Exploring New Network Architectures
Deconstructing the End-to-End Argument in terms of Changing
Economics, Technology, and Policy
This issue examines changes in Internet architecture and business case approaches used by Vistabeam, Matt Larson's rapidly growing WISP, with coverage in Nebraska, Colorado, Wyoming, and Montana. It demonstrates a new found strength on the part of the fixed wireless ISP, a business with low cost of entry that is particularly well adapted to those who can develop a close knowledge of their local economy and create innovative and cost-effective solutions to their local telecommunications problems.
Part 2 uses a discussion from our Arch Econ community of interest toexamone continuing development of an as yet relatively unknown architecture that the proponents of which contend may yield a pathway to a more stable and secure foundation for global TCP/IP architecture. We speak here of RINA (Recursive Internetworking Architecture).
This issue examines changes in Internet architecture and business case approaches used by Vistabeam, Matt Larson's rapidly growing WISP, with coverage in Nebraska, Colorado, Wyoming, and Montana. It demonstrates a new found strength on the part of the fixed wireless ISP, a business with low cost of entry that is particularly well adapted to those who can develop a close knowledge of their local economy and create innovative and cost-effective solutions to their local telecommunications problems. In the states of the Rocky Mountains and Great Plains serve by Matt a fixed wireless WISP can deliver a triple play every bit as good and more cost-effectively than the three great vertical monopolies Comcast AT&T and Verizon. After a gap of about five years Matt updates us on the advances in technology that have enabled him to open new markets in his local territory by bringing hitherto unimaginable speeds into the most remote landscapes.
The second part of the issue presents a high altitude survey from our Arch Econ list in January and February 2017 of the changes in the overall Internet inherent in the creation of a purely digital PSTN. While the process has been extended and drawn out, it now looks very much ready to be brought completion.
But, seen in the context of the economic and business models of telecommunications, it has given birth among the Arch Econ community of interest to a fruitful ongoing discussion that sheds light on the continuing development of an as yet relatively unknown architecture that the proponents of which contend may yield a pathway to a more stable and secure foundation for global TCP/IP architecture. We speak here of RINA (Recursive Internetworking Architecture).
Simplicity, it seems, wins. This goes all the way from Matt Larson's wisp where he has a toolbox of technology-based economic solutions that he applies to solve communication problems in the immediate geographic area that he serves. We see the technology that he uses becoming flexible enough that he can put together and deliver, in a short amount of time, technology solutions that do a superior job of handling his customer’s needs when compared the solutions offered by vertically integrated telcos.
Given what is available to Matt, he can develop and implement cost-effective solutions to his customer problems that are much superior to what the AT&T, Verizon, and Comcast of the world have to offer. In expanding into a customer base once served by the Universal Service Fund, he has prepared his company for the requirements of the Connect America replacement and his problems now flow primarily from regulatory issues involving the implementation of the predigital PSTN to the question of how a TCP/IP-based PSTN is put in place.
Some terminology: Noting the difference between fixed an mesh wireless. If you have a fixed wireless network, that is to say what you call a wireless Internet service provider, is the only alternative is a mesh wireless network very likely using the protocols of guifinet? But even those mesh networks have super nodes. The mesh generally fills in between the super nodes and super nodes generally connect to the Internet via some kind of wireline connection unless a super node is directly at an Internet exchange point.
Some call it middle mile conductivity. So how does it work? Customer premises radios connecting to access points? And access points connecting in turn to “the connections between towers on the backbone” and finally, depending on geography, a middle mile tower connecting to a fiber backhaul usually found at the edge of the fixed wireless Internet service provider territory. Matt explains the remaining details: “We will carry a microwave signal up to 120miles to get back to a fiber point, but we generally do that over our own radios and also maintain connections going in many different directions for redundancy. Matt adds that the fiber proponents make it sound like there needs to be fiber to every wireless tower location, and that is not true. Once we have a fiber connection in a region, we can deliver about 500MB within a 50-100 mile radius using licensed microwave between our towers. If we have a fiber connection into a more densely populated area, we can deliver gigabit backhauls to our towers and end users using millimeter wave and 24ghz.”
Looking to where things are headed in a few years, thanks to the recursive nature RINA, we conclude that the new technology will likely be too modular and hence way too flexible to support any longer the old top-down end-to-end argument of the TCP/IP world. Finally, we also offer a lengthy edited discussion of RINA (Recursive Internet Architecture) from our Arch-Econ list. A highlight —
Goldstein: The FCC banned retail inter-networking, at least as we know it, by turning the Internet into a Public Packet-Switched Network [PPSN] or (Super-Minitel). Pai might undo that somewhat, though by not requiring monopolists to make their facilities available, there's nothing to build a decent internet with.
Nonetheless it is possible to run RINA recursively, if necessary, on top of whatever PPSN exists, in order to build additional internets. It won't have any of the good QoS abilities if it's stuck atop the big-I Internet, which only offers BESQR(“best efforts”, scare quotes required) transmission, but QoS could be managed over enterprise Carrier Ethernet circuits or properly-engineered (stick to "layer 2" in the radio!) WISP links.
So with RINA there is only one application protocol, CDAP. An application is not a piece of software that does “something”. It is a chunk of information that is acted upon according to the sharply defined verbs or processes of CDAP: namely Read, write, create, delete, start, stop. According to the Github entry in the Pouzin Society’s web pages: “After an application connection is established, the applications can create/delete/ read/write/start/stop” consequently, RINA can transport traffic of any protocol (it is agnostic about this), while the theory behind RINA says that you could construct any distributed application in the world using a single application protocol, CDAP.
And then at the ProtoRINA home page we find:
“RINA is a clean-slate internet architecture that builds on a very basic premise, yet fresh perspective that networking is not a layered set of different functions but rather a single layer of distributed Inter-Process Communication (IPC) that repeats over different scopes---i.e., same functions / mechanisms but policies are tuned to operate over different ranges of the performance space (e.g., capacity, delay, loss). Specifically, a scope defines a Distributed IPC Facility (DIF) comprised of the set of IPC processes, running on different machines, that collaboratively provide a set of well-defined flow services to upper application processes. Application (user) processes can themselves be IPC processes of an upper DIF that is providing services over a wider scope.”
A DIF is best thought as a distributed application whose members have the single task of providing and managing IPC services over a certain scope (point to point link, network, internetwork, VPN, etc.). All DIFs have the same structure and mechanisms, configurable via policies that optimize the operation of a specific DIF for its target deployment environment.
To conclude -- RINA at the most simplistic level may be thought of as a set of rules that describes 1) data 2) what you can do to the data and 3) a tally of what you have done to the data. Finally, because it is recursive, the data carried does not populate a single global namespace where an incursion by a hacker can expose virtually everything, RINA’s advocates maintain that it can be applied on an as-needed basis to applications where security is absolutely critical in a way such that some of the first expected instantiations of RINA will occur in applications where security is paramount.
It seems reasonable to suggest that, although TCP/IP has had a fantastic world-changing run, it has created a global infrastructure that is now plagued by serious problems. In other words there are things that it needs to do that it is having great difficulty doing. And that one of the reasons for this difficulty is the layer and layers of “patches” that have been created and applied in order to keep the whole thing working.
Finally, in Barcelona Eduard Grasa maintains that current prototypes are useful for researchers, academics and R&D teams at companies that want to set up RINA proof of concepts to better understand the potential of the technology. They are definitely not yet products. These prototypes are being improved and tested under the ARCFIRE project.
Executive Summary p.3
Editor’s Introduction - Updating the Fixed Wireless Wisp
Individualized Acts of Problem-solving p. 7
How Fixed Wireless Business Meshes with American Character p. 7
Advice for the Would-be Entrepreneur p. 9
Daily Report of Key Indicators p. 10
Over all Performance Dashboard p. 11
Infrastructure p. 12
Tower Revenue and capacity planning p. 13
Fixed wireless and Sandbox like Freedom to Experiment p. 16
Gigabit Service to Small Towns p. 17
Millimeter Wave Hardware p. 18
Wireless Cowboys p. 20
A Different Example – Casper Wyoming p. 21
Universal Service Kinds of Issues in Canada p. 22
Portable Towers and 100 meg to a Ranch p. 23
Prototype Tower p. 24
From The Arch Econ “list”
The Internet is the New PSTN & it’s nearly Impossible to Alter p. 28
The real story over the last 10 years has been Access technology p. 29
Martin Geddes: Telecoms is dying. Yippee! 1/22/17 p. 32
Telecom no longer a Growth Business but Need to Avoid Concentration of
Flow of Information in Hands of a few p. 34
Wheeler Should Not Have Applied Re-Classification via Administrative Law p. 35
How Does RINA Work in the Context of Shockey’s Law? p. 38
How to Deploy RINA between Offices p. 40
Current Concept of internet Based on Outmoded Point of View p. 40
Viability of RINA Prototype? p. 42
WebRTC Dependencies draft-jennings-rtcweb-deps-16 p. 45
Common Distributed Application Protocol p. 46
The search for simplicity in physics p. 49
Simplicity Wins p. 51
Let’s face facts: we need a new Industrial Internet p. 54
The fundamental architecture of the Prototype Internet is broken, and
cannot be repaired p. 55
This is new ground in the human condition p. 59
Shaping new approaches to capture cost information p. 63
Yet another security issue p. 68
Strategy for dissemination of Paradigm Change p. 69
ARCFIRE is the Third EU Funded Development Program p. 72