Computer Networks: Topologies, Devices and OSI Layers

The correct answer is A, Attack Surface. The attack surface of a system is the total collection of every point at which an unauthorised person could try to enter it or pull data out of it: open ports, running services, user accounts, login pages, removable drives and even the employees who can be tricked. Network administrators reduce risk by shrinking this surface, that is by closing unused ports and removing software that is not needed. Option B is wrong because an attack signature is the recognisable pattern of bytes or behaviour by which antivirus and intrusion detection software identify a known attack. Option C is wrong because an attack pattern is a general description of how a class of attacks is carried out, used while designing defences. Option D is wrong because an attack path, also called an attack vector chain, is the particular route a single intruder actually followed from entry to target, not the whole set of possible entry points.

The correct answer is A, Modulator. A modulator changes the form of a signal: it loads the digital bits produced by a computer onto an analog carrier wave so that they can travel over a telephone line, a cable or a radio link. The reverse device, a demodulator, recovers the bits at the far end, and the two together make the modem, whose very name is short for modulator-demodulator. Option B is wrong because a mouse is a pointing input device; it sends position and click data to the computer but does not convert one form of signal into another. Option C is wrong because a plotter is an output device that draws large line diagrams, maps and engineering drawings with pens on paper. Option D is wrong because a keyboard is an input device that turns keystrokes into character codes for the processor, which is coding rather than signal conversion between digital and analog forms.

The correct answer is A, General Packet Radio Service. GPRS was added to the 2G cellular system so that the same radio channel could carry data in packets instead of holding a circuit open the way a voice call does. Because the user is charged for the volume of data rather than for the minutes of connection, it was called an always-on service, and it made mobile browsing, WAP pages and picture messaging possible; the faster EDGE and then 3G grew out of it. Option B is wrong because the R stands for radio, the medium used, not for route. Option C is wrong because the unit of data in this system is a packet, and package is not a networking term here. Option D is wrong because radiation has nothing to do with the expansion; it only resembles the word radio and is placed there as a distractor.

The correct answer is C, Star topology. In a star network every computer has its own cable running to a central hub or switch, so the shape resembles the spokes of a wheel. Adding or removing one machine disturbs nobody else and a broken cable affects only that one node, which is why almost every office and laboratory LAN is wired this way; the weakness is that failure of the central device stops the whole network. Option A is wrong because a bus topology has all nodes tapping one common backbone cable with terminators at its two ends. Option B is wrong because in a ring the machines form a closed loop, each joined to two neighbours, and data circulates in one direction using a token. Option D is wrong because a mesh links every node directly to every other node, needing n(n-1)/2 cables for n nodes and having no central device at all.

The correct answer is B, 15. In a full mesh every node is joined directly to every other node, so the number of links equals the number of pairs that can be formed from the nodes, given by the formula n(n-1)/2. Putting n equal to 6 gives 6 multiplied by 5, that is 30, divided by 2, which equals 15 cables; each machine also needs 5 network ports, so the port count is n(n-1) in total. Option A is wrong because 12 would be twice the number of nodes and has no basis in the formula. Option C is wrong because 18 is three times the number of nodes and again does not follow from counting pairs. Option D is wrong because 30 is the value of n(n-1), which counts each link twice, once from either end, and must be halved since one cable serves both directions.

The correct answer is C, Network layer. A router joins two or more different networks and decides, for every packet, which outgoing line will carry it towards its destination. It makes that decision by reading the destination IP address, and the IP address is a network layer address, which is the third layer of the OSI model counting from the bottom. Option A is wrong because devices of the physical layer, such as a repeater or a hub, only pass on or regenerate electrical signals and cannot read any address. Option B is wrong because data link layer devices, namely the switch and the bridge, forward frames using MAC addresses and work within one network rather than between networks. Option D is wrong because the transport layer is where TCP and UDP break data into segments and manage acknowledgements; a gateway may operate up to that level and above, but an ordinary router does not.

The correct answer is C, 48 bits. The Media Access Control address is the permanent physical address written into the network interface card by its maker, and it is 48 bits long, shown to users as twelve hexadecimal digits in six pairs such as 00:1A:2B:3C:4D:5E. The first half identifies the manufacturer and the second half the particular card, so no two cards in the world should carry the same value. Option A is wrong because 16 bits is the width of a port number, which identifies an application and not a machine. Option B is wrong because 32 bits is the length of an IPv4 address, a logical address that changes when the machine moves to another network. Option D is wrong because 128 bits is the length of an IPv6 address, adopted because the supply of 32-bit IPv4 addresses was running out.

The correct answer is B, Ring topology. In a ring the machines form a closed loop and a small control frame called a token keeps circulating around it; only the computer holding the token may put data on the ring, and when it finishes it releases the token to the next machine. Because permission is passed around in turn, no two nodes transmit together and collisions cannot occur, which was the chief attraction of the old Token Ring networks. Option A is wrong because a star network depends on a central hub or switch and uses contention or switching, not a circulating token. Option C is wrong because a bus network uses CSMA/CD, in which a node listens to the backbone, sends when it is free and backs off if a collision is detected. Option D is wrong because a tree is only a hierarchy of star segments and inherits their access method.

The correct answer is B, RJ-45. The registered jack numbered 45 is the eight-pin plastic plug crimped on the end of an unshielded twisted pair Ethernet cable, and it is what clicks into the network port of a computer, a switch or a router. Option A is wrong because a BNC connector, with its bayonet twist lock, belongs to coaxial cable and was used in old bus networks and in test instruments. Option C is wrong because RJ-11 is the smaller six-position jack of an ordinary telephone line, which carries at most two pairs and cannot serve an Ethernet link. Option D is wrong because SC, the subscriber connector, is one of the push-pull connectors used on optical fibre, along with the LC and ST types, and it terminates glass strands carrying light rather than copper pairs carrying electricity.

The correct answer is C, Telephone conversation. Full duplex means both ends can send and receive at the very same moment, and on a telephone call each speaker can talk while still hearing the other, so traffic flows in both directions together. Option A is wrong because a keyboard only sends characters to the computer and never receives any from it, which makes it simplex, a one-way mode. Option B is wrong because a walkie-talkie is the standard example of half duplex: the user must press the button to speak and release it to listen, so the two directions take turns on one channel. Option D is wrong because a television broadcast travels only from the transmitter to the viewer's set, with no return path from the viewer, and is therefore simplex as well.

The correct answer is B, TCP. The Transmission Control Protocol first sets up a connection through a three-way handshake, then numbers every segment, waits for an acknowledgement for each and sends it again if none arrives, and finally hands the data to the application in the original order. Because of this care it is used for web pages, email and file transfer, where a missing byte cannot be tolerated. Option A is wrong because UDP is connectionless: it fires datagrams without a handshake, acknowledgement or reordering, which makes it fast and suitable for live voice, video and DNS queries. Option C is wrong because IP is a network layer protocol that only addresses and routes packets and gives no delivery guarantee at all. Option D is wrong because ICMP carries control and error messages, such as the replies used by the ping command, and does not carry user data.

The correct answer is B, the internet. The internet joins networks in every country through undersea cables, optical fibre on land and satellite links, all of them following the common TCP/IP rules, and no single organisation owns it, so it is rightly called the largest wide area network and a network of networks. Option A is wrong because Ethernet is the family of standards, numbered IEEE 802.3, used to wire a local area network inside a building, not a wide area network. Option C is wrong because Bluetooth is a short range radio technology standardised as IEEE 802.15, which reaches about ten metres and builds a personal area network. Option D is wrong because the intranet of a company is a private network restricted to its own staff; it may be spread over several offices, yet it is far smaller than the public internet and usually rides on it.