Wireless Community

Icon

How to Bring an Affordable Broadband ISP into New York City

I was recently asked by a colleague how an ISP like Earthlink can provide low cost broadband in New York City. The truth is that without help, they can’t. Let’s look at the two possibilities that exist: using existing infrastructure and building new infrastructure.

*Existing infrastructure*

There exist a handful of wires that enter most businesses and homes: copper/phone, coaxial, and electrical wiring.

Copper cabling is owned by Verizon. Because of an FCC requirement they must provide “common carrier” access at a competitive rate to other ISPs, which means that Earthlink can gain access to the copper and provide DSL. Earthlink would have to pay Verizon an access rate somewhat equivalent to (but probably a little less than) Verizon’s own DSL service, about $30 a month. Earthlink would need to charge at least this much in order to make any revenue on the service, so there’s no way they could underbid Verizon’s service. Earthlink, by the way, offers DSL in New York City at a rate of $19.95 for the first 6 months, and $45.95 per month for the second 6 months, which is an average of $32.95 per month for the minimum 1 year of service.

Coaxial cable is owned by either Time Warner Cable (TWC) or Cablevision, depending on which area of the city you are in. Over cable, Earthlink currently provides a cablemodem service at $44.95 a month. This service is essentially a rebranding of the Time Warner Cable Road Runner service, and as such, Earthlink must pay TWC an access rate that is close to TWC’s own cablemodem service, which is $44.95 per month. Furthermore, due to the recent “Brand X Supreme Court decision”:1, TWC is no longer required to let Earthlink provide this service.

Electrical lines are owned by ConEd. While technologies like Broadband over Powerlines (BPL) are starting to be deployed, they are still in their infancy. Even if such technologies were to be made available over ConEd’s power lines—and there are indications that much of New York City’s power lines and in building wiring won’t support this technology—ConEd the electric company is new to broadband service provision, and it is likely that they either will not allow a third party ISP like Earthlink to offer service, or will charge an uncompetitive rate for providing access.

*New Infrastructure*

Earthlink can build its own infrastructure to provide service. There are generally two components to a broadband service: internet connectivity/backhaul and last mile connectivity. Generally, the last mile connectivity—the line that brings the internet into your residence or business—is the most expensive part. There are options for backhaul that Earthlink can already use, such as dark fiber or leasing communications lines from a number of providers (Verizon included).

Last mile connectivity, in the case of wired service, requires digging up streets and sidewalks, bringing one or more cables into the basement of a building (in New York City, at least), and then bringing that cable up to the apartment or store that is receiving the service. This is expensive because it requires a significant amount of human labor, and because it requires lots of construction and politicking to get permits to do the work. It is unclear what such a network would cost to the end user subscribing to Earthlink service, but it would require an enormous investment on EarthLink’s part to even get 1 person hooked up. Furthermore, with the exception of perhaps fiber, such a network would just be a duplication of an existing copper, coax, or electrical network. Many buildings in downtown Manhattan already have a fiber line, so this might be a starting point, but fiber lines are absent exist from most residences in New York City.

The alternative is wireless last mile service. In this case, Earthlink would have to both select a technology—there are plenty out there that would work, including WiMax and Wi-Fi, and Earthlink is trialing them in other cities—and deploy the antennas. New York City, which is a large land area (303 square miles), has a secondary difficulty that isn’t shared by any other city in the USA: height. You can read a bit more about height issues that New York City presents in a “previous blog post”:2. These issues mean that there would have to be sufficient density of antenna deployment throughout the City in order to provide service.

Wireless solutions, however, are vastly cheaper than any wireline solution. Orders of magnitude cheaper. So much cheaper, in fact, that even a full City-wide deployment might be made available at an affordable (read $20 or so per month) rate to every resident.

Where can these antennas be placed? One option is to use the same deployment strategy that cellphone networks use, contracting with private land owners in the City and mounting them on buildings. This would work, but prices for antenna location rental are high, residents don’t want antennas nearby their apartments (the “not in my backyard” problem that plagues most infrastructure deployment), and there just aren’t enough enough locations. In fact, antenna deployment is so difficult that New York City has “_horrible_ cell phone service”:3. This is the primary reason why DoITT, New York City’s government IT organization has created a plan to lease City light-poles. Earthlink could use City light-poles for their network (they have already suggested that they’d like to), but the cost of doing so is still too expensive and too onerous to make it a viable option, at least under DoITT’s leasing requirements (the reasons for this are saved for another blog entry).

*Solutions*

In the end, the only way to bring affordable broadband to New York is to involve the Government. Either through requirements that prevent existing network owners from price gouging and other monopolistic behaviors, or through programs that enables cheap infrastructure to be created by new network providers. In the latter case, we can solve the problem for one or a couple of private companies by reducing the cost of leasing public space, but this really discounts the value of the investments we as residents of New York City have made in our public infrastructure.

Really, the best solution is for us, as a City, to invest in building more public, shared infrastructure. We have already done so by “laying down lots of fiber cable”:4 (there’s still lots that has gone unused). We should invest in more infrastructure that will enable lots of companies to compete to provide broadband to New Yorkers. This way, we don’t create another monopoly broadband provider, but rather create a means for lots of market competition.

[1]http://www.washingtonpost.com/wp-dyn/content/article/2005/06/27/AR2005062700415.html
[2]http://www.wirelesscommunity.info/2005/06/22/why-mesh-based-wireless-networks-are-ideal-for-new-york/
[3]http://schumer.senate.gov/SchumerWebsite/pressroom/press_releases/PR01333.html
[4]http://www.nysernet.org/about

Filed under: Community Wireless, Mesh, Muniwireless, New York City, Urban Wireless

Why Mesh-based Wireless Networks Are Ideal for New York

Mesh networks are wireless networks, based on Wi-Fi technology, where each wireless “node” or “access point” connects with a number of other wireless nodes. Information flows from node to node, winding up either at its destination or the internet in general. Internet connections, in a mesh-based network, are provided by one or more nodes that connect to the internet directly (referred to as “backhaul”).

bq. Robust and reliable, mesh wireless systems offer multiple points of connection to the network and no central tower. Mesh users can bypass obstacles like hills and trees by using different signal paths. Mesh networks are easily expandable at very low cost, and they have no single point of failure. Mesh networks also feature shorter distances between nodes, which means each antenna can broadcast at lower power, creating less interference and allowing more users to communicate simultaneously. (from “http://www.freepress.net/wifi/guide2.php”:1)

The benefits of mesh technology are many:

* self-organized, dynamic routing and connection
* little or no centralized configuration
* each node is interchangeable with every other node
* overlapping wireless coverage areas ensure that no node is a point of failure
* no wires are necessary, as the network is entirely wireless
* organic build-out of the entire network is possible
* multiple separate networks can be built independently and grow into a single cohesive network
* redundant paths to backhaul
* easy/seamless addition of additional backhaul

The alternative for wireless networks are “hub-and-spoke” networks, so called because they resemble bicycle wheels in their architectural diagrams. In hub-and-spoke networks there is one central distribution node with which all client nodes communicate. This centralized architecture is like the cell phone networks: even when client nodes are close to one another, they must route through the central tower in order to communicate. This creates a single point of failure, and requires significant infrastructure at these tower nodes.

Mesh networks are designed to survive the failure of a single node (and sometimes more than single node). Like the internet itself, mesh networks have significantly more resiliency built in. Its untrue that they are more likely to go down than non-mesh networks; This is like saying that cars with airbags are more likely to crash. The fact is that mesh networks have extra safeguards from failure, and reduce the likelihood of network failure.

Furthermore, since there is built in overlap between the service area of neighboring mesh nodes, if one should go down, service should still be available in its coverage area (albeit at a possibly lower data rate). Because mesh nodes are interchangeable, servicing the network is as simple as plugging in a new node where one has failed. This isn’t nearly so easy to do with a non-mesh network, even one that has a centralized network management system. Mesh networks need no such network management infrastructure, since their management intelligence is built into each node, and is spread dynamically throughout the entire network.

So, what does this mean for New York City? Let’s look at a couple of examples, which “NYCwireless”:2 has had the opportunity to investigate:

*In Building Wireless Network*

New York City has a plethora of apartment buildings, many of which date from the early or mid parts of the 1900′s (and some even earlier). These buildings have electrical and phone systems, but even cable networks are an add on. Creating an additional network is both expensive and difficult for these buildings, since it often requires significant construction to create a complete wired network infrastructure that is safe from the elements. This is the reason why Wi-Fi has been so popular with New Yorkers — cable and DSL modems don’t often reside where people want to use their computers.

Using a mesh-based wireless network, such buildings are easy to “light up”. Traditional hub-and-spoke wireless networks rely on a distributed wired infrastructure (like cable and DSL networks), whereas the only requirement for a mesh network is electricity. Using a centralized backhaul located in one or a few places in the building, and inexpensive mesh wireless devices, an entire building, servicing 100′s of units, can be brought merely by plugging in the nodes at appropriate locations (3-10 per floor, on average).

Since most buildings have thick walls separating apartments or floors, if there is a spot in the building where the wireless signal is weak, just add another mesh node nearby. And since the nodes are usually powerful enough to communicate between adjacent floors, the network will have multiple routes up and down the building, ensuring that even if part of the network goes down, the rest of the network will still work.

Distributing a wired backhaul (T1 or fiber internet connection) throughout a building is cost effective, but it may not even be necessary all of the time. If there is another nearby building with lots of available bandwidth — like a school, library, or even business — setting up a higher-powered mesh node at that location will enable the extra network capacity to be wirelessly beamed into the building through the windows. With this type of network, there might not even be any wired internet at all within the building!

*Business District Network*

New York City is filled with Business Districts. Whether you live on the south side of Staten Island, Central Harlem, or along Northern Boulevard in Queens, there are plenty of areas where small businesses (and sometimes large ones) line the streets for blocks on end. Especially for small businesses, bringing in an internet connection can be both costly and time consuming (waits of months for a T1 are not unheard of). Considering that many small businesses would rather place their resources into running and building their shops, spending all of that time and money, month over month, on simple internet access can be a wasteful investment.

Creating a wireless hotspot all along a business district benefits residents, businesses, and visitors to that area. This is the primary reason why the “Alliance for Downtown New York”:3 partnered with “NYCwireless”:2 to help create its “wireless network in downtown Manhattan”:4, and why other Business Improvement Districts (BID) and Parks have worked with “NYCwireless”:2 in similar capacities. Creating a hotspot that exists outside (throughout the cold New York winter) is hard enough. Creating a hotspot, using traditional non-mesh wireless technologies, that is larger than a standard City block is even harder. Consider that such a hotspot likely needs more than one DSL connection, that its hardware must be mounted outdoors, and that if a local business is a participating host for all or part of the network, that the network becomes dependent upon that business staying in that location for years on end.

All of these difficulties are easily solved by using a mesh network. Instead of mounting the hotspot access point and large antenna on the top of a short building, the multiple nodes of the mesh network can be mounted inside or next to local storefronts, lining the street. Backhaul can reside inside of a building where it is easy to get to, and more DSL lines can be added as needed. Since there are more mesh nodes than the few traditional access points, coverage is better.

The best part of this architecture is that the business district network no longer is reliant on one or a couple of businesses. Instead, the backhaul can be provided by a long-standing resident of the area — most likely a BID or City organization — and the network can flow down the street. As local businesses come and go, the BID can merely unplug a mesh node from a departing business, and plug it in next door. The network continues to operate, and service never has to go down for more than a few minutes (if even that). In this type of community architecture, a traditional hub-and-spoke network would be unmanageable.

*City-wide Network*

New York City represents a great challenge for any wireless network. The dense metropolitan area covers about 240 square miles, and much of that area is covered by buildings 20 or more stories in height. New York is truly a huge, three dimensional city.

Creating a large scale wireless network throughout New York represents a great challenge — one that “NYCwireless”:2 would love to help solve. In all likelihood, any solution that makes this endeavor tractable will need to make use of multiple network technologies, and combine network deployments from a number of providers. To be sure, New York will require more ingenuity than that of Philadelphia’s, San Francisco’s, and Boston’s wireless city plans.

Here again, mesh networks can play multiple roles. Nodes can disburse wireless internet backhaul from the city’s dark fiber and existing excess bandwidth via a mesh network mounted on lamp posts. That network can feed separate mesh networks that draw the internet up into and throughout a building like a tree drinks water. And people and businesses can move their lives and their livelihoods from building to building without causing network disturbances.

With the help of mesh networks, New York can become a living, organic city, whose lifeblood is the packets of information that flows freely from point to point, person to person, bouncing around automatically finding its way to and from the internet.

[1]http://www.freepress.net/wifi/guide2.php
[2]http://www.nycwireless.net
[3]http://www.downtownny.com
[4]http://www.downtownny.com/?sid=49

Filed under: Community Wireless, Mesh, New York City, Urban Wireless

Twitter

Follow

Get every new post delivered to your Inbox.