Presentation 2

The Future of Dynamic Circuit Networking
A pleasant residual of the dot-com boom is that many parts of the world were left with high-speed fiber optic cables that link many points together. This glass-based system of transport has capabilities of supporting transmissions as fast as 1,000 Gb/s and is 1,000 times larger than the total radio bandwidth of planet Earth. Unfortunately, fast optical networks are significantly bottlenecked at routers that are reading and evaluating every packet of information that is sent through them. Although throughput of routers has increased significantly over time, they are still a far cry away from being able to process information at fiber optic speeds of a Terabyte per second. Internet2 is paving the wave of the future to full utilization of fiber optic transmission lines through the creation of the Dynamic Circuit Network (DCN). Using IPv4 or IPv6 protocol, this revolutionary re-design of packet transmission will the first step of a complete re-structuring of the world’s data network as we know it, discussed throughout this paper.
In traditional IP networking, the packet transmission of data is randomly sent over data lines (this packet takes this route, that packet takes another route, etc.). This mechanism necessitates that every packet generated and sent has to: a.) Contain origination and destination information, b.) Be evaluated by each router along the path, and c.) Be deconstructed at origin and then re-assembled at the destination once all packets have arrived. This has proven efficient in the past for redundancy, in that a single downed line would not prevent the transmission of data, but it does not make the best use of resources available. Conversely, in DC networking the entire path from origination to destination is pre-determined before any information is sent. This virtual circuit is dynamically created for the transmission of the defined data. The circuit is established and reserved for this data, the data is transmitted directly from origin to destination using this path, and the transmission line is freed up after the completion of the transmission.
The technical aspects of a Dynamic Circuit Network can be found in U.S. patent No.: 6,707,820 as well as in the presentation originally done by the author on March 6th, 2008. Moving forward with the research on the DCN and the great capabilities it can facilitate, the author began to wonder if there may be a future for router-less networking. If packet transmission is no longer random and follows a predetermined path, routers would only need be aware of their place in any dynamic circuit created and transmit that data accordingly. They would no longer have to evaluate each packet and determine the best path to forward each one to its next destination. Router-less networking, in its simplest form, simply consists of a cross-over cable between two computers. What if each host in a given geographic area connected to a DC domain, established a virtual circuit to that DC domain? The DC domain relays that virtual circuit to the receiving host, completing the virtual circuit. If all of these entities could be recognized as being in the same segment, then the communication could be created completely without routers. The same as a crossover cable creates a connection between two hosts, that same kind of connection would be created dynamically between two hosts over high-speed fiber optic lines.
Moving even further ahead and discovering very little research information is what this author believes to be the next great emerging technology, even beyond the router-less network: packet-less internet transmission. Some of the greatest emerging technologies have gone back to the past to find their roots. Think of Sky Sails and how they simply modernized utilizing wind to move marine vessels. Packet switching networks provided the means to move data quicker and more reliably than standard analog phone circuits. Now that technology has caught up to the speed requirements of data, we will someday retire packet switching in a way to fully utilize the new dynamic virtual circuits. There will be no need to packetize the data in order to put it on transmission lines, as the data will have the capabilities to be streamed directly from the application layer from host to host. This idea of nullifying the creation and transmission of packets was first introduced to the author in a presentation of the capabilities and future of the Windows Server 2008 Hypervisor for Virtual machines. Currently, each operating system running within a virtual machine must access the network through a software-virtualized network interface. ii Even if the network transmission is not transmitted over any physical lines, it is still broken into packets by the originating operating system and reassembled by the receiving operating system. As the world works towards more server consolidation and the virtualization of servers, the network responsibilities of each guest server will be passed to the host hardware to be transmitted to other hosts. If the Microsoft Engineers are right, and packet-less network traffic can be transferred via this hypervisor that the guest sit upon, why could it not be successfully achieved on a larger scale? Overheads of network virtualization can be overcome using CDNA (Concurrent Direct Network Access), and using this technology to directly access virtual machine NIC drivers via the hypervisor, is really just a small-scale version of what could be created with a large scale, dynamically created circuit. The creator of this circuit would act as a type of hypervisor, from only a network perspective, by having location information of all guests that is under its domain. Machine A makes a request to the DC domain acting as a large scale hypervisor requesting a network to Machine B. The DC domain knows if Machine B is up to receiving requests, and if it is then the DC creates the circuit between Machine A and Machine B and securely manages the transfer at a super high rate of speed. See Exhibit A.

Packet switching and T3 network lines are going to become obsolete in the upcoming years. Internet2 and the Dynamic Circuit Network is going to break out of academia and research and become mainstream, much like Arpanet did. The advantages to be taken of this super high-speed network are going to be plentiful to those brave enough and ingenious enough to take full advantage of it.