We've been running point-to-point links for 9 years, upgrading technologies as the market evolves. Currently have 90+ radios servicing about 40 facilities in 23 towns, links range from 500 feet to 15 miles and carry all data and VOIP traffic, so yes, we have some experience, mostly positive. Our most distant location hops through 6 links before it reaches our data center and gets there with 5 - 6 ms delay. Everything currently is unlicensed radio frequencies, no optical links. We only have a couple links where we're trying to push near 100 megabit of actual throughput (averages closer to 70 meg), that for supporting off location backups and redundancy for our main data center (30 mile separation), the balance we try to achieve a 20 - 30 meg TCP throughput which is adequate for our small, remote office needs and very cost effective compared to attaining 100 meg links. We self install and self maintain all of it and wouldn't trade any of them for VPN's (we have several of those also) as we prefer being in control of our network and taking the middle man out whenever possible. We're dealing with very rural locations with limited options for quality, affordable data connectivity, which is what led us to investigate fixed p-t-p in the first place.
What you're looking at is very feasible, provided you have roof access and ability to run cable to the radios. Be careful with vendor and manufacturer claims of throughput, most show raw, theoretical numbers and don't account for the hit you take on overhead and management of the signal (figure 50% minimum), sounds like you will however have benefit of short distance which makes for an easy link. This has been a very fluid technology the past few years and the competition is helping price points, but you should still be prepared to pay some significant dollars to achieve 100 meg+ of true throughput in a radio that is robust, self healing and secure. Also a lot of junk out there so be sure and get customer references to back up vendor claims. Conversely, if you can get by with less throughput, there are a number of fairly economical solutions available. If you're in an area that may already be congested with competing equipment, might be a good idea to hire someone to do a quick site survey with a spectrum analyzer. That's typically less of a problem on very short links and the better quality (read more expensive) equipment can monitor spectrum and self adjust as needed. Dennis -----Original Message----- From: Ben Scott [mailto:[email protected]] Sent: Tuesday, November 10, 2009 9:49 AM To: NT System Admin Issues Subject: Short PtP LoS fixed wireless (was: Connecting two offices...) On Tue, Nov 10, 2009 at 10:03 AM, John Aldrich <[email protected]> wrote: > Well, if nothing else, do you have LOS between the two buildings? If > so, maybe a point-to-point wireless circuit would work. I'm actually somewhat interested in this. Do people here have any experience, positive or negative, with short-haul point-to-point line-of-sight fixed wireless? (Wireless could mean radio or optical.) The common scenario would be: Two buildings across the street from each other. Not feasible to run a hard line. But you can get a clear line-of-sight from the roofs. I wouldn't want to run my LAN over a common 802.11 link -- there are problems with throughput (especially when under heavy load), reliability, and security. (So a Pringles can is not an option, sorry.) But I'm assuming by taking advantage of this specific scenario, something better would be possible. Proprietary gear is acceptable, since if you're just linking two points, you don't have to worry about interoperability. Ideally, I'd like something achieving at least 100 megabit switched LAN speeds, with solid encryption ("VPN class", if you will). -- Ben ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~ ~ Finally, powerful endpoint security that ISN'T a resource hog! ~ ~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/> ~
