Information Technology Reference
In-Depth Information
in the simulation: the instructor stipulates during the exercise whether a packet is transmitted
and received correctly, and the system (i.e., the students) responds accordingly.
IP Layer
Packets at the IP (Internet protocol) level are called datagrams. While segments may be
broken into multiple datagrams, experience has shown that this introduces too much
difficulty, so it is not included in the simulation. IP is primarily responsible for routing
datagrams, that is, selecting the route they will take to the recipient. If a packet is destined
for the current machine, IP sends it to TCP, otherwise, IP forwards the packet.
Two key ideas for forwarding are the use of the routing table and routing protocol. The
routing table contains data about the network and, in particular, about the first hop to each
destination. The protocol is the rule that updates the routing table as data about the network
is received. Exchange of network data is handled notionally.
Network Access Layer
The NAL is responsible for converting an IP address into a hardware or MAC address.
NAL also contains error correction data to ensure that frames are correctly received. Again,
error correction is handled notionally.
Physical Layer
The physical layer contains the network interface cards (NICs) and the cables/hubs,
etc., connecting machines.
SIMULATION DESIGN
The design of the game is most easily described by referring to the handout materials
and then walking through its operation. Figures 1 through 5 illustrate the handout materials.
Figure 1 contains a diagram of a hypothetical network. Each is labeled with its IP address and
MAC address. Nodes also have DNS names. Two machines were given duplicate names
(Martha) to illustrate the importance of root domain names. Each of the machines has one or
more network ports connected as shown. On those machines with more than one NIC, all are
assumed to have the same MAC address (another simplification).
The routing table (Figure 2) shows for each source the first hop to each possible
destination. The format is slightly different from a real routing table: * is used as a wild card
in order to simplify the discussion. Each node is assumed to know that the nearest router is
its gateway. Figures 3 and 4 include a port map to determine the port used by each application
and the name map used by the DNS resolver. Figure 5 shows the message form and the header
form. The use of all of these is described below.
OPERATION
The class is asked to separate into groups, each representing a network device. Several
students per device enables them to each play a different layer in TCP/IP, passing messages
between them, thereby emphasizing the nature of a layer protocol. The devices are “con-
