OSI

Both the OSI (Open Systems Interconnection) and the TCP/IP (Transmission Control Protocol and Internet Protocol) models function in similar manner. However, they do have some small differences and appreciating these variances is critical in understanding computer networking and troubleshooting local area network (LAN) and wide area network (WAN). For instance, the TCP/IP referencing model uses specific protocols which can be used to troubleshoot problems related to Domain Name Server (DNS) (Roark n.d). Understanding the OSI model at the Physical and Data Link layers is important in correcting possible problems with the physical connections. It is not unusual to encounter problems with network links on routers such as the Cisco router. When this happens, network administrators should be able to troubleshoot these issues using the OSI model by carefully checking for possible causes, layer by layer, from the Physical layer and Data Link layer to the Network layer. The OSI reference model was the first to be adopted before the TCP/IP model. The OSI model was created by the Advanced Research Project Agency (ARPA) to enable the grouping of similar working network components into various layers of the protocol in a logical manner. However, with the advent of the Internet, a need for a more streamline protocol arose (Roark, n.d). This protocol would address a number of needs of the rapidly-growing Internet. For instance, it was important to address the problem when two dissimilar computer systems were communicating. With this need in mind, the Defense Advanced Research Project Agency (DARPA) then decided to develop the TCP/IP suite to address several issues that were found limiting with the OSI reference model.

OSI Model and TCP/IP Suite

The OSI reference model is made up of seven layers of protocols or rules with each layer, providing services to the layer which is above it (Roark n.d). These layers include the Physical layer, Data Link layer, Network layer, Transport layer, Session layer, Presentation layer and the Application layer. The TCP/IP reference model on the other hand is made of four different layers. Again, just like in the OSI reference model, each layer in the TCP/IP reference model is responsible for providing services to the layer which is above it. Each of the layers is critical for a set of computer network tasks and the four layers include the Application layer, Transport layer, Internet layer and the Link layer. Among these layers, there are general differences which can be noted. The OSI reference model was defined before the advent of the Internet; the TCP/IP reference model was defined after the Internet was developed. Service interface and the protocols in OSI reference model are clearly differentiated unlike in the TCP/IP reference model. While TCP/IP reference model supports internet functioning, the OSI reference model does not support internet working. The TCP/IP consists of loose layers while the OSI reference model has strict layering (Roark n.d). Although the TCP part of the TCP/IP reliably delivers packets, the IP does not. In the OSI reference model, each packet is reliably delivered.

Physical and Data Link Layers

Physical Layer

The Physical layer is also known as layer 1 of the OSI reference model (Wetteroth, 2003, p.4). This layer defines the hardware characteristics which are critical in delivering data transition signal. Examples of specific characteristics defined in this layer include the voltage levels and the number and locations of various interface pins (Roark n.d). In relation to the TCP/IP, this reference model does not define any physical standard and instead makes use of the already existing standards. TCP/IP reference model defines the manner in which data is transmitted in the network. The Physical layer directly communicates with the communication medium and carries out two critical missions: sending the bits and also receiving them. In the Physical layer, a wide variety of media are common in data communication, these include electric cables, microwaves, light waves, fiber optics and radio waves. In the OSI reference model, the type of medium used may vary from one layer to another and, therefore, different media may necessitate different sets of rules or protocols of the Physical layer (Roark n.d). This means that the upper layers of the OSI model are totally independent from the processes used in the delivery of bits (Wetteroth, 2003, p.4). In general, the Physical layer will define the patterns of the bits used but will not define the medium, in which data is transmitted. The layer can also describe how data is encoded into the media signals as well as the characteristics of media known as the attachment unit interface.

Data Link Layer

The main function of the Data Link layer is to provide full control of the Physical layer. In the OSI reference model, Data Link layer transfers data from the network layer of one device to another (Reynolds & Tymann 2008, p.128). The layer also converts raw bit stream made of 0's and 1's of the Physical layer into data frames and thus providing error-free transfers from node to node. The Data Link layer detects and seeks to correct any errors that may arise from the Physical layer. Two networks are important in Data Link layer of the OSI model: the broadcast networks and the point-to-point networks. In broadcast networks, all the stations tend to share a common channel while in point-to point networks, pairs of hosts or, sometimes, routers, are directly connected to each other. Typically, all wide area networks (WANs) are known to be point-to-point connected and the LANs are broadcast. Layer 2, which is also known as the Data Link layer ensures that data is delivered to the next layer without errors. The layer formats all the packets for transmission immediately they are delivered. Data link layer is responsible for reliably delivering data across the Physical layer in OSI reference mode (Roark n.d). Data is framed and then retransmitted until it is acknowledged after arrival using the control mechanisms. However, the TCP/IP does not create protocols under this layer. In TCP/IP reference model, layer 2 is referred to as the network interface layer or sometimes known as the link layer. The TCP/IP does not have the Data Link layer per se but has the subnet protocols which share striking similarities with the Data Link layer of the OSI model. The subnet protocols have the IMP-IMP protocols which seek to provide more reliable connections between two IMPs. Data link protocol known as the logical link control (LLC) can be used in both OSI and TCP/IP networks for Ethernet-based networks such as the LANs (Roark n.d).

Data Link and Physical Layers in OSI as they compare with Subnet in TCP/IP

The two layers of the OSI, Data Link and Physical layers correspond to the subnet layer of the TCP/IP reference model as shown in Table 1. In most times, the other lower layers of the TCP/IP model below the network layer are rarely discussed. What the TCP/IP model does is show that the host has to connect to the network by using some defined protocols, so that it can transmit the IP packets through it. Since the protocol used is not well defined, the transfer of IP packets will differ from network to network and host to host. Following much deliberation by several organizations, it was finally decided that the TCP/IP reference model’s network interface layer perfectly corresponds to a number of specific functions of the OSI network layer and the OSI Data Link Layer. Because the two layers deal with the functions fundamentally specific to each individual computer networking technology, the principle of layering dictating that the two be grouped together in terms of their functions is absolutely inappropriate (Roark n.d).

The use of standardized model has enabled the addressing of a number of challenges that have emanated due to the rapid increase in user demands. The TCP/IP has become a more popular standard widely used today for internetworking, mainly because it is the protocol suite which the Internet depends on (Roark n.d). It is also relatively simple and stable compared to other alternatives like the OSI. Finally, the TCP/IP is available on almost all hardware and operating systems environments and is often provided free of charge. The Internet, as we know it, is not centrally governed in their policies for accessing and usage of the services or their technical implementation (Leiden, Wilensky & Bradner, 2009, p.24). Each of the network constituents will set their own standards. What is directed by the maintaining organizations which the Internet Corporation for Assigned Names and Numbers (ICANN) is the two name spaces which are the DNS and the Internet Protocol address. Basically the standardization of the main protocols in TCP/IP which are IPv4 and IPv6 has been made possible by the Internet Engineering Task Force (IETF) which is a not-for-profit organization. IPv6 was rather a later version whose main objective was to correct issues with addressing in IPv4. Individuals all over the world needed more addresses and the designers of IPv6 demanded a manner of interpreting, assigning and applying the addresses which were more application for modern version of internetworking. While IPv6 addressing scheme functions in a similar manner just like IPv4 addressing, addressing has been overhauled to create a new form of addressing system which can support continues internet expansion. With the expansion, it is expected that there will be novel applications in the future.