For those who might have missed the story, LUS put out a call for proposals to supply what was described as a wireless network for LUS and city use. No mention of public access was made, though locals familiar with the way that the LUS fiber project evolved from purely utility purposes are reasonably hopeful that a wireless network will evolve in the same way.
The RFP itself is pretty simple as such things go and you have to think that bidders will need to request further specifications. But there is enough there take a stab at answer the questions I asked earlier.
Note: this is an 802.11 "WiFi" mesh network. That's the same architecture that is being used in metro wireless installations from Philadelphia to San Francisco. For the technically inclined: the hardware standard described involves two radios operating in two different bands. Specifically, the equivalent of Tropos' most advanced access points, and its software, is specified. (Tropos is the market leader in metro WiFi.)
1. Does it include a very strong backbone "supply" element?
- Yes, It is hung directly off LUS's current fiber ring. --It will not be crippled by running off a wireline supply source that has less capacity than it is able to use. (The expense of providing for adequate "backhaul"--and sometimes the ability to find such at any price has been a major limiting factor in most public muni WiFi efforts.)
- Yes, the request makes it clear that there will be at least a "phase 2" (official protestations aside) and that proposal should take into account the networks eventual expansion to full coverage of the 45 square miles of the city. The access point model specified is the first of a new generation from its maker and future models in the family are promised to be interoperable with these and to support emerging technology and standards like MIMO and WiMax and older standards like public safety.
- Hmmn...well maybe or at least implicitly. Nothing beyond the first layer, "phase 1" is specified. But assuming that what is described for phase 1 sets the pattern for the future it looks like the plan is to make full use of the fiber. Wireless mesh networks are built around ratios between aggregation access points that are connected to backhaul networks and simple mesh network which are only connected to other access points via wireless. Common acceptable ratios are 5 or 6 mesh nodes per aggregation point. All too many systems are using larger ratios and putting up with the resulting performance issues. A gold-plated system would use a slightly smaller number. The ratio LUS is suggesting for phase 1 is 1:1.3. That is astonishingly low and only makes sense where the wireless owner also owns the backhaul network (in our case fiber). Other users would have to pay per drop for their microwave, WiMax, T1, fiber link, or the like and such per drop costs would run up the expenses very quickly. Maintaining such a low ratio would mean deploying a system of pretty astonishing capacity. While policy might limit the bandwidth allowed, nothing in the network itself limit network speeds. They could conceivably run at near the rated speed of 802.11 protocols that underpin it--currently about 54 megs.
- Fiber. This is certifiably yummy. See above.
- Yes. This is a "good thing," for it means that a multitude of low-cost hardware will be able to access the network. Proprietary spectrum has some advantages for local governments and, generally, some is available to it for various safety functions but such networks cannot be practically be opened for public use.
- WiFi is specified. The suggested hardware is software upgradeable.
- Support for a wide range of applications including surveillance video, voice, data, mobile communications-seamless roaming, VPN, and meter reading are in the specs.