Bulletins from the Pacific Packet Radio Society - page 053

The internet sublayer (called layer 3B) is used to connect the individual networks together so data from an individual on one network can be sent to someone else on a different network. This layer is NOT used instead of the network layer, but rather in addition to the network layer. It creates its own headers, and has separate information from the network sublayer.

We recommend use of the NBS Internet Protocol. We do this while realizing that the NBS IP is subject to change on its way to become an ISO standard.

Again, we recommend that static call sign addressing be used at the Internet level. Unfortunately, the format has to be different from those used at the link and network levels.

The call sign is good in one respect - it's unique. It's poor in another - it doesn't designate the location. So, just the call sign would identify the entity but not where it can be found. A look-up table would be theoretically possible, but it's no t practical to have every network node keep a table of every ham in the world, where he is, and how to route to him. In fact, most everything in print says that imbedding the routing in the address is a mistake - it ties that location to a given routing unless alternative routing schemes are devised.

So, a good address would include two elements:

( 1 ) who (unique to the host ham station), and

(2) where (unique designation to the area).

That's crowding things to get both of these things into the NBS Internet address field, but it can be done.

The NBS Internet address field consists of 64 bits overall These 64 bits are subdivided as follows:

The first 16 bits comprise the Digital Network Identification Code (DNIC), further broken down as follows:

4 bits: Zone (continent or ocean)

8 bits: Country (each country has one or more country numbers - the U.S. has 3 now)

4 bits: Network (assigned by the country authority - the FCC in the U.S.)

These 16 bits are organized into BCD quartets, corresponding to the digits 0-9; A-F are not used. The rules for the above DNIC scheme are governed by CCITT X.121.

The remaining 48 bits are terminal addresses and a re to be assigned by the specific network.

Click for Original Graphic Image of this page.