112
Introduction to Basic Packet Networking
A network node is a "collection point" in a packet network. It is at these node locations
that packets are collected and routed to other nodes and end-users. Many network
nodes operate at high data rates, usually at 9,600 or 19,200 baud. End-user access to
the network is usually at a lower baud (1200 baud) and on a different frequency. The
advantage of a network system is two-fold: automatic routing of connections to other
network nodes and increased data throughput. A network node "knows" how to
automatically route data through neighboring nodes to distant nodes on the network.
They do this by "hearing" other nodes broadcast this information. This means that a
user connected to a node does not need to specify the particular route to take in order
to reach a distant node as long as the local node contains a listing for that distant node.
Using a network node can also dramatically increase the data throughput to distant
locations. The packet data and the acknowledgments travel from node to node (rather
than from end-to-end), greatly reducing the inevitable data collisions that will occur
when attempting to communicate over long distances.
The defacto standard networking protocol, NET/ROM, was developed in the late 1980's
by Software 2000, Inc. Since then, other node derivatives such as TheNET, TheNET
Plus, TheNET X-1J, and G8BPQ have evolved. Although these variants differ in their
features, they all utilize the same basic networking protocol.
Some network nodes have more than one radio port in order to provide a high-speed
network "backbone" on one frequency, as well as Local Area Network (LAN) access to
the high speed network at a lower data rate on a different frequency. This approach to
packet networking makes good sense for a number of reasons. First, the high speed
backbone is free to pass large amounts of data without competing with end users for the
channel. Generators of large amounts of data are called servers. Many times servers
are busy communicating with other servers, so dedicating a channel to this function is
desirable. Second, users are "shielded" from the server-to-server activity. Many users
could peacefully co-exist with a server if all the server did was to respond to user
queries. User input to a server generally consists of very small amounts of data that
generate fairly large responses. In a local area, this works quite well since all users can
detect the presence of the server's data. Carrier Sense Multiple Access (CSMA) allows
the channel to be shared nicely. However, servers, and in particular many Bulletin
Board Systems (BBSs), spend much more time forwarding and receiving messages
from neighboring bulletin board systems than they do actually serving the local user
population! It is not unusual for a BBS to attempt forwarding (and reverse forwarding) on
an hourly basis. This situation quickly evolves into quite a paradox - BBSs provide an
extremely popular service that has been a boon to the growth and popularity of packet
radio. As user numbers increase to take advantage of the services a BBS provides,
BBS systems also increase in number in order to provide the service to more and more
users. More and more BBS systems must now forward to each other, generating huge
amounts of data on the channel, leaving precious little time for the very source of their
existence - the user community. User-to-user (keyboarding) QSOs become, at best, a
very frustrating proposition. As a result, many users become ex-users, interest declines,
and packet usage stagnates. With the K-Net function enabled in your KPC-3, cost-
effective networking can benefit packet users in your area immediately.
To Next Page
Loading...
