For over 30 years, the Internet has been coping with ever increasing traffic and new applications, including voice and video, while retaining its original architecture drafted over 40 years ago.
Despite its enormous success, the Internet is suffering from several key shortcomings that stem from a design that appears increasingly unfit to support current (unanticipated) usage trends.
Some of these shortcomings include…
- The send-receive communication paradigm, which inherently causes an imbalance of powers in favor of the sender of information, who is overly trusted.
- Locator-identifier aggregation, whereby a node’s topological locator (i.e. it IP address) also serves as its unique identifier, leading to severe limitations on mobility.
- Host-centric design, which emphasizes the topological location of content as opposed to the content itself.
- A lack of built-in multicast and caching solutions, resulting in wasted network resources and sub-optimal content delivery.
- A lack of built-in security and other critical functionalities, with most features being added as afterthoughts in response to new usage requirements (e.g. VPN, DNS, MIP, HIP, IPv6, PGP, DHTs, DiffServ/IntServ, BGPSec, DNSSec, IPSec, SSL/TLS etc.)
Some unanticipated usages include…
- Mass usage (billions of devices and users, multiple devices per user and multiple users per device)
- The need for human-friendly naming conventions and interaction
- Mass content storage, retrieval, and delivery
- Secure site-to-site transport, secure user-to-user messaging
- Varying degrees of quality-of-service
- Mobility, multi-homing, multi-casting etc.
- Unified IP architectures for audio, video etc.
The worst consequence is that the full range of possibilities offered by the Internet is not being exploited and trust in its proper operation has been lost.