IPv6 and the Future Internet: Problems CreatedIPv6 and the Future Internet: Problems Created

Changes in how the Internet is used have meant changes to the crucial underlying protocols. That's mostly been good, but there are pitfalls.

We are on our way to implementing IPv6. But this is not the end of protocol changes. The Internet went through multiple versions up to IPv4, which is still in use. The looming address problem--i.e., not enough addresses--was solved by IPv6. But changes to IP also meant changes to a whole range of devices, routers, servers and endpoints, and also to other protocols like TCP, UDP, and ICMP applications, as well as the Domain Name Service (DNS).

The November/December 2012 issue of the IEEE Computer Society has four articles that look at the state of the present Internet protocols and what new problems and issues will be faced in the near future. Three of the articles pose specific problems and suggested solutions, and the first article is a free download. If you are a network architect, then it is worthwhile to download each of the other three articles (for which there is a fee).

The first article, "Future Internet Protocols" provides an overview of this issue and the other three articles. It sets the stage for the other three articles by reviewing Internet history and some of the old predictions.

Ironically, even as IPv6 was solving a big problem, the new, vastly expanded address space available with IPv6 creates new concerns as well--specifically the potential for malicious exploits against this huge new address space. There is also a need for better control of quality of service and the delivery of larger packets.

Part of the need for a vast new address space was the emerging "Internet of Things". In case you're unfamiliar, the "Internet of Things" is the idea that everything that has power--AC, DC and battery--will be connected and drive even more Internet traffic. This will also stimulate development of new protocols, mostly for applications that are semantically rich and very useful for humans to interact with devices and each other.

One of the comments in the article is: "In the face of such overwhelming change, we can easily see that the protocols underlying the Internet that were developed in the last century won't be sufficient, or at least not efficient for the next".

Another article, "Designing a Deployable Internet: The Locator/Identifier Separation Protocol" (LISP) discusses the issues of dealing with billions--not just millions--of devices all connected to the Internet. The article offers a detailed discussion of LISP, a protocol that aims to provide greater flexibility within the Internet architecture; it does this by decoupling two key elements that are combined within the IP address: The user's identification and location. That, in turn, supports multi-homing, virtualization, and host/virtual machine mobility.

The next article, "IPv6 Deployment and Spam Challenges" focuses on spam and the possibility that the larger IPv6 addressing space could make this issue worse rather than better. We all have encountered the e-mail spam problem. The article discusses solutions that IPv4 has introduced to reduce spam. But the vast addressing capabilities of IPv6, such as temporary addresses, could wind up contributing to spam. This makes confronting and solving the spam problem even more difficult.

IPv6 poses a big issue for ISPs because they will have a difficult time distinguishing between good IP addresses and those addresses that are known to generate spam. Spammers can use temporary IP addresses, thereby nearly eliminating the ISP's ability to blacklist IP addresses.

Finally, "All-Weather Transport Essentials" discusses the original design of TCP and UDP and how the Internet world has changed. The authors of this article present arguments for updating the design of the Internet protocols, especially TCP. The proposed approach does not focus on reducing overhead, which has traditionally been an issue when TCP retransmissions proliferate. In fact as Internet speeds increase, protocol overhead becomes less of an issue.

The proposed approach instead favors moving some of the policies that govern the Internet from the transport protocols to the application protocols. This would have less effect on the network and would bring more changes to the endpoints and their applications.

Another free paper worth reading on this subject of improving the operation of TCP is "Next Generation TCP, Accumulated Acknowledgements on advanced TCP Implementations".