Internwtworking Via TCP/IP Model

Question: Describe about the Internwtworking for Via TCP/IP Model. Answer: 1. The differences between the repeater, switches and a router are described below. Repeater: Each and every transmission media create signal losses while transmitting data. After a particular distance, the signal travelling through the transmission media will lose all the energy and hence identification of that signal and the identification of the data brought by that signal will be difficult. So signal cannot be transmitted after a certain distance in any transmission media. Repeaters are used to remove this distance limitation. Repeaters are the devices used to regenerate the weak signals to its original energy level so that the signal can be transferred some more distance (McQuerry, 2004). The repeater can be used a as a noise remover. One cannot say it is an amplifier but it is a regenerator (McQuerry, 2000). Repeaters cannot split the host domain. It doesnt have the filtering capacity. It cannot differentiate the analog signal and digital signal while doing repeater functionality. Based on the signal type the repeaters can send the data (McQuerry, 2004). It works in the physical link later in OSI model. It checks logical link control and packets of the sender segment and receiver segment. If they are same in both segments then repeater can send the data. The collision stopping power in repeater is limited. It cannot filter the signal too (McQuerry, 2000). Switch: This device is used to transfer data from one host to another host. In another words switch can do one-to-one data transfer between devices. It operates in layer-2 of OSI model usually. Intelligent switches can operate in layer-3. It can learn and store the MAC address of the devices connected with the switch and that data will be used to do one-to-one data communication. It can split the collision domain and hence avoids collision in the network. Filtering capacity is limited. It wont do broadcasting but it got the capability (McQuerry, 2000). Router: Routers are layer-3 devices, used to communicate between two different networks. IP Address is the important addressing parameter in routers (McQuerry, 2004). Routers can create communication between devices of different protocols and architectures. Router does pocket forwarding. It can exchange protocol related information too. Routers got the filtering capabilities. It avoids collision. It controls the traffic and filters the type of traffic. Using routers the traffic can be denied using routing tables. Routers are using special type of protocols called routing protocols. Usually routers are more costlier than most of the other networking devices (McQuerry, 2000).. Some of the differences are listed below (McQuerry, 2000). Repeater Switch Router It doesnt have any knowledge about the devices connected with the device. It got limited small level of intelligence about the devices connected with the switch. It got more knowledge about the devices connected with the router. Layer-1 device Layer-2 device mostly Layer-3 device Multicasting Initially broadcasting. Then unicasting and multicasting follows. Initially broadcasting. Then unicasting and multicasting follows. Does not remember any data (McQuerry, 2004) Remember MAC addresses of the devices in the lookup table. Switch can learn the new incoming MAC addresses (McQuerry, 2004) It remembers the IP addresses and stores them in routing tables (McQuerry, 2004). Data collision occurs Low collision possibilities Low collision possibilities are there Speed range can be 100 MBPS Speed range is upto 1000 MBPS Speed range is upto 80 MBPS No routing decision is taken For complex network , a delayed routing will be done Faster routing decision will be taken. No NAT operation possible No NAT operation possible Can do NAT operation LAN device LAN device LAN and WAN device Mostly few ports will be there Number of ports will be more. Number of ports will be less Slow Faster in LAN switching Faster in WAN/MAN routing A router is more complex and expensive than a switch: Routers are more complex because of its functionality and the complexity. Router is a WAN device mainly. It is used to connect two different networks. It collects the location information and hop information about the network nodes in multiple networks, analyse them, forms routing tables and routes the incoming pocket based on the routing table. Routers take care of filtering operation too. Routers are providing security like firewalls. It can do NATing too (McQuerry, 2004). Since router is exposed to internet, the security level of the router must be high. Router decides the routing path based on routing protocols. So router OS got the capabilities to understand routing algorithms and to take decision (McQuerry, 2004). Switch is a LAN device. It collects the MAC address details of the devices connected with the switch and got the details of who is who. Switch forwards the data from the sender to receiver. Switch will not provide the best path from the sender to the receiver. It simply broadcasts. Switch cannot be used to control the internet access (McQuerry, 2004). A switch is more complex and expensive than a repeater: The switches are intelligent devices. It learns and memorizes the MAC addresses. It stores the MAC address values in its switching table. Usually switches are Layer 2 devices. Layer 3 switches are doing more than usual Layer-2 switches. Layer-3 switches can do part of routing operation too. It can understand the IP address details too. Switches got buffers to control the data flow. It controls the data transfer too. Switch got the capabilities to stop the collisions too (McQuerry, 2004). Repeater cannot do all the above activities. It just boosts the energy level of the signal and forwards to the output node. It is a layer-1 device. It will not store and memorize any data for the future use. It will not do anything to stop collisions. Collisions are very much possible in repeater traffic. Switch is capable of managing the traffic on the transmission line and splits the collision domain of the host, whereas a repeater can send the packets only to the next segment and it creates collisions on the network. This is because, a series of repeaters are connected on the same collision domain (McQuerry, 2004). The switch must know the MAC address of each end devices that are connected on the port of the switch to send the message to its end devices. Therefore, the switches are more complex than the repeaters. On the other hand, the repeater allows only one transmission at a partic ular time. Thus, switches are more expensive and complex than the repeaters. 2. For these IP Address calculations (Cowley, 2012) The Network address of the ISP is150.80.0.0/16. It is a class B network. It got 16 host bits and 16 network bits. It got 2^16-2 IPs = 65534 Its Network ID is 150.80.0.0 and SNM=255.255.0.0 From the IP Requirements the Required IPs(Min) can be 200*128+400*16+2000*4 = 40000 For providing 40000 IPs at least 16 host bits are needed. Remaining 16 bits will go as a network bits. Since 16 bits can give 65534 IPs out of which only 40000 can be used, Remaining IPs will be free IPs. Free IPs = 65534 40000 = 25534 Segment 1 needs 200*128 = 25600 IPs. 15 host bits are needed to provide these IPs. Segment 2 needs 400*16=6400 IPs. 13 host bits are needed to provide these IPs. Segment 3needs 2000*4=8000 IPs. 13 host bits are needed to provide these IPs. Segment 1 sub block can be = I stgroup sub block=150.80.0.0/17 Segment 2 sub block can be = II nd group sub block=150.80.0.0/19 Segment 3 sub block can be = III rd group sub block=150.80.0.0/19 Class B IP range starts from 128.X.X.X to 192.X.X.X and hence 150.X.X.X falls under class B. So by default the given IP and ISP's network address 150.80.0.0/16 falls in class B (Cowley, 2012). Class B can give 2^16-2 IPs. The IPs in which all the host bits are zero will act as a net id and the host ip in which all the host bits are 1 act as a broadcast IP. Hence these two IPs cannot be used for host computers. So 2 IPs are deducted from 2^16. As the required IPs is only 40,000 and the available IPs are 65534. The remaining free IPs is calculated by deducting 40000 from 65534. So 25534 IPs are free IPs available in ISP's hand. All the 2^16-2 IPs can be used very efficiently by doing sub netting. While doing sub netting host bits will be less than 16. It cannot be 16 as it will leads to classful Class B IP (Cowley, 2012).. So host bits can be made 15,14,13,12...1 like that. And the number of IPs resulting from this wills as if that IPs belongs to a separate network. Number of host bits in sub netted IP address depends on the requirement. Hence for the segment 1 where the requirement is 25600. Hence 15 host bits can give this IP address count. Hence segment 1 can use 17 network bits. Similarly the segment 2 can use 19 network bits. Sub blocks are calculated accordingly (Cowley, 2012). 3. R1 and R2 both routers are forming a big network. If R1 goes down, then R2 handles the routing between the hosts. If R2 goes down then R1 handles the routing. But H3 and H6 hosts are dropped down from the network when R1 goes down. In case if R2 goes down, one host H4 gets drops down from the network. Since the type of traffic and purposeof the traffic are not known , the frequency of the traffic also not known. Hence it is very difficult to tell which router is important (Singh, 2010). As a general view the router which causes maximum damage to the network when it is failed will be considerered as an important router. Hence R1 is considered as an important router. 4. The distance vector routing (Cs.bu.edu, 2016) knows the cost of best path and neighbour route address. It updates the routing table when Distance Vector route changes. So, it updates frequently. The distance vector and link-state are the two primary classes of routing protocol. The distance vector routing comprises of two factors they are, distance and direction. The Link-state routing (Cs.bu.edu, 2016) has the knowledge of all the nodes and routers on the network. It uses Dijikstras algorithm (Cs.bu.edu, 2016) to update all information on the switch. Link-state routing performs reliable flooding. This is the process of sending information to all the nodes on the network. When all nodes collect the information on the network it will lead to stop the flooding. Link-state routing contains four fields. They are ID of the node, sequence number, address list of all the connected nodes, time to live (TTL) of the packet(Singh, 2010). Link-state routing needs more memory space. Because it should have knowledge of the whole network topology. It does not need frequent update. It requires more processing power because of its knowledge about topology on the network (Cs.bu.edu, 2016) The Distance vector routing changes the information in an iterative manner. It changes local link cast and update Distance Vector. When the Distance vector changes that time only the node notifies the neighbour. First, the routing determines the cost of neighbours then estimate the least cost of all. All the time the Distance Vector sends to the neighbours and updates. Each time of transmission the neighbour estimate and determine the least cast path. All neighbours contain distance on its list. In this routing, the router table initially set the distance at infinity. Then, the router receives packet and determine the destination. Then only it finds least cast path or shortest path on the network. Each time the router checks the neighbour routes cost or distance. It also sets the path to known after visiting each link on the network. Due to its routing table updates at all time this require more bandwidth than the Link-state routing. But it did not require more memory space. Because it did not needs to know about the overall topology of the network (Singh, 2010). The below table explains the differences (Graziani and Johnson, 2008) (Cs.bu.edu, 2016) Distance Vector Routing Link-State Routing In distance vector routing, router needs only the information of the next hop. In link state routing, router knows the complete information of all other host. It need not know the complete network topology. It must know the complete network topology to make sure the reachable paths by all the other routers in the network. It advertises the complete routing table. It can be difficult to configure the routing table. Management is not required in distance vector routing. More management is required for link-state routing. It requires more bandwidth. It requires less bandwidth. It uses direct addressing method to find the best path on the network. It uses the reachable paths on the network. It is mainly used for identifying the best destination path in the network. It also used for identify the best destination path in the network. It doesnt support classless routing. Classless routing supports to link-state routing. It updates the routes frequently. It makes use of triggered routing updates. Broadcast is used to update the routing. Multicast is used to update the routing. Routing is performed in a decentralised manner. Routing is performed in a centralised manner. It affects the routing loops. It less affects the routing loops. Only routing table is created. It creates neighbour table, topology table and routing table. It can be easily configured. It requires more processing power for configuration. It requires less memory and CPU usage. It requires more memory and CPU usage. It requires less processing power. It only forwards and advertises the information about the network topology to the directly connected links and adjacent routers. Best destination path is calculated by Bellman-Ford algorithm. Best destination path is calculated by Shortest Path First algorithm. 5. Role of ICMP (Internet Control Message Protocol) in IP protocol Architecture: The ICMP is a communication protocol in between the two connected systems of IP protocol. The Internet Protocol (IP) is not a reliable protocol in communication environment. It does not guarantee to deliver the packets on the required destination. Sometimes the destinations have the problem of missing packets. So, high level protocol is required to meet the problem of missing packets. The ICMP protocol is used to report the errors, if any packets miss on the destination device. (Cisco, 2016) Figure: ICMP Header Various tasks are performed by ICMP protocol. ICMP is used to report the routing failures. It helps in testing the node reachability within the internet. It helps to redirect the ICMP messages to the source IP address. It helps to generate the ICMP Time Exceeded message and various other messages. (Cisco, 2016) ICMP is used by the hosts and routers to communicate with the network-level information. ICMP creates and sends messages to the source IP address. The ICMP message consists of router and the IP address. Any IP network device has the capability to send, receive or process ICMP messages (Cisco, 2016). The ICMP lets denotes the internet gateway through router, service or a host. This is used in case if it the packets are not delivered to the destination. Figure: ICMP Error Message Purpose of IP header and first 8 bytes of datagram data The Internet Protocol has a format of 8bytes in the header part which is used to send the error messages to the sender. It contains an error message and an IP address of the sender. It helps the sender to match the error message of the host (Cisco, 2016) Figure: IP Datagram Structure ("The IP datagram structure", 2016) An ICMP message consists of 8 bytes long in every transmission to the sender if error occurs on the network. The IP datagram header contains at least 20 bytes. The ICMP message received by the sender performs Path MTU Discovery to identify the IP address of the destination packet, which caused the error (Singh, 2010). References Cisco. (2016).TCP/IP Overview. [online] Available at: https://www.cisco.com/c/en/us/support/docs/ip/routing-information-protocol-rip/13769-5.html [Accessed 4 Sep. 2016]. Cowley, J. (2012).Communications and networking. Heidelberg: Springer. Cs.bu.edu. (2016).Distance-Vector Routing. [online] Available at: https://www.cs.bu.edu/fac/byers/courses/791/F99/scribe_notes/cs791-notes-990923.html [Accessed 4 Sep. 2016]. Graziani, R. and Johnson, A. (2008).Routing protocols and concepts. Indianapolis, Ind.: Cisco Press. McQuerry, S. (2000).Interconnecting Cisco network devices. Indianapolis, IN, USA: Cisco Press. McQuerry, S. (2004).Introduction to Cisco networking technologies (INTRO). Indianapolis, Ind.: Cisco Press. Singh, V. (2010).Computer networking course. New Delhi: Computech Publications. Tuxgraphics.org. (2016).network address calculator / ip address calculator. [online] Available at: https://tuxgraphics.org/toolbox/network_address_calculator_add.html [Accessed 4 Sep. 2016].