(Wireless) Connectivity from the Edge
While I applaud the idea of municipal Wi-Fi in spirit, in practice it is problematic and we have a far better alternative in connectivity from the edge by taking advantage of existing paths and building from there. The danger is that these well-meaning efforts will give us more of the past rather than the opportunities inherent in a dynamic and ever-changing Internet.14-Dec-2006

We should view connectivity as part of cities’ fundamental infrastructure. But we need to be careful about our metaphors and not view the implementation as a grand project similar to building a water system. We must be careful to avoid the notion of the city being the exclusive provider and instead, in the spirit of the Internet it can and must come from the edge with each of us contributing to create the whole.

The Internet is not a service that is delivered; it is something we create by contributing to Our Internet.

I’m enthusiastic about wireless connectivity but believe we can do far better by taking control from the edges of the network rather than modeling as a service in the style of broadband and the earlier digital phone networks.

We already have abundant connectivity – at least potentially. But we have it locked down into exclusive “broadband pipes”. Rather than building another service delivery system we should take control of what we already have. It’s shameful we have no ownership to our basic infrastructure. It’s owned by carriers so who use it to sell us services. Broadband is just another product priced to maximize their revenue even it is creates an artificial barrier that keeps it out of reach for vital city services and for discovering new possibilities and value.

Rather than framing municipal connectivity as just another exclusive service delivery platform we have to make existing capacity available and add to it.

I was once an enthusiastic supporter of municipal connectivity but now I fear it. The basic dynamic of the Internet is one of taking advantage of opportunities and discovering what’s possible and this can lead to conflicts with the model of the city as a service provider. The cities have an important role as a contributor rather than exclusive provider.

This is very good news for we already have the elements of wireless connectivity in place just like we had modems in place in 1993 when we discovered the Web. Today’s broadband is not really in the spirit of the Internet but we can take advantage of it and build out from this transport to provide connectivity without additional investment and thus without making any commitments. We can also take advantage of the existing cellular network. Even if the price is high it gives some of us a chance to learn about mobile connectivity. My HP-HW6945 is not just a phone – it’s my portable device with telephony as an application.

If we look at this real map (of an office park in Waltham Ma) we find there are already many open (green) access points available. This doesn’t mean they all provide connectivity with the rest of the Internet but it does give a sense of what is possible.

Today’s access points are problematic because they can expose internal network traffic. But we can address that problem by improved NAT software which partitions segregates internal traffic and perhaps gives it a preference. VPN technology allows mobile users to maintain safe and stable connections to the rest of the Internet without depending upon meshes or other centralized implementations. Of course such approaches are transitional. Over time we can develop protocols that are native to this dynamic view of the Internet. This approach is consistent with cities providing amenity access in some public spaces and buildings just like hotels and some restaurants do now.

The problem comes when we want to do more than that and ask a city to make a major investment in providing coverage.

One problem is that the Internet is not a network like the phone network or broadband. It’s just the result of our taking advantage of opportunities and discovering what’s possible. The Internet was created to connect, or Inter-network LANs (local area networks) using links that were often much slower than these networks. Users also joined in using very slow dial up connections. Initially this hodgepodge was good for little more than email and transferring modest sized files. Since we had little control over the path any packet would take there was no way to add special capabilities for particular applications. Each time people try they discover that it can’t work because no one has enough control to make promises for the network as a whole. Wi-Fi Map

If we promised that voice would work in 1995 we would have had to turn the Internet into a requirements-driven infrastructure like today’s very expensive phone network. We would have and, in fact, do have, today’s broadband. Local connectivity is limited to the speed and pricing of the backbone. It’s as if your home network were limited to the speed of your upstream connection rather than the gigabit speed of wired networks and 100 megabit speed of today’s wireless networks.

We can look at France’s Minitel which led the world when it was deployed in the 1980’s but by the 1990’s it delayed the adoption of the Internet in France.

Perhaps the biggest question for Muni-Wi-Fi (MWF?) is why – what are we trying to achieve and at what price? I applaud projects like New York City’s plan to provide amenity access in public areas like Central Park without having to make specific promises.

Note that I said “plan to” – the actual implementations of these projects tend to be difficult because cities tend to be process heavy – to say it nicely. This alone would be reason enough to be skeptical even if there were no other issues.

But such amenity access is not the same as providing complete coverage with promises of adequate performance to deliver on promises. When we ask the city to make a major investment we have to justify the cost and provide funding.

Even if a city is willing to simply do what’s necessary to provide complete coverage as an amenity we are still in a situation that is little different from taking advantage of available access points. We still need to be mobile when we travel between cities and even within the city we need to move between the city’s coverage and local coverage. One big advantage of using IP for connectivity is that I can extend the coverage – for cellular I’m dependent upon the carrier for extending coverage.

If the city is the exclusive provider then we’ve taken a step backwards.

We then need to ask why – what is the purpose of MWF? It’s certainly not for people walking around browsing on their laptops looking at web pages.

The cities are up against virulent form of the Christensen’s The Innovator’s Dilemma. Not only are they constrained to deliver what is asked for, they typically try to deliver it as asked. Thus we should expect a “wireless” task force to eschew solutions using wires even when that’s a far better approach. In fact, making the absence of a wire a requirement misses the point of the Internet and connectivity.

We often hear about the “digital divide” as a rational but that doesn’t make sense. Why spend a lot of money to build an entirely new infrastructure when we already have existing “broadband” infrastructure that we’ve paid for? It makes a lot more sense (and dollars) to simply share an access point among nearby houses and even not so near if we have a relay. More likely we are dealing with MDUs (multi-dwelling units, AKA apartment houses). It makes a lot more sense to provide access by taking advantage of the aggregated demand and using existing wiring and wireless to share that connection. Perhaps the focus on Wi-Fi does damage and prevents providing vital access in more effective ways.

One approach for MWF is to give franchise rights to a provider who can then deploy access points and sell subscriptions. But that’s just a recreation of the cellular model and doesn’t give us connectivity – we have to be very aware of the billing path each packet takes. Any billing model creates a price hurdle for applications –how much do we want to pay for medical monitoring or monitoring parking meters?

Unless we provide abundant connectivity we can easily find ourselves trying to prevent “abuse” (and, worse, defining abuse) in order to keep explicit or implicit promises. Once we start monitoring usage either to limit use or to enforce policies there’s the risk of creeping liability – akin to having a responsibility to inspect each car driving through the city.

Despite all these concerns the city can play a very positive role as a contributor to our infrastructure by providing opportunities without making promises. This isn’t to say that the city just provide access and sit back. As people discover the value of connectivity they will demand more. This works because they would also understand the value of having a connected infrastructure. At the same time any organization can contribute to the commons. This is a very positive dynamic.

Perhaps the most important step the city can take is to drop the word “wireless” (as in the DD (Digital Divide) example above). It doesn’t really matter how we provide connectivity. Wireless is just part of the mix. Public housing should be connected as a matter of course. In fact future housing codes should presume connectivity.

The city must start by healing itself and assuring that it not only has connectivity among its own facilities, it must take advantage of connectivity both at a transport level and a policy level. Rather than building a separate network for each purpose it should take advantage of the common infrastructure. At a technical level, for example, rather than community access TV it should be providing video content. That would be far more valuable than funding a toy TV station in return for giving carriers exclusive access to the city’s rights of way.

In fact, why not demand some of the capacity to provide wireless coverage thus having to spend money on yet another infrastructure? The idea that IPTV is unrelated to the internet is offensive and an affront to reality. Yet that is the kind of dissonance which leads us to pay again and again for the same thing while limiting ourselves to the intended purpose for each path rather than finding opportunity in repurposing what is already available.

Untethered access is just part of the mix of connectivity and so is video and telephony.

The hacking and research community should address the issues I’ve cited above as participants in this dynamic. It’s not about fixing the Internet and making it mobile – it’s about taking advantage of connectivity with today’s Internet just being an element of the system.

We have to start viewing connectivity as part of the cities’ fundamental infrastructure. It seems foolish to build yet another infrastructure in addition to the multiple broadband and cellular systems we already have. Instead we should leverage what we have an extend it. We must recognize that connectivity is connectivity. City fiber is just part of the mix and not special.

Most important, perhaps, is working with the dynamic process and discovering what is possible rather than demanding we wrestle the Internet into the ground and make it fit into today’s mold at the price of tomorrow.

Today’s broadband/telecom model is an artifact of the FCC service model. We should instead be providing connectivity rather than services.

We have to be very wary about attempts to give us “more” without asking more “what”? For example FON seems to be about openness yet they are explicit about being complicit in the carrier model. That’s not the same as giving us control.

When we do invest in municipal systems, including those with sufficient capacity to provide police and fire departments with connectivity, we shouldn’t fall into the trap of assuming scarcity. A good model is the Interstate Defense Highway System – a vast system of roads that the military can be used to deploy its resources. We know it better as the Internet Highway System because it was not built for exclusive use.

This is even truer of connectivity – we have trouble imagining how much capacity there really is and the importance of sharing a single infrastructure. If you’ve got firefighters working in a basement they must be able to take advantage of any available connectivity. Better improve all streets than create a parallel network of streets for each purpose.

We should assume that we have abundant common infrastructure funded as a commons rather than as billable services and then work to transition to this model. The cost will be less than we pay now for redundant transports and the capacity will be available via fibers and wires as well as without wires. Without having to contain the bits for billing purposes we will view those providing access as contributing to the common good rather than competing with a provider.

This is the model against which we should measure our city’s efforts. Is the city’s effort consistent with the Internet Dynamic or is at odds with it?

The Internet’s strength comes from discovering and rediscovering its purpose. Public works projects are built for stated known purpose. And that’s the problem.