Web of the Future: Building a Faster, More Secure and Egalitarian Internet

The origins of the Internet reach back to the 1960s, when exploration of a technology called "packet switching" was pursued by several innovative thinkers. This area of research got a major boost when the US Defense Advanced Projects Agency (DARPA) invested in the development of the ARPANET. This successful experiment in computer resource sharing led directly to the development of radio- and satellite-based systems and to their integration with the ARPANET into what we now call the Internet. Skipping over 40 years of development and inventions (including Tim Berners-Lee's World Wide Web), the Internet has become a global infrastructure for the support of commerce, social interaction, research and governmental services. An estimated 1.6 billion people use the Internet and probably twice that many may be advantaged indirectly by its services. In addition, there are over 4 billion mobiles in use today, of which an estimated 15 to 20 percent have access to Internet-based services. This trend appears to be continuing and highlights the increasing importance of improving the reliability and security of the Internet's infrastructure.

Written large, this means more than securing the underlying network technology (routers, domain name servers, resolvers, network transmission facilities). The prevalence of e-mail spam, denial of service attacks, viruses and worms are in large measure attributable to weaknesses in browser and operating system software. Overly permissive in their ingestion and execution of web pages, many personal computers and laptops have become infected with viruses and worms that allow these resources to be controlled by so-called "botnet generals" who rent these botnet armies out for a variety of nefarious purposes. We need browsers with better ability to resist infections and operating systems that confine browsers to only those resources needed to function properly. The movement toward "cloud computing" introduces strong motivation for these mechanisms.

It seems clear that a resource that is becoming as important as water, power and highways must be made more resistant to the depredations of malware makers. This is not merely a technical matter. Abuses of the Internet occur daily and frequently on an international basis. In consequence of the supra-national nature of the Internet, law enforcement is challenged by the lack of common international agreements as to what constitutes abuse and what international actions can be reliably expected upon identification of a perpetrator. We can anticipate the need for a kind of International Law of the Net in the future. Such concepts can also provide better support for international and domestic electronic commerce.

Of late, there has been considerable attention paid to the notion of accountability in cyber-space. For some, this means requiring that any user be absolutely identifiable before he or she can make use of this global resource. There are reasonable arguments for and against such a requirement, but it seems useful to have available the means to strongly authenticate one's identity when the need arises. Anonymity or pseudonymity can play an important role especially in "whistle-blowing" situations but equally important is the ability to strongly identify oneself when conducting electronic transactions that are intended to have contractual and legal weight. Digital signatures mean little unless the parties are strongly identifiable.

These thoughts suggest that it is timely to explore ways of improving the security and resilience of the Internet and providing strong means for authentication of its various components and users. Efforts along these lines are already under way for the Domain Name system (using digital signatures to validate the Internet addresses associated with them) and to validate the assignment of Internet addresses to users. Similar methods can secure the confidentiality and integrity of all communication on the Internet. While we are considering avenues for improving the Internet we must also consider improving access to it for all users. Increased availability of broadband access will open new opportunities to applications that need low latency or high capacity to move data around. The open nature of the Internet has served to stimulate innovative uses of the Internet and it seems vital to preserve this aspect of the Internet's services. FCC Chairman Julius Genachowski's leadership in this direction serves the interests of all Internet providers and users and deserves support as a key milestone in the evolution of Internet policy in the U.S.

With over 75 percent of the world's population not yet online, we have much work to do to secure the benefits of the Internet's information sharing capacity for ourselves and future generations.

Comments

[vintcerf]

I would sure like to have the ISPs answer that question. It seems to me that they are in the business of providing Internet connectivity and to do that they need addresses to assign to users. Network Address Translation is a long-term dead end. Moreover, when the IPv4 address space is exhausted in 2011, anyone with an IPv4 address only will need an IPv6 address to communicate end-to-end with anyone having only an IPv6 address. The ISPs sometimes claim no one is asking for this service. That's frankly not true any more. Google has invested two years in the implementation of IPv6 in parallel with IPv4 to be ready for the time when there are IPv6 only users on the Internet. It would help if the ISPs were to move promptly to implement and to provide full international interconnection with IPv6 in parallel with IPv4.l

[NickHP]

Dr. Cerf

Why haven't the ISP's gone to IPV6 yet? In itself that seems like it would help for DOS and spoofing of addresses, such as email return addresses.

Nick Maryland