Exercise 9 (21.24.3.) (Tanenbaum pp. 374 -424)

1. Answer shortly the following questions ("review type" problem: answers are usually found in lecture notes or in Tanenbaum's book).
a) Why is flow control much easier than congestion control?
b) Buffers of a router are overflown. Which packets will be destroyed, the arriving ones or some others?
c) What is tunneling?
d) Why a firewall has both routers and a gateway?
e) Both IP and ATM protocols count checksums only for the header? Why for the header and why only for it?
f) Is fragmentation needed in concatenated virtual circuit internets, or only in datagram systems?
g) Why are ARP and RARP protocols needed? What is common to both these protocols? How do they differ from each other?

2. a) What kind of errors the network layer has to prepare for? Is it possible to have the data link layer or transport layer to take care of these errors?
b) What all things have to be negotiated when establishing a virtual circuit across many networks? Explain in each case what happens when wishes/ requirements are not compatible?

3. A PC is connected to a 6 Mbps network, where token bucket method is used for congestion control. Tokens are generated at a rate of 1 Mbps. The token bucket is initially filled with to maximum capasitity of 8 megabits. How long can the computer transmit at the full 6 Mbps. (The answer is not 1.33 s!)

4. a) Tunneling through a concatenated virtual circuit subnet is straightforward: the multiprotocol router at one end just sets up a virtual circuit to the other end and passes packets through it. Can tunneling also be used in datagram subnets? If so, how?
b) A person with a portable computer travels to Berkeley and to his surprise finds out that there is a wireless LAN. So he doesn't have to plug his computer in, but can he use it? If he can, how is it done?
c) In Berkeley there is also a GSM network. Supposing the local operator has a roaming contract with some Finnish operator (calls from one operator are transferred to another). Explain how a travelling person can read his email from Berkeley, or is it not possible?

5. a) Convert the IP address whose hexadecimal representation is C22F1582 to dotted decimal notation.
b) A class B network on the Internet has a subnet mask 255.255.240.0. What is the maximum number of hosts per subnet?

6. a) Describe, in suitable detail, how IP fragments are reassembled at the destination.
b) Some fragments can get lost in the transmission. For this reason, the reassembly of fragments uses a timer. In case of timeout the unfinished packet is destroyed. Suppose a datagram is fragmented into four fragments. The first three fragments arrive, but the last one is delayed.
Eventually the timer goes off and the three fragments in the receiver's memory are discarded. A little later, the last fragments arrives. What should be done with it?
c) An IP datagram using the Strict source routing option. has to be fragmented. Do you think the option has to be copied into each fragment, or is it sufficient to just put it in the first fragment? Explain your answer.