You are on page 1of 5

Name: Jinesh G.

Sanghvi CIS617 Final Term Question Bank Q1 Compare the Connection oriented and connectionless oriented internetworking approaches !ns Conne"tion #riente$: Advantages: 1. Buffers can be reserved in advance 2. Sequencing can be guaranteed. . Short headers. !isadvantages: 1. "o alternate routing around congestion. 2. #ulnerabilit$ to router failures along the path. . %equirement of table spaces at the routers. &. !ifficult to implement if one of the networks involved is an unreliable datagram network. Conne"tionless #riente$ a%%roa"h: Advantages: 1. Can be used to connect subnets without #C's inside. 2. %obust to router failures. . (ore potential for adapting congestion !isadvantages: 1. "o sequencing. 2. )onger headers.

/n this wa$ passage through the small3packet network has been made transparent and the subsequent networks are not even aware that fragmentation has occurred. (ultiple e+it gatewa$s can be used because reassembl$ occurs onl$ at the destination host& and thereb$ higher performance can be achieved. Advantages: 1. Non'Trans%arent Fragmentation: 0his strateg$ refrains from recombining fragments at an$ intermediate gatewa$s. /n this approach when an oversi. *ach fragment in addressed to the same e+it gatewa$& where the pieces are recombined. !isadvantage: 0he total overhead increases& as each fragment must have a header.e is too small to handle& then the packet is break up into small chunks called fragments and each fragment is sent as a separate /nternet packet.ragmentation *+plain different t$pes of .ragmentation with their advantages and disadvantages !ns Fragmentation: -hen a large packet wants to travel through a network whose ma+imum packet si. 4verhead required to repeatedl$ reassemble and then re3fragment a large packet passing through a series of small3packet networks. *g : A0( !isadvantages: 1. .ed packet arrives at a gatewa$& the gatewa$ breaks it up into fragments. *+plain . 2. All fragments must e+it via the same e+it gatewa$.. 4nce a packet has been fragmented& each fragment is treated as though it were an original /5 packet. Trans%arent Fragmentation: 0his strateg$ makes fragmentation caused b$ a 1small packet2 network transparent to an$ subsequent networks through which the packet must pass on its wa$ to ultimate destination. 0he e+it gatewa$ must know when it has received all the pieces& so a count field or an 1end of packet2 bit must be provided.Q2. .

Q&. )an$om *arl+ $ete"tion: -hen the average queue length e+ceeds a threshold& packets are picked at random from the queue and discarded. 0he urgent pointer is used to indicate a b$te offset from the current sequence number at which urgent data are to be found. through1< will cause a gap at the receiver that ma$ force packet : through 1< to be re transmitted8if the receiver routinel$ discards out3of3order packets9. 0here are two policies: a. /t is also used to re7ect an invalid segment or refuse an . -ine 8drop new packets and keep old9 : . (ilk 8drop old packets and keep new9 : .Q . 0he receiver is hereb$ kindl$ requested to deliver the data to the application upon arrival and not buffer it until a full buffer has been received %S0: 0he %S0 bit is used to reset a connection that has become confused due to a host crash or some other reason.or file transfer an old packet is worth more than a new one because dropping packet : and keeping packets . 6ive the different techniques of dropping packets to disappear congestion. 5S?: 0he 5S? bit indicates 5=S?*! data.or multimedia a new packet is more important then the old one because a user ma$ not mind dropping few packets while using real time applications. *+plain the : one bit flags in the 0C5 header !ns =%6: =%6 is set to 1 if the urgent pointer is in use. AC>: 0he AC> bit is set to 1 to indicate that the Acknowledgement number is valid. !ns (oa$ She$$ing: A router drowning in packets can 7ust pick packets to drop depending on the application running. /f AC> is <& the segment does not contain an acknowledgementso the acknowledgement number field is ignored. b.

. .B. Since $ou need three subnet bits& which leaves $ou with five bits for the host portion of the address.1B. 6iven the Class C network of 2<&. 0wo bits would onl$ allow $ou four subnets 8229. QB. /n essence the S@" bit is used to denote C4""*C0/4" %*Q=*S0 and C4""*C0/4" ACC*50*!& with the AC> bit used to distinguish between those two possibilities. /t specifies that the sender has no more data to transmit. S@": 0he S@" bit is used to establish connections.<C2&& subnet the network in order to create the following network with the host requirements shown./": 0he . ?ow man$ hosts will this support 2B A 2 8 < usable9. 0he largest subnet must support 2D host addresses. 0he connection repl$ doesnot bear an acknowledgement& so it has S@" A 1 and AC> A1. /n order to create the five needed subnets $ou would need to use three bits from the Class C host bits. )ooking at the above network& $ou can see that $ou are required to create five subnets. /s this possible with a Class C network and if so& then how @ou can start b$ looking at the subnet requirement. 0his meets the requirement. 0he connection reques has S@" A1 and AC> A< to indicate that the pigg$ back acknowledgement field is not in use.attempt to open a connection./" bit is used to release the connection.

B.:&C2. An e+ample of how $ou might assign the sub3networks is as follows: netA: 2<&.1B. net!: 2<&.B.B.0herefore $ou have determined that it is possible to create the above network with a Class C network.B.<C2.1B. host address range 12E to 1BD .B. netB: 2<&.12DC2. to 12: net*: 2<&. host address range 1 to < host address range to :2 host address range :B to E& host address range E.1B. 2C2. netC: 2<&.1B.E:C2.1B.