Ntc_Week_4_Individual

Topologies Bus The one feature of bus network that defines this topology is the existence of a backbone. Every host on the network (or network segment) is connected to this cable. Although bus networks are the most inexpensive and simplest to install of all the network topologies, one break in the backbone renders the network segment useless. Because each end of the backbone needs to be terminated to prevent signal bounce, a break in the backbone results in the backbone not being terminated and an endless cycle of signal bounce that cripples the network. Star The distinguishing feature of a star network is the existence of a central hub that all hosts on the network (or network segment) connect to. Some hubs are passive and forward all network traffic to all hosts on the network, and some hubs are intelligent and forward traffic only to its appropriate destination. Installation of a star network is a bit more expensive than a bus network, but the tradeoff is that a failed network cable does not result in the loss of network connectivity for any other network nodes except the one directly attached to the failed cable. Network connectivity problems can quickly be isolated and corrected. If star networks have a disadvantage, it is that the hub can present a bottleneck in network throughput. Should the hub fail altogether, the entire network (or segment) will lose connectivity. Ring Ring networks are distinguished by the fact that each computer is connected to the next in a closed loop. Data packets that are not meant for a specific node are regenerated by the hosts and sent along to the next host on the network until the data reaches its intended destination. Ring networks are also capable of utilizing token passing media access control, but regardless of the method of access control used, ring networks provide equal opportunity for all network nodes to access the transmission medium. One node cannot monopolize a ring network in the same way that can happen on star and bus networks. FDDI networks can even remove a failed node from the network, meaning data can travel in the opposite direction around the ring. Ring topologies can be either physical or logical. A physical ring is easy to identify by the closed loop of network medium, but a logical loop cannot be identified by simple observation. A physical star topology can function as a logical ring topology by using any of the token passing media access control methods like FDDI. The primary disadvantage to ring networks is their limited bandwidth when compared to other topologies. In addition, specialized network interface cards are needed to build a ring network. Mesh Mesh networks are the epitome of reliability but because every node is connected to every other node, installation is both expensive and difficult. One failed node has absolutely no effect on the rest of the network aside from not being able to accept traffic. Reliability is the strong point, and any organization that must have the utmost in network reliability and can spare the resources will want to install a mesh network. The cost of a mesh network grows exponentially with growth, so careful consideration must be given to whether or not a mesh network is truly required. When considering any of the network topologies talked about in this paper, one must take into consideration, future growth, money, cable media, and the length of cable. Below in figure 1.1 is a detail understanding and comparison of the topologies, cable media and other variables used when trying to decide on a network for an organization. Table 1.1 Network Topology Comparison Topology | Information Transfer | Setup | Expansion | Troubleshooting | Cost | Cabling Concerns | Star Bus Each computer connects to a central connection device. | All information passes through the central network connection. | Each computer must be close to the central device. 100 meters maximum cable length. Up to 24 computers per network. | Add a new computer by plugging in a new cable from the computer to the connection device. | When one computer goes down, the rest of the network is unaffected. If the connection device goes down, then the network is down. | More expensive of the simple topologies, it requires costly connection device. Usually cheaper than a hybrid network. | Uses twisted pair cable. Requires large amounts of cable. No more than 100 meters from the computer to the connection device. | Bus Single cable connects everything. | One computer at a time sends information. Information goes along the cable and the computer accesses the information off the cable. | Connect the cable from one computer to the next and so on to the end. A terminator is placed at each end of the network. | To add a computer, you must shut down the network and disconnect the cable from the existing computers. | If one computer malfunctions, the entire network goes down. | A cheaper network since there is usually one continuous copper cable. | Single continuous cable connects the devices. Terminator is required at each end of the cable. Uses coaxial or twisted pair cabling. | Ring Single cable configured in a ring. | Information goes in one direction around the ring and passes along the ring until it reaches the correct computer. | Computers are located close to each other. Setup is easy. There is no connector. The ring has no beginning and no end. | Cable between the computers must be broken to add a new computer, so the network is down until the new device is back online. | If there's a break in the cable or an error in the network, information continues to transfer through the rest of the ring until reaching the point of the break. This makes troubleshooting easy. | One of the more expensive topologies due to high cable costs. | Requires more cabling than other topologies. Uses twisted pair. | Hybrid Mesh Combines two or more different structures. | Often used across long distances. Information transfer can happen in different ways, depending on the other topologies. | Often created when expanding an existing network. Can use a variety of connection devices. | Connection devices make combining  different networks and different topologies easy. | Troubleshooting is most difficult in this topology because of the variety of technologies. | Expensive, large, and usually complicated. | Cabling depends on the types of networks. Can use twisted pair and coaxial cable. Also incorporates fiber optic cabling over long distances. | Standards Ethernet IEEE 802.3 defines a collection of standards for copper and fiber optic cabling (the physical layer) and media access control methods. Among the more recognized standards are 10Base-T and 100Base-TX. Such Ethernet standards make it possible for manufacturers to develop networking equipment that will be interoperable with networking equipment from other manufacturers. The only disadvantage to standards like IEEE 802.3 is that they can sometimes inhibit innovation that might drastically improve today’s networking environment. Token Ring IBM was the forerunner in token ring media access control, but modern token ring protocols have since been defined by IEEE 802.5. Token ring networks utilize a three-byte packet of data that serves as a signal to the network host possessing it that it can transmit data over the network. Because only one host is transmitting data at a time on a token ring network, network performance is excellent under heavy traffic. Data retransmissions very rarely occur on a network controlled by token ring media access control because data simply does not collide (and therefore become corrupt). Network bandwidth was always limited and far behind Ethernet standards so when token ring’s 100Mbps standard was introduced, very few manufacturers took interest; many companies had already adopted Ethernet standards of 100Mbps years ago. FDDI FDDI networks transmit data at a rate of 100Mbps over a physical or logical ring network, a much higher speed than a typical token ring network. Two counter-rotating tokens control access to the transmission medium, so redundancy is built into the system (increasingly reliability). In contrast to typical token ring networks, FDDI implementations allow for different nodes to be assigned a different priority regarding media access. If FDDI networks have a disadvantage, it is that the cabling is expensive and the network interface cards must be specialized (making them expensive as well). Wireless Wireless networking is chock full of advantages and disadvantages. The primary advantage of wireless is that the data transmissions are carried through the air on radio waves; therefore, it is never necessary to wire a host to another network component so long as that host is equipped with a wireless network interface card. Not having to install physical cabling can be a tremendous cost saver, but radio signals are always susceptible to eavesdropping and interference. Anyone eavesdropping on a wireless network using the proper tools can decipher the data transmissions, so wireless should not be considered where very sensitive data will be transmitted over a network. Wireless bandwidth is lower than in wired Ethernet networks, which presents another consideration to be made when throughput is a concern. TCP/IP and its relationship to the OSI Model TCP/IP is the suite of protocols on which the Internet is built. Each protocol operates on separate layers of the OSI model and complements the other protocol perfectly. Internet Protocol (IP) operates primarily on the network layer of the OSI model by providing addressing and routing functions as well as accepting packets from TCP and UDP and packaging them for transmission across a network. IP is connectionless, which means it simply sends the packets without worrying about if a connection is present; ensuring that a session is available is the job of TCP. TCP is the half of the protocol that ensures delivery of IP packets; this is done through acknowledgement packets, or ACKs. ACKs can be compared to thank you notes that a receiver sends back to the sender of a data packet to let the sender know the data has been received and does not need to be retransmitted. TCP operates on the transport layer, whose job it is to ensure data is received and in such a way as to not overflow the receive buffer of the receiving node.