Led by Dr Klensin and Marie Zitkova, the .aero manager, the workshop explored how Internet technologies can help create an integrated interline environment. Particular attention was paid to the latest developments in DNS technology as well as the security considerations of using the public Internet for business applications, the concept of the industry naming structure in the .aero domain and the general potential of the Internet for air transport specific applications.
With the evolution of online technologies, the industry transition to Internet-style solutions is accelerating. Airlines find themselves with increasingly computerized systems for ticketing, itinerary and passenger tracking, baggage processing etc. To be effective, all of these systems must deal with each other on an interline basis.
IATA's recently announced Simplifying the Business initiative recognizes the potential of online technologies to achieve lower costs, higher efficiency and improved passenger services. For each initiative, innovative and coordinated use of Internet technologies will be the key to exploiting the full potential of the technology and avoid the associated inefficiencies of systems that grow through "add-ons".
Practical implications
Dr Klensin explained that the Domain Name System (DNS) is an important building block of the Internet. It is a distributed database, which maps structured names to "things". A wide variety of information can be distributed using various records defined in DNS. The records may contain host IP addresses or email server addresses – as well as VoIP addresses, phone numbers, fax numbers, individual e-mail addresses, URIs which identify available services, or even public cryptography keys.
The database is strictly hierarchical, so that administration of each node within the "tree" can be fully controlled by the "owner" of each node independently of the actions of others. The information contained in the DNS database itself is publicly available to all users of the network although distinctions can be made between the public Internet and various private networks with regard to what information is available.
These three concepts have important practical implications. The owner of a domain name can configure all information relevant to the domain (such as VoIP, e-mail, host address, URI for web service or public key) without involving a central authority. This information is then automatically distributed by the DNS.
In terms of reliability and security, the Internet is no worse than private networks but requires users to employ a different approach to ensure the same degree of reliability and security. For example, with private networks, users tend to assume that the presence of a message on the network indicate that it is authentic. That assumption may be less valid as private networks evolve from dedicated wires to shared fiber and virtual network arrangements. With the Internet, the user must take responsibility for end-to-end authentication of all messages and ensure that only certified messages pass through to the application. Cryptography based on public/private keys is the only 100 percent reliable technique, assuming the secret keys are properly managed. Others, such as hiding names, using secret addresses, or restricting access to certain addresses, are generally less effective or suffer from poor scaling.
Possible applications
"The idea behind .aero is to facilitate industry transition from a single centralized network to an Internet environment by developing a predictable naming structure for industry systems to communicate with each other." Explains Marie Zitkova .aero Business Manager.
"A number of business applications that might benefit from DNS technologies and structured domain names were discussed at the workshop such as."
Interconnection of VoIP systems between airlines. Domain name holders could configure a VoIP address associated with a given predictable name so that callers can use their VoIP application to place a call. For example gva.dcs.lh.aero zone could contain a VoIP address for Lufthansa's departure control at Geneva Airport.
One domain name could be used as an identifier for an entire array of services, since multiple records of different types can be associated with one domain name. The same name (such as gva.bag.lh.aero) could be used by different systems to locate a phone number, fax number, e-mail address or web service URI relating to baggage processing in Geneva Airport.
An e-ticketing application assumes that each airline has a database of all tickets and that each airline using the ticket at different stages updates the source database. Currently, these databases are held at large airlines or in interconnected hubs. Application of a standard naming convention to access these applications could allow more airlines to use independent solutions, while retaining the same "name" used by other systems.
There are already DNS applications in place that "parse" RFID values to locate a system that supports data provided by the manufacturer. Baggage management systems may follow the same method of data processing as interline e-ticketing – storing all data relating to a bag in the database of the originating airline. In this scenario, the industry application could use the same technology to parse bag identifiers (RFIDs or barcodes) to locate the database of the originating airline. The same processing logic could apply to ticket identifiers or any other identifiers used in the industry.
The DNS could be used to distribute public keys needed for secure communication between airline application systems, otherwise a very complex exercise. Details of this concept were presented by .aero to IATA's Information Management Council (IMC) in November 2004.
The wider picture
The workshop participants supported proposals to use DNS based naming structures, providing the technology is considered in the wider context of industry communications standards and protocols, as well as the deployment of new technologies such as XML.
It was felt important to approach this activity as a development of a framework for intra-industry communications in the IP-enabled world rather than a "push" for more use of DNS technology. For example, while a majority of airlines have moved to an IP network environment, some airlines continue to rely on legacy networks. So from a community perspective, development or standardization of messaging formats, naming schemes or communication mechanisms should also provide for transition from and translation to Type B. At the same time, planning for the future should avoid the expensive "two transitions" trap of simply migrating old formats and ways of doing things to the new environment, only to be faced with redesigning to take advantage of the other special properties and strengths of the Internet later.
It was also felt that a smart approach to deployment of the new technology and its related standards would help to close the gap between small and big airlines – a useful part of the overall value proposition.
Note: A fuller version of this report is available on request simply by using the .aero feedback form or through contact us.