INT21CN Computer Networks

Tutorial #15

1. Modems are sometimes said to "convert digital data into analog form". This is highly simplistic. Give a somewhat better description of the basic function performed by modern modems.

2. The fastest external modems which can be purchased at present operate at 56Kbps, yet the serial ports on most home computers are set up to operate at 115Kbps or faster.
   1. Why the disparity?
   2. Why does the serial port operate at a "weird" speed like 115200bps?

3. What is a null modem and why is it sometimes needed where RS232 interfaces are used to built point-to-point data links? Describe briefly the connections required in a minimum RS232 null modem (ie: one which uses only pins 2, 3 & 7).

4. What is the link efficiency (or utilisation) in an asynchronous system which sends 8 bits of data with one start bit and one stop bit? What if the data was only 7 bits, as in ancient ASCII data links? How many 8-bit bytes per second can be transmitted using a 28.8kbps modem (ignoring the possibility of compression)?

5. The Department of Information Technology at Bendigo has a dial-in router (for staff use only, sorry!) which is connected to subnet 20, ie 149.144.20.0/24.
   1. When registered users dial in to this router, what would you expect the network/subnet part of their home machines IP address to be?

   2. The dial-in router is connected to subnet 20 and has IP address 149.144.20.20. A separate router (149.144.20.254) provides the link to the university backbone from subnet 20. Draw a labelled sketch of this portion of the network.

   3. (Optional research question -- you do not need to understand the answer to this part!) Now, consider when a dial-in user connects to the dial-in router. A route must somehow be established to their machine, and announced to the backbone router. The two ways in which this can be done are by the use of "proxy-ARP" or by the dial-in router "announcing a host route". Attempt to describe how each of these might operate.

6. In a dial-in situation, IP addresses are usually "dynamically-allocated", and therefore different for each dial-in session. It's usually possible to pay a somewhat higher rate but have a "static" IP address, which is the same for every dial-in. Why do you think dial-in accounts with dynamically-allocated addresses are cheaper?

7. How are IP addresses normally allocated in the situation where a point-to-point link is used to connect two routers together? Can you imagine a more address-space-efficient way of allocating these addresses?

8. What is the function of PPP in a data link using modems?

9. Research Question: when a dial-in Internet user connects to an ISP using PPP, their machine has to somehow discover its own IP address. How do this think this happens? Harder research question: in the olden days, when SLIP was commonly used for dial-in access, the protocol provided no support for allocating an IP address to a remote user. How did the dial-in machine discover its own IP address using SLIP?