.

  • Promotes uniform network addressing across both Local Area Networks (LANs) and Wide Area Networks (WANs), facilitating seamless communication.

Debate:

  • Connectionless Service:

    • Proponents argue that network reliability should not be managed at the network layer and should instead be handled by host systems, encouraging rapid transmissions.

  • Connection-Oriented Service:

    • Advocates stress the essential nature of QoS management, particularly for real-time data such as audiovisual communications, where reliability is critical.

Routing Algorithms Overview

Function:

  • Facilitate the efficient routing of packets from their source to their intended destination.

Routing Algorithm Role:

  • Determines the most appropriate output line for an incoming packet based on current network conditions.

Properties of Routing Algorithms

Desirable Properties:

  • Correctness: Accurate delivery of packets.

  • Simplicity: Ease of understanding and implementation.

  • Robustness: The ability to withstand and adapt to errors or changes.

  • Stability: Consistent performance over time despite changes in the network.

  • Fairness: Equitable treatment of all packets.

  • Optimality: Minimizing throughput and delay in packet delivery.

Algorithm Classification:

  • Non-Adaptive:

    • Employs static routing with precomputed routes, suitable for stable traffic environments.

  • Adaptive:

    • Utilizes dynamic routing that adjusts decisions based on real-time traffic and network topology changes.

The Optimality Principle

Concept:

  • Aims to minimize packet delay while maximizing throughput, crucial for maintaining network efficiency.

  • This results in a tree structure of optimal routes from source to destination, known as a sink tree.

Shortest Path Routing (Dijkstra's Algorithm)

Mechanism:

  • Constructs a graph where routers are represented as nodes and links as arcs.

  • Determines the shortest path using various metrics, such as hops and queuing delays, ensuring efficient packet delivery.

  • Illustrated with comprehensive examples highlighting node interactions and path calculations.

Flooding

Definition:

  • Involves transmitting every incoming packet on every outgoing link except the incoming one, facilitating rapid data dissemination.

Challenges:

  • Generates numerous duplicates unless managed effectively, leading to network congestion.

Control Measures:

  • Hop Counter:

    • Limits the number of hops a packet can take before being discarded to prevent infinite looping.

  • Sequence Numbers:

    • Tracks packets to prevent repeats and maintain orderly delivery.

Selective Flooding

Definition:

  • Routers selectively forward packets only towards appropri ate directions, enhancing efficiency in communication.

Practical Uses:

  • Particularly beneficial in military applications and distributed databases, where resilient and controlled communication is essential.

Distance Vector Routing

Overview:

  • Routers maintain a table (vector) of the best-known distances and the optimal line to reach each destination, allowing informed routing decisions.

Routing Table Structure:

  • Each entry consists of the preferred line and an estimated distance, facilitating efficient data transfer.

Challenges of Distance Vector Methods:

  • Count-to-Infinity Problem:

    • Difficulty in resolving incorrect distance information leading to prolonged convergence times.

Link State Routing

Transition from Distance Vector:

  • Addresses delays and efficiencies by implementing five core functions:

  1. Discover neighbors and their addresses.

  2. Measure the delay to each neighbor effectively.

  3. Construct link state packets containing this information for distribution.

  4. Distribute these packets to all routers via controlled flooding.

  5. Compute the shortest paths based on the accumulated and updated data from neighbors.

Hierarchical Routing

Purpose:

  • Aims to manage extensive routing tables by organizing routers into regions or clusters, enhancing scalability.

Trade-offs:

  • While offering space savings, this strategy may increase the overall path length for data transmission, potentially impacting latency.