Whenever a TMC Labs engineer takes a vendor meeting, attends a trade show, corresponds
with a reader, or talks with an analyst, the engineer always asks if there are new things
we should be covering. Our queries, we’ve found, have elicited many, many comments
about wireless technologies.
So, we’ve embarked on a product-ordering spree, inviting vendors to send us their
latest wireless networking products. The first vendor to respond, SilCom Manufacturing,
sent us Freespace Turbo. The version we got uses a line-of-sightlaser beam to transmit
data up to 1,000 feet, at 155 Mbps ATM, 100 Mbps Ethernet (full- and half-duplex), 45 Mbps
T3, and 100 Mbps FDDI.
Starting at about $17,000, FreeSpace Turbo is an effective (but pricey) solution for
campuses where buried utility lines, natural obstacles, or severe weather might complicate
the deployment of wireline solutions. FreeSpace Turbo includes features such as built-in
telescopic sights, optional lens defrosters, fiber-optic connections, and wall/pedestal
mounts.
INSTALLATION
The whole of the laboratory staff had zero experience with fiber optics, but our
unyielding optimism made us attempt a solo installation anyway. To our delight, the
installation turned out to be easier than we had expected.
The main components for each end are the transmitter (“link head”), control
box, and mounting hardware. We started by assembling the mounting hardware and then
attaching the link heads to them. This was a simple process, accomplished with regular
hand tools.
We should point out that we tested the product indoors over a mere 30-foot distance, so
we dispensed with the mounting hardware. Instead, we carefully rested the transmitters on
the laboratory floor. (Of course, in a real-world environment, the mounting brackets are a
requirement, not an option.)
We sent an e-mail message to the whole office, cautioning the staff to avoid walking
through the laser beam underfoot. We were sure to repeat the warnings in the setup manual
about how blindness could result from staring at the laser for too long. These warnings,
as you might guess, occasioned no little concern.
The next step was attaching a PC to each hub and attaching each hub to a transmitter.
The physical connections between the computers and the hubs were standard network wires,
but the connections between the hubs and the transmitters were fiber-optic wires. We found
that they connected easily on the hub end, but they required a lot of patience and a close
reading of the instructions for the transmitter end. After some experimenting, we got it
right. We accessed the hubs using Windows’ HyperTerminal and an RS-232 connection,
and found the command line interface to be very easy to use. We configured each hub with
an IP address and a subnet mask matching those of the computers.
The final step was to test each transmitter’s built-in UPS and to properly align
the laser. To do so, we had to unlock the control box using the key that came with the
equipment. Testing the UPS involved pressing a few switches and turning a couple of keys
— a procedure eerily reminiscent of the famous “dual key” configuration
used to launch nuclear missiles.
While testing the UPS was quick and easy, using the telescopic sights to align the
laser was very tedious. Each transmitter unit, much like a microscope, has a coarse
adjustment and a fine adjustment. The goal is to have the crosshairs of each telescope
pointing directly at the crosshairs of the other. Fortunately, the units have a feature
that compensates for natural building movements. In the real world, organizations would
hire professional surveyors to assist with this process, but to expedite our test we
cheated a bit and used a level and a few shims.
Among the things we learned was that our office floor is not level. We also learned, in
a conversation with a SilCom representative, that the laser warning notes are there mostly
for liability reasons. The laser’s power is no more than that of the laser in a CD
player. Still, because we were using the transmitters in such close proximity, we were
advised to attenuate the beam by covering each laser apparatus with a piece of clear
plastic or cellophane.
The system also includes a special telephone handset for each transmitter. It’s
not a regular telephone, but when you’re not sending any data across the network, you
can test the connection by having a conversation across the wireless network. The quality
of this connection was like that of a low-end voice over IP circuit, but it works well
enough for installers to communicate with each other while standing on twelfth-floor
rooftops 300 meters apart.
DOCUMENTATION
We had one manual for the hub (Fast Ethernet switch) and one for the transmitter. We also
found some good information at the SilCom Web site, including data sheets, suggested
applications, etc.
Overall, the manuals were good. They were of the no-frills variety, but they included
all of the necessary explanations and illustrations at a level designed for average
network personnel. We’d like to see improvements to the sections regarding the
cable-to-transmitter setup and the transmitter height/angle adjustments. These sections
could have been clearer.
FEATURES
The designers of the Freespace Turbo thought of some features that are particularly
appropriate for an outdoor wireless network. For example, since the network path is the
air, there is built-in encryption and security so that no one can intercept your data. The
transmitter units are sealed from the weather elements, and there is an optional lens
defroster that automatically enables itself when the outside temperature falls below 37
degrees Fahrenheit.
Beside the telescopic sights and telephone link discussed above, there is a
tone-alignment signal and there is a signal strength meter. There are built-in diagnostic,
loopback, and fallback controls, and there is a built-in three-hour UPS system. The data
path is guaranteed for 99.9 percent statistical uptime, and since the unit makes no use of
any time or date information, users needn’t worry about Year 2000 issues.
We noticed another interesting feature when a TMC employee walked through our laser
beam path. (This fellow hadn’t read our cautionary e-mail message.) Essentially, the
disruption brought down the network connection — but very briefly, mind you. When the
connection was reestablished, no reboot was necessary to reinitiate communication. Thus,
the reconnection after the walkthrough was no more disruptive than it would have been had
we reconnected a regular network wire to a hub. We did hear a few quiet clicking noises,
but the reconnection was seamless.
The hub/switch units that SilCom shipped with the Freespace Turbo were their p8400
models. The chassis units fit two cards, which are available in several configurations.
The cards we received included a 16-port 10/100Base-T with SNMP management, and an 8-port
combination card that was one-half 10/100 TX and one-half 100Base-FX (fiber optic). The
cards cost between $925 and $3,095 each.
Users are not required to use the SilCom models; any standard hub with a 100Base-FX
port will work. The SilCom hubs do have some nice features though, like serial connections
for SNMP and Telnet management.
OPERATIONAL TESTING
Our testing spanned a two-week period. During this time, we noticed that the network speed
was greatly affected by the speed of the endpoints. Our configuration, we should point
out, was a lowest-common-denominator setup. We had one Windows computer attached to each
hub, with the wireless link in the middle. This arrangement eliminated any outside factor
that could have affected the data speed.
Our experience was that the data connection was fast, impressively so, considering the
connections was wireless. However, we would recommend this kind of solution on the basis
of its functions, not just its performance.
As for functionality, we tested the control box’s built-in UPS, the diagnostic/
loopback features, and the remote management. We’re pleased to report that for our
first-ever wireless network test, everything worked the first time, and everything from
the mounting pedestals to the beam alignment to the network configuration was extremely
easy to set up. We can’t see any reason why the installation and performance would
not be just as good across a large campus as it was in our laboratory.
One last observation: The wireless connection is invisible to the network, which is
good because it means that you don’t have to spend much a lot of time catering the
network settings to the transport medium. (So many people want network intelligence, but
sometimes dumb computers are better!)
ROOM FOR IMPROVEMENT
The knobs for adjusting the transmitter’s height angle and side-to-side
angle were — not to mince words — a pain to use. Also, they were poorly
documented. Attaching the fiber connections to the hub was as easy as plugging in a
standard RJ-45 connector, but attaching them to the link head was difficult. Even worse,
this part of the installation was also poorly documented.
So much for the petty frustrations. To make a more positive suggestion, we recommend
adding zoom controls to the telescopic sight.
CONCLUSION
We strongly recommend this product. It works well, it is easy to install (our wireless
neophytes installed the product in less than an hour), it has an impressive feature set,
and it is (for the most part) well documented. The optional hubs/switches are of very high
quality as well.
The main drawback is the price. By the time you invest in hubs with fiber ports and
hire professional surveyors to conduct the link head alignment, the product will become
very expensive (you can do the alignment yourself, but for serious business applications,
you’ll want to account for natural building movements, birds, weather, etc.). If you
have a multi-building campus and need outdoor, high-speed network connections, then this
might be the right product.
|
TMCnet LOGIN
Webinars
