Topic: Radio
Guest speaker: Prof. Jim Binkley, PSU
Attendees: ?JameySharp, Andrew, Glen, ?JamesPerkins, Richard, ?LarryLeach, and a couple more whose names I did not get.
Notes:
- James introduced Richard, BS Physics guy
- Andrew gave his "two minute introduction" which was more like 30 minutes... but it covered a lot of ground, especially regarding radio gear and antennas.
- Goal of LV2 radio is to go faster than 19.2kbps (spec: 100kbps), and bidirectional (uplink & downlink); uplink can be slower.
- Glen is the radio guy
Richard, it turns out, has some familiarity with antennas, too.
Previous design had a 900 Mhz and a 144 Mhz link.
- Telemetry data was modulated and put in the Television audio channel
- Donut baluns?
- Turnstile reflector on ground.
- Vertical polarization prelaunch, circular polarization overhead.
- LV2 will go Mach 3, 14g maximum accel, -10 maximum decel.
LV2 will use AMD Elan SC520 (586), has 128M SDRAM, CAN bus
Radio card may be a PC*MIP 802.11 card (one is due out in a month or so), or maybe on ISA-like Internal General Purpose Bus.
- Jim says the Lucent (Orinoco) cards are the most familiar to him, and sport a plug-in antenna override.
- 802.11 sounds cheap. Note it is near the sub-2.4 Ghz ham band... perhaps a modest freq. shift and we could amplify the signal for the onboard radio.
Jim says if we use std. cards either we need a better antenna, or more power. More power necessitates moving to the ham band.
Jim describes 802.11 in some detail, including:
- IBSS mode, where all traffic goes through an Access Point
- ad hoc mode, where any station can talk to any other
- maximum rate is 6 Mbps UDP, 5 Mbps UDP either way
- there are beacons being sent at a rate of 10 Hz -- this is probably adjustable -- the beacons are how cards find each other.
- Some access points handle roaming, but it's different from brand to brand.
- RTS/CTS mode adds more packet overhead (3 more packets per data packet)
- Access Points have a power saving feature where the RX shuts down periodically -- seems to use scheduling of next transmission. This seems to be a waste of bandwidth.
- some of the early ad-hoc designs simple stick an ethernet chip on the back end.
- Access points have a 2-3 Mbps slideoff
- at its best, 802.11 will gert ~5Mbps. It is CSMA/CA -- no way to do CD on radio!
- tool for throughput testing: ttcp (test tcp) -- test bandiwdth.
- if two systems are talking, bandwidth is haved to 2.5Mbps
- BAWUG is the Bay Area Wireless User's Group -- folks trying to do massive home-to-home networking to circumvent the high cost of T1s. These people may have antenna geeks that are a resource for us.
- Lucent Access Point can add a "network id" code to each packet, to ignore packets on "other" networks. There's also an "any" code, which has been exploited by people driving through corporate parks and surfing on the open networks.
- Cost of PC cards are around $125. Other formats are ISA and PCI cards.
- On ISA you can set up a PC to run as an IBSS/Adhoc - instant router
- Lucent and Aironet look the most interesting. Note a big reason for this is open source support.
- Could use a Yagi antenna. Jim has gotten data without loss from 271/26 interchange to North Plains (~10 miles) using a Yagi on the base station, and the laptop in a car. Summer weather.
- 915 Mhz has been bounced off of billboards on top of a hill to get around line-of-sight restrictions.
- It's hopeful to use an approach to track the rocket, note this requires using an omni antenna onboard.
A ground antenna that tracks the rocket?
- Place antenna below rocket, so that steering at launch is minimal
- Place antenna a couple miles off, so that steering at launch is slow
Discussion of planar vs. cylindrical/helical patch antenna ensues. I am not worthy to understand.
- Could use GPS telemetry from rocket to point ground antenna
- commercial offering promises 10-20 miles using a dish on each end.
- Larry wrote the ARRL regarding a power variance.
- Jim says 1 Mbit is a resonable expectation for the lowest speed we'll get if the radios can talk.
- "part 15" connectors on boards are not possible to acquire.
- The 915 ham band is right next to the cell band, so it's easy to shift freq to the band and amplify like crazy.
- Jim has two 2.4 Ghz dishes, and will let us borrow them. He wants to see how they test out at a distance. We need to acquire pigtails for them, however.
- The power budget may be a harder problem than the antenna-pointing problem.
- Do the cards have an RSSI?
- could use triangulation for recovery
- Jim will come talk to us again about drivers
- James will do a presentation on TWiki (maybe next meeting?) -- need a phone line to connect to the web.
- next meeting all team meeting July 10th 7pm PCAT 103
- Need to talk to BAWUG
- Roster should go up on TWiki
- Can runs at ~500kbps (need at least 200kbps)
- Decision to use Eagle cadsoft.de -- free for educational use
-- ?JamesPerkins - 28 Jun 2001