HIGH SPEED NETWORKS
B.E / B.Tech Degree Examination Nov/Dec 2009
ANSWER ALL THE QUESTIONS
PART A – (10 * 2 = 20 marks)
1.What does the term ‘asynchronous’ indicate ATM networks?
2.Mention the encoding techniques followed in Fast Ethernet.
3.Mention any four important parameters that are associated with a single server queue.
4.Write short notes on Committed Information Rate in Traffic Rate Management.
5.Write down the expression by which the RTT is computed in the original TCP implementation (i.e., without any refinements).
6.What is meant by silly window syndrome?
7.Define Quality of Service in a data network.
8.What is meant by Active Queue Management Scheme?
9.Is RTP an application layer protocol or transport layer protocol? Justify your answer.
10.What is meant by Forward Equivalence Class in MPLS networks?
PART B – ( 5 * 16 = 80 marks )
11. (a) (i) Explain the various cabling schemes and the encoding scheme followed in Gigabyte Ethernet. (8)
(ii) Discuss the relevance of CSMA/CD in Gigabyte Ethernet. (8)
(b) (i) Explain the action that take place in the receiver upon receiving the HEC field in ATM networks. (8)
(ii) Explain how the GFC is used to provide the flow control at the UNI of ATM networks. (8)
12.(a) (i) An engineering firm provides each of its analyst with a personal computer, all of which are hooked up over a LAN to a database server. In addition, there is an expensive, standalone graphics workstation that is used for special purpose design tasks. During the course of a typical 8-hour day, 10 engineers will make the use of the workstation and spend an average of 30 minutes at a session. Manager is satisfied with this arrangement since the utilization factor of the work station is only 5 hours out of 8. The engineers complain that the wait time for using the work station is long, often an hour or more, and are asking for more workstations. Explain the queuing analysis that should be done by the engineers to convince the manager. (10)
(ii) Explain the choke packet congestion control mechanism. (6)
(b) (i) Consider a LAN with 100 personal computers and a server that matintains a common database for a query application. The average time for the server to respond to a query is 0.6 seconds, and the standard deviation is estimated to equal the mean. At peak times, the query rate over the LAN reaches 20 queries per minute.
Answer the following questions.
(1) What is the average response time ignoring line overhead? (2)
(2) If a 1.5 second response time is considered the maximum acceptable, what percent growth in message load can occur before the maximum is reached? (3)
(3) If 20% more utilization is experienced, will response time increase by more or less than 20%? (3)
(ii) Discuss the issues of fairness. Quality of Service and reservation in traffic management. (8)
13. (a) (i) Explain the various congestion control mechanisms followed in ATM networks. (8)
(ii) Discuss the ABR and GFR service categories in ATM networks. (8)
(b) (i) Explain the Additive Increase Multiplicative Decrease behavior of TCP congestion control algorithm. (8)
(ii) Explain the basic TCP flow control mechanism in detail. (8)
14. (a) (i) Discuss the motivation and design goals of Random Early Detection scheduling policy.
(ii) Explain the Random Early Detection algorithm in terms of the estimation of queue size and determining packet discard.
(b) (i) Describe the functional modules that should be present in ingress and egress routers of Differentiated Services Networks. (8)
(ii) Describe the components of a core router of Differentiated Services Network. (8)
15. (a) (i) Explain the soft state based receiver initiated reservation in RSVP. (8)
(ii) Discuss the flow specifications and filter specifications in RSVP. (8)
(b) (i) Discuss the labeling and label switched path of MPLS networks. (8)
(ii) Explain the stacking rules followed in the label switched path of MPLS networks.(8)