In this paper we present a candidate routing
architecture for the future IP-based Airborne Network
(AN). The advantages and disadvantages of this
architecture are presented. We focus on the issues of
mobility and the separation of routing domains.
The future AN will consist of multiple IP-enabled
airborne platforms with wireless connections to each other
and to multiple surface network domains. For example,
Air Force platforms may connect to each other and may
connect directly to Air Force domain ground sites. These
Air Force platforms may wish to use Navy platforms as
relays to access Navy ground sites and connect via
SIPRNET to Air Force ground sites. Similarly, the best
path may be to connect via an Army ground site to gain
access to the Air Force Intranet via SIPRNET. Given the
dynamically changing topology and the bandwidth-limited
channel conditions corresponding to airborne networking
versus terrestrial networking, it will be critical to develop
effective link access protocols, routing protocols, and
management strategies which can accommodate the
unique characteristics of the Airborne Network.
Various IP-enabled radios (e.g., TTNT) would be used to
establish these connections. Military satellite links, as
well as various commercial satellite links, such as
INMARSAT and Iridium, would also be available. The
specific IP-radios that will be available for both the air
and the ground nodes of the AN certainly will influence the
eventual choice of routing architecture. We consider some
of these IP-radios as part of our candidate architecture to
ensure that our analysis is consistent with the planned
infrastructure.
