PRESENTATION COMPUTER NETWORKS

UNIT 4: NETWORK LAYER

CONTENTS Design Issues Routing Algorithms Shortest Path Routing Flooding Broadcast & Multicast Routing Congestion Control Internetworking

INTRODUCTION The role of network layer is to establish end to end connectivity between the source and destination. The fundamental job of network layer is to route packets from source machine to destination machine. If too many packets are present on the same route, then this may lead to congestion or collision. It is the duty of network layer to manage congestion and to protect packets from congestion.

DESIGN ISSUES Store & Forward Packet Switching – it is a technique in which information is sent to an intermediate station where it is kept and send at a later time to the final destination when the route gets free. The packet is stored at the intermediate node in its queue until it is fully arrived. Services provided to the Transport Layer – Two kinds of services are provided by the network layer to the transport layer – - Connection Oriented - Connectionless (To be continued)

Virtual Circuit & Datagram Subnet – In Virtual Circuit dedicated direct link is established between the source and destination. Each packet follows the same route to reach destination. Delay is at the initial stage only. In Datagram Subnet no direct link is established between source and destination. Each packet follows independent route to reach the destination. Delay is at each and every packet.

ROUTING ALGORITHMS Routing Algorithms are the act of moving information across an internetwork from source to destination. Routing Algorithms play a vital role in network and are used to define route for packets. In routing the path with the lowest cost is considered the best. There are two major classes of routing algorithms - Adaptive Routing Algorithms - Non Adaptive Routing Algorithms (To be continued)

PROPERTIES OF ROUTING ALGORITHMS Correctness Simplicity Robustness Stability Fairness Optimality

SHORTEST PATH ROUTING Shortest path is the best route taken by a data packet travelling between two network nodes. This algorithm is one of the most popular algorithm for computing the shortest path between two nodes. This algorithm is also called Dijkstra Algorithm because it was proposed by Dijkstra in 1957. This algorithm computes the shortest path from a node called root node to every other node on the network. (To be continued)

The algorithm starts with determining the source and the destination node on the graph. All the nodes in the network are divided into two groups Permanent (P) & Temporary (T). At the time of starting, ‘P’ is initialized to the current node and ‘T’ is initialized to null. In group ‘P’ we have nodes to which the shortest path has already been found and in group ‘T’ we have nodes to which we are considering shortest path i.e all the remaining nodes. In the beginning, algorithm makes source node ‘q’ as permanent node i.e put it in ‘P’ group. (To be continued)

For this source node ‘q’, the algorithm adds each of its neighbor ‘n’ to ‘T’ group along with its cost to reach it through the previous node ‘q’. After adding ‘q’s various neighbors in ‘T’, the algorithm picks up one of the neighbors which has the smallest cost in ‘T’. If this neighbor ‘n’ is not present in ‘P’, the algorithm adds it to ‘P’, the algorithm adds it to ‘P’ thus making it permanent. The algorithm stops when ‘T’ is empty.

SHORTEST PATH ROUTING

FLOODING Flooding is an non-adaptive algorithm and requires no network information. In flooding the final destination of the outgoing line is not checked, simply packets are duplicated onto all destinations. Since flooding naturally utilizes every path through the network, it will also use the shortest path. The main advantage of flooding is the increased reliability provided by this routing method. Since the message will be sent at least once to every host, it is almost guaranteed to reach its destination. (To be continued)

The problem with flooding is that it generates large number of duplicate packets at the final destination. While a message may have only one destination it has to be sent to every host. The following techniques an be used to stop duplication of packets – - To keep the track of the packets that have been flooded, to avoid sending them a second time. - Another solution is to use ‘selective flooding’. In this the router do not send every incoming packet out on every output line. Instead packets are send only on these lines which are going in the right direction.

BROADCAST ROUTING Broadcast Routing is a process of sending a message to all the nodes in a network simultaneously. In broadcast communication there is only one source and all other hosts are the destination. The relationship between source and destination is one to all. Examples of Broadcast Routing – Stock Market updates, Radio programs might work best by broadcasting to all machines. Simple Broadcasting, Flooding, Multi-Destination Routing, Reverse Path Forwarding are methods which have been proposed for Broadcast Routing.

BROADCAST ROUTING

MULTICAST ROUTING Multicast Routing is a method of forwarding datagrams to a group of interested receivers. The sending of a packet from 1 sender to multiple receivers with a single transmit operation is called Multicast Routing. In multicast communication there is one source and a group of destinations. Multicast design applications can send one copy of each packet and address it to a group of computers that want to receive it. Examples of Multicast Routing – Videoconferencing, a group of processes implementing a distributed database system.

MULTICAST ROUTING

CONGESTION CONTROL In any network when there is too much data traffic at a node that the network slows down or starts loosing data, it is known as congestion. Congestion is the situation where traffic is more on the line than the carrying capacity of the line. Congestion Control concerns controlling traffic entry into a communication network so as to avoid collapse by attempting to avoid over subscription of any of the processing or link capability of the nodes and the network. (To be continued)

The ideal situation is when packets sent are directly proportional to packets delivered. When one part of the subnet (e.g. one or more routers in an area) becomes overloaded, congestion results. Because routers are receiving packets faster than they can forward them, one of two things must happen: The subnet must prevent additional packets from entering the congested region until those already present can be processed. The congested routers can discard queued packets to make room for those that are arriving.

CONGESTION

CAUSES OF CONGESTION Packets to the same outgoing line build up a queue. Slow router’s or CPU’s can cause queues to build up. Low Bandwidth lines. Low buffer capacity of node. Deadlock. Bursty traffic. Insufficient memory to store arriving packets. Packet arrival rate exceeds the outgoing link capacity. Low processing capacity of subnet as compared to sending capacity of hosts.

STEPS TO AVOID CONGESTION By monitoring vacant buffer at each node, so that incoming flow of packet can be stopped when buffer nearly becomes full. By discarding some packets and for this a life time of packets should be fixed. Flow control can be used to avoid congestion. Traffic shaping is another concept that can be used to prevent from collision. Upgrading router’s and CPU’s can lead to faster processing of packet which may avoid timeout for the packets. Increase in bandwidth and performing QoS on the traffic can be used to avoid congestion.

INTERNETWORKING Interconnecting two or more networks to form a single network is known as internetworking and the resulting network is called internetwork. Build a single network (an interconnected set of networks, or internetwork, or internet) out of a large collection of separate networks. Each network must stand on its own, with no internal changes allowed to connect to the internet. Communications should be on a best-effort basis. “Black boxes” (later called routers) should be used to connect the networks. No global control at the operations level. (To be continued)

Internetworking is the practice of connecting a computer network with other networks through the use of gateways that provide a common method of routing information packets between the networks. The resulting system of interconnected networks is called an internetwork, or simply an internet. The most notable example of internetworking is the Internet, a network of networks based on many underlying hardware technologies, but unified by an internetworking protocol standard, the Internet Protocol Suite, often also referred to as TCP/IP. Implementing a functional internetwork is not a simple task.

INTERNETWORKING DEVICES Routers Bridges Repeaters Gateways Note - Internetworks are networks composed of several network segments which may differ in topology and protocols.