Digital Modes of Wireless Communications

Spread Spectrum Ham Radio

 

 

 

NODEDB 

Worldwide APs
Kitchener APs

 

ARRL 

HSMM
ARRL 802.11
more ARRL 802.11

 

Antennae 

Helical 2.4g Antenna
NEC2 Antenna Profiler Software
Superpass 2.4g Antenna Manufacturer
Bidirectional Amplifier

 

Ham Projects 

IEEE 802.11 & Ham Radio

 

WLAN In The News

802.11 Planet
Wireless News
Electronically Steered, Planar-Phased Antenna Array
Non-ham Antenna Experiments
Community Centre Project UK (021105 Slashdot)
IEEE 802.11 & Bluetooth
802.11b Community Network
802.11 Networks
Free Networks Org
Waterloo Wireless
Wireless News (Globe)
Toshiba Wireless Service (Globe)
Reflector Mountain (PBS)

The best point of reference when describing radio based digital networks, is the residential Internet connection. A digital subscriber line (DSL) tends to be asynchronous, offering download speeds of about 1 mbps, but upload speeds might not be much better than 128 kilobits (kbps). The best DSL speeds I've seen are in the 3mbps download, with about 380kbps upload. Cable Internet speeds are roughly analogous.

Wireless internet connection speeds depend on the quality of connection, which in turn is influenced by line of sight propagation, by attenuating factors such as atmospheric humidity, trees and other intervening objects. A good wireless connection, point to point, can yields speeds that are several times that of a DSL download, and more than ten times the that of a DSL upload. IEEE 802.11b is nominally 11mbps, and people have reported getting between about 3 and 5mbps in either direction.

I am interested in wireless, because I am a Ham, and I am licenced to make the airwaves my experimental laboratory. This also means that I am interested in harnessing technology for the good of humanity. Technology is good as long as the good are free to use it in a beneficial way. Since Hams are relied on for emergency communications, the wireless networks of interest are not necessarily gatewayed into the wired system. Certainly, it would be our objective not to rely on such interconnections from such a perspective. Some of the things I talk about and link to here, are not restricted to licenced Hams, while other things are. I don't always identify the things that require a licence. Even if you are a casual wireless experimenter, I would encourage you to become licenced and enjoy even greater privileges associated with a great hobby.

The first thing to note, is that the strongest commercial access points are about 100 milliwatts in transmit strength. This doesn't sound like much but with a quality high gain directional antenna, this can get you about thirty kilometers node to node. Do get the antenna fairly high, above the trees for sure. Also, use high quality LMR-400 feed line, or better -- with only 100mw to play with, we can't afford to be frivolous. Most people will not need more than this, however. If in doubt, use the "link analysis" calculator (margin) to calculate your specifics. Even though Hams are permitted more than one watt, there are special rules which apply, and the circumstances that require this kind of power are extremely rare I am told.

The next thing to note, is that most access points will produce a far weaker signal than say a Cisco, or a SMC. An Apple Airport AP, with a built in antenna will have an omni directional range of about 50meters. Normally, these are used inside a building where the range is further attenuated by the walls, floors, ceilings, etc.. These are great for walk-about coverage for your laptop computer, but are less useful in building a wireless infrastructure. They are a great easy to use product, nevertheless, for infrastructure I would probably go with a very low cost Linux AP, an SMC, or if money is secondary, a Cisco.

Next question is polarization. Most non-hams are using vertical omni-directional antennae. It has however been shown that circular polarization is better on average, due to the effects of multipath and faraday rotation. Solid dishes and reflectors tend to do a better job than equivalent grids.

Probably, the long range goal for this wireless network would be to run IPv6. The fact that IPv4 is reasonably well understood at present should lull us into using it for anything more than an initial stepping stone. IPv6 solves many problems which hamper v4. Use a standard text on v6 if you don't understand what I mean.

Hams need to identify every few minutes, mandated by law. Many suggestions have been made as to the use of IPv6 facilitating this, but until IPv6 gets serious mindshare, the best answer seems to be ping payload in IPv4. To that end here is a script that a Ham circulated on the tamu mailinglist:

#!/bin/sh
# pingcall by n9ssa@arrl.net 3.3.3 (3.march.2003)
# usage: pingcall 
#
# Shell script that
#  finds default gateway,
#  converts supplied callsign to hex,
#  and pings the default gateway once with callsign in payload
#
echo Pinging default gateway with callsign in payload
echo Ascii payload is $1
gw=`route -n | grep ^0.0.0.0 | awk '{ print $2 }'`
echo Default Gateway is $gw
hex=`printf %s "$1" | od -An -txC | sed 's/ //g'`
echo hex is 20$hex
ping -p 20$hex -c 1 $gw
	
Use this on Linux and similar machines, invoked by ./pingID.sh MYCALL. It assumes gw is some other machine across the wireless network, I think. A followup suggested:
	ping -b -p 20$hex -c 1 255.255.255.255
	
You probably want to put this into a loop and have it cycle every 10 minutes. Alternatively, those who like to cron can add it to their crontable entries pretty much as is.

What sort of throughput can you expect with 802.11g? According to most reviews, with pristine signal quality you can get about 14mbps. At the extreme fringes of signal, you can expect about 1mbps.


SUMMARY AP SPECS

	
Date:    Sat, 5 Apr 2003 08:53:33 -0800
From:    Ron Curry rec at CURRY dot ORG
Subject: AP/Bridge TX/RX specs

I though I'd repost this since there have been a couple of questions
recently on this topic.

Ron Curry
N6QL


Hi Guys,

Here is my compilation of TX/RX data for some popular hardware:

I collected these over a period of about the last two years and just
updated them recently before publishing here:


- Dlink DWL-2000AP 15 dbm TX, -89 dbm RX (1mb), -77 dbm RX (11mb)
- Dlink DWL-900AP+ 15 dbm TX, -89 dbm RX (1mb), -77 dbm RX (11mb)
- Dlink DWL-810+ 15 dbm TX, -89 dbm RX (1mb), -77 dbm RX (11mb)

- Linksys WAP11 18 dbm TX, -84 dbm RX (1mb), -75 dbm RX (11mb)
- Linksys WAP54g 15 dbm TX, -89 dbm RX (1mb), -80 dbm RX (11mb), 
	-65 dbm (54mb)

- Cisco BR-342/350 20 dbm TX, -94 dbm RX (1mb), -85 dbm RX (11mb)
- Cisco AP-350 20 dbm TX, -94 dbm RX (1mb), -85 dbm RX (11mb)

- Smartbridge SB2110 17 dbm TX, -92 dbm RX (1mb), -83 dbm RX (11mb)
(Note: integrated N connector!)
- Smartbridge SB2510 17 dbm TX, -92 dbm RX (1mb), -83 dbm RX (11mb)
(Note: integrated N connector!)

- Enterasys (all) 15 dbm TX, -94 dbm RX (1mb), -82 dbm RX (11mb)

- Proxim (Orinoco all) 15 dbm TX, -94 dbm RX (1mb), -82 dbm RX (11mb)

- Intel 2011 AP 20 dbm TX, -90 dbm RX (1mb), -83 dbm RX (11mb)

- SMC 2482W 18 dbm TX, -91 dbm RX (1mb), -82 dbm RX (11mb)
- SMC 2655W 20 dbm TX, -91 dbm RX (1mb), -82 dbm RX (11mb)
- SMC 2682W 13 dbm TX, -81 dbm RX (1mb), -76 dbm RX (11mb)

All figures came from the vendors website or directly from the
vendor via phone call. Please note that the DWL-900AP+ and the
WAP11 v2.2 are supposedly same hardware. I don't know why they
have slightly different specs though. I have verified a few of TX
numbers with my
spectrum analyzer (Cisco, Linksys, Dlink). Have no way to measure
the RX numbers but actual experience reflects the numbers on the
ones I experimented with (again Cisco, Linksys, Dlink).

Stay away from any brand that will not release RX figures or only
publishes "range"! Netgear and 3com are two examples - they won't
release their specs which can only mean they must be terrible! I really
had
to dig to find the Enterasys and Dlink numbers.

You can see from this there is a broad range of capability here. In
general, you get what you pay for. Linksys and Dlink have the worst
specs of all - hence their low price. The Linksys receive specs are
horrible and these are penny-wise pound foolish units because you'll
need to add a $150-$300 500mw amp+pigtails+powersupply to your initial
$120 purchase just to get them to approach the Cisco or Smartbridge in
range.

Please note that 20 dbm = 100mw, 18 dbm = 63mw 17 dbm = 50mw, 15 dbm
= 32mw. You get what you pay for!

One last note, an AP can ONLY talk to clients. An AP cannot
communicate with other AP's unless it has a bridge mode. A bridge with
MAC/IP filtering is
highly desired for point-to-point links. Linksys, Dlink sell AP's that
have bridge mode (but then again their specs are horrible).
Enterasys has a bridging mode as I recall also. Cisco, Proxim,
Smartbridge have separate bridge products that can also be used in
AP mode. Bridge targeted products usually have more security and
network management capability than the low-end stuff. Also, and this
is very important, I've not found any interoperability between
vendors in a bridging configuration. In otherwords a Linksys will
not talk to a Dlink or a Cisco, etc. Based on vendors data sheets
and those that I have test they are all proprietary therefore you
need to stay with one vendor for a backbone or point-to-point links
or build "repeaters" using two different units back-to-back (read
expensive).

Hope this helps some of you out.

73, Ron Curry N6QL


	

 

AP Technology 

Linux Access Point
Link Analysis
Hardware Comparison
Airport Mods

 

Admin Utilities 

Kismet Wireless Sniffer
Netstumbler
802.11 Frame Tutorial

 

Quick Links 

IEEE 802.11b Patents
Software Radio
Digital Modes
Amateur Radio Reference Guide
My Projects Page

 

Clubs 

AMRAD


va3xto@rac.ca
Copyright © E. J. Ritzmann.
Last Change $Date: 2003/12/24 21:36:17 $

Viewable With Any Browser