The following questions are an approximate superset of the question
topics that you will
find on the midterm exam. The exam will be mostly short answers with
a couple of essay type questions.
I will continue to add questions to the end of the list, so check
back later for new questions. They will be added to the bottom of the file.
- How many bits will be "airborne" in a STS-48 link between two
ground stations through a geosynchronous satellite
(assuming a 270 msec round-trip time)?. Calculate the minimum RTT given
the distance,
time to send an image, and number of bits in a sliding window sequence
number if each packet is 2Kbytes long.
- If we use x^3 + x^2 + 1 as our CRC generating polynomial and are
sending "0xB6", what is our Q and R? What
is transmitted?
- An STS-1 channel has a one-way propagation delay of 10 ms.
For 1 Kbit packets, what window size is
needed to achieve an efficiency of at least 50%?
- Explain why the ethernet protocol does not scale
to cross-country distances. Be precise.
- Assuming DMA in and out of main memory,
how many interfaces to 45-Mbps T3 links could a workstation-based switch
handle if its I/O bus speed is 800 Mbps and the memory bandwidth is 2 Gbps?
- Know the definitions of the following acronymns: FTP, RFC, HTTP, URL, DAN, LAN, MAN, WAN,
STDM, FDM, RTT, FDDI, IP, TCP, UDP, ISDN, ATM, CSMA/CD, TDM, FDM, RTT
- What are the advantages and disadvantages of statistical multiplexing?
- What are the three kinds of failures that network designers have to
worry about?
- Why is the dealay x bandwidth product important?
- What is jitter? Why would it be difficult to send video over a
network with a high jitter value?
- What are the 7 layers in the OSI network architecture? What is an example of each layer?
- What layer is IP, Ethernet, TCP, email.
- Messages are normally represented by a complex DAG. Why
isn't the message just copied into consecutive memory locations? How much
of a performance hit do you take for each memory copy?
- Contrast the Process-per-protocol and the Process-per-message models.
- Be able to encode a message using NRZ and NRZI encoding.
- Be able to use 4B/5B to encode a message given the table.
- Why are 4B/5B and NRZI used together in encoding data?
- What is the sentinal approach to framing? Given a bit stream, perform
bit stuffing to generate the correct message. Be able to take a stream and
unbitstuff it.
- What are the advantages and disadvantages of Ethernet and FDDI
technologies?
- How does CSMA/CD work?
- How does FDDI recover from a lost token?
- Given a source routing example, determine the destination
that a message will go to. (Something like Figure 3.7)
- Given a virtual circuit routing example, determine the destination
that a message will go to. (Something like Figure 3.5)
- Compare datagrams, virtual circuits and source routing. What
are the advantages and disadvantages of each approach?
- What is the spanning tree algorithm and which links would be eliminated in a problem similar to Figure 3.11?
- Why does an ATM cell have 48 bytes of payload?
- What kind of routing does ATM use?
- What is packetization delay and how much packetization delay is there for an ATM cell?
- Given a banyan network similar to Figure 3.33, decide if there will
be a contention free route for two packets.
- When using a crossbar switch, contention can occur within the fabric.
Would you use input buffering to solve these problems?
- What is the hidden node problem (page 138) and how does 802.11 deal with it?
- Describe the process for selecting an AP with a mobile node (pages 140-141).
- What is an internetwork?
- What is the service model for IP?
- How is fragmentation performed with IPv4?
- Assume a host with an IP address of 222.123.250.3.
To what class does this IP address belong? To what
network is it connected?
- What is an ARP packet and how is it used? How would ARP be implemented
in a network such as ATM that doesnt have broadcast capabilities?
- What does an ARP cache contain? How does the ARP protocol work?
- Is source routing practical? Explain.
- What type of IP address class would be adequate for a
corporation that needs to establish up to 100
sub-networks, each of which with up to 500 hosts? Justify your choice.
- Assume the following routing table:
| Subnet Number | Subnet Mask | Next Hop |
| 128.96.39.0 | 255.255.255.128 | Interface 0 |
| 128.96.39.128 | 255.255.255.128 | Interface 1 |
| 128.96.40.0 | 255.255.255.128 | R2 |
| 128.4.153.0 | 255.255.255.192 | R3 |
| <default> | --- | R4 |
What does the router do with the following packets?
a) 128.96.39.10 b) 128.96.40.12 c) 128.96.40.151 d) 192.4.153.17 e) 192.4.153.90
- What are the problems with stop and wait, and how are they solved?
- How do bridges learn the location of a node?
Created Feb 11, 1998 by Mark Clement
Updated Oct 21, 1999
(clement@cs.byu.edu)