Internet Networking With TCP And IP For Open Systems Interconnection

Open Systems Interconnection (OSI)

Discuss about the Internet Networking with TCP and IP for Open Systems Interconnection.

The Open Systems Interconnection (OSI) and TCP/IP are internet models that define the transmission of data and information over networks, and interrelations between protocols, as well as how they work. The OSI model traces its roots to 1970s and is just a model whose use is not mandated in networking; although several protocols closely adhere to it. The OSI model is used mainly for the describing, discussing, and understanding individual functions in networks. It has seven layers, although in simpler applications, not all models are used (Alani, 2014). The TCP/IP can be traced to the 1960s by the Department of Defense and is also a layered protocol although it doesn’t use all the layers described by OSI despite the layers being equivalent in terms of function and operation.  The IP protocol corresponds to the Layer 3 of OSI model (network layer). TCP corresponds to Layer 4 of OSI (transport layer) as well as some layer 5 functions. Ther are no assumptions in the TCP/IP model about what happens above a network session level (part of OSI layer5); OSI on the other hand defines several more standardized functions layers. OSI specifies two link layers below IP, while TCP/IP makes no prescriptions (Tetz, 2018). In TCP/IP applications can supply functions it needs if they are not found but this is not possible with OSI. OSI is also a generic standard independent of protocol while TCP/IP is regarded the internet standard. TCP/IP is a more reliable protocol than OSI  since the internet is developed around TCP/IP. The OSI data link and the physical is combined by TCP/IP into a network access layer. While the OSI remains a reference model, the TCP/IP model is an OSI model implementation. The OSI model is more useful in working with and describing network models because it is a theoretical model that describes and defines networks deeply; how they work, the encapsulation and De-encapsulation of data, making it very suitable for learning, describing and understanding networks and how they work. TCP/IP, on the other hand, is less suitable because it is a working model that has been simplified (Edwards, 2009).

Address Resolution Protocol (ARP) was defined by RFC 826 in 1982 and is a protocol used in the mapping of internet protocol (IP) address to physical machine addresses that the local network recognizes. It is used in discovering the address of the Link Layers, for example, the MAC address, that is associated with with given address of a network layer, usually an IPv 4 address (Dixit & Singh, 2013). ARP is a request-response protocol that has a link layer protocol encapsulating its messages and is communicated within a single layer boundary and so it is defined in the IP suites’ link layer. ARP makes use of a simple message format with a single response or resolution request. The ARP message size is dependent upon the upper ad lower layer address sizes determined by the network protocol type. ARP cache plays an important role in networking because it is a data repository used for connecting IP addresses to MAC (media access control) addresses for a physical device within a local network.  The ARP cache then keeps track of of hardware addresses for devices managing it (‘Cisco’, 2018).

TCP/IP Model

The network will have a local area network (LAN) for each of the six locations, interconnected to form a wide area network (WAN) or CAN.  The addressing solution to be used is based on TCP/IP, which is universally supported. The computers within the offices/ locations are private hosts, ad so do not require pubic addresses and will be designed any valid addresses. The addressing solution for the network based on the 10.0.0.0 private address means that the RFC IP address block A will be used for addressing the sites. The Class A range of RFC 1918 has the internal address range 10.0.0.0 to 10.255.255.255 (Sequiera, 2013). Addressing the hosts will entail assigning decimal values for every octet 8-binary bits for values between 0 and 255 . Each computer in the organization is a host that requires a host address; a unique number to identify it within the LAN and enable data packets addressed to it to be delivered.  The assumption  is that the network will use IPv4 addressing and values between the broadcast and the network addresses are assigned to devices (Bartz, 2015).

The 10.0.0.0 address becomes the entire network address and because the hosts will number about 1200 in total, the Class A addresses as set aside by RFC1918 will be sufficient. The address 10.0.0.0 belongs to the network, and so is not used for assigning to the hosts. The hosts are assigned from 10.0.01 to 10.0.0.254 because 10.0.0.255 is a broadcast address. The subnet mask gets the address 255.255.255.0 and the 24 bit net bits will be used, so that each subnet will have a total of 256 unique addresses and a classful mask of 10.0.0.0/8 is used that will allow for a total of 16 777 214 hosts to be addressed. This will ensure present needs are taken care of and future expansion of the hosts within the entire network and the various LAN in given locations is possible; that is several more hosts can be added to the network. 

Address	10.0.0.0
Netmask	255.0.0.0 =8
Wildcard	0.255.255.255
Network	10.0.0.0/8
Broadcast	10.255.255.255
Host minimum	10.0.01
Host maximum	10.255.255.254
Hosts per network	16777214

Addresses as per location; 

Location	Workstations	Private Address range
Finance Office	260	10.0.0.1 to 10.0.1.6
IT Call Center	520	10.0.1.7 to 10.0.2.12
Research and Development Office	120	10.0.2.13 to 10.0.2.133
Marketing Department	40	10.0.2.134 to 10.0.2.174
Information Technology	130	10.0.2.174 to 10.0.3.50
Head Office	60	10.0.3.51 to 10.0.3.111

The private addresses cannot be used over the internet, yet the offices need to communicate; a solution for this would be to mix the private addresses with public addresses, but this can result in dis-contiguous subnets which can cause problems such as confusion when updating from both systems. As such, the Network Address Translation (NAT) solution is adopted to enable communication over the internet for all the different locations and branches (‘Cisco Technology Support’, 2014). As defined by RFC1631 refers to the process of swapping a given address for another address in the header of the IP packet. In practical application, NAT is used to enable privately addressed hosts using the RFC1918 addressing system to be able to access the internet. The routers to be used, such as Cisco and UNIX computers within the network operate on a stub domain corder (Ferguson, 2015). The entire network can have a single connection to the internet so any host within the sub domains will forward packets it intends to transmit to other hosts in other locations.

Address Resolution Protocol (ARP)

The NAT device then looks at the IP header and replaces it with a unique global address. A response from the host outside the sub domain sends a response, causing the NAT device to receive the response, check the present table of NATS and replaces destination address with the original found on the inside source. NAT routers use PAT (Port Address Translation) enabling multiple addresses inside each domain to map similar global addresses, so hundreds of private address nodes can use a single global address for internet access. By mapping the UDP and TCP port numbers, the NAT router is able to track different conversations on the network.  The design will have 14 hosts for each subnet (Johnston, 2009). The diagrammatic representation of the network design is shown in the image below;

When the hosts per building jump to more than 1024, the simplest and most practical solution would be to use a subnet mask of 24 bits. The subnet mask makes it possible for IP networks to be subdivided to enhance security and performance by enabling efficient routing of traffic within subnets. By applying the 24 bit subnet mask, the IP addresses for each host is split into two, namely, the host address and the extended network address to support a scheme of two level addressing. This will enable all the hosts to be addressed uniquely within each sub domain

References 

Alani, M. (2014). Guide to OSI and TCP/IP Models. Cham: Springer International Publishing.

Bartz, R. (2015). Mobile computing deployment and management: Real World Skills for CompTIA Mobility+ Certification and beyond (p. 418). Indianapolis, Ind.: Sybex.

‘Cisco’. (2018). IP Addressing: ARP Configuration Guide, Cisco IOS Release 15M&T – Address Resolution Protocol [Support]. Cisco. Retrieved 7 April 2018, from https://www.cisco.com/c/en/us/td/docs/ios-xml/ios/ipaddr_arp/configuration/15-mt/arp-15- mt-book/arp-config-arp.html

‘Cisco Technology Support’. (2014). Network Address Translation (NAT) FAQ. Cisco. Retrieved 7 April 2018, from https://www.cisco.com/c/en/us/support/docs/ip/network-address- translation-nat/26704-nat-faq-00.html

Dixit, V., & Singh, V. (2013). Essentials of computer networks, internet and database technologies. Oxford: Alpha Science International.

Edwards, J. (2009). Networking Self-Teaching Guide [electronic resource] : OSI, TCP/IP, LAN’s, MAN’s, WAN’s, Implementation, Management, and Maintenance (1st ed.). Hoboken, NJ: Wiley.

Ferguson, B. (2015). CompTIA Network+ review guide (3rd ed., pp. 76-77). Indianapolis: John Wiley & Sons.

Johnston, A. (2009). SIP: Understanding the Session nitiation Protocol (3rd ed., pp. 235-237). Boston: ARCTECH House.

Sequiera, A. (2013). Interconnecting Cisco network devices (2nd ed.). Indianapolis, Ind: Cisco Press.

Tetz, E. (2018). Network Basics: TCP/IP and OSI Network Model Comparisons – dummies. dummies. Retrieved 7 April 2018, from https://www.dummies.com/programming/networking/cisco/network-basics-tcpip-and-osi- network-model-comparisons/