It is almost Thanksgiving and time for another entry. November has been enormously busy and fun.
I spent the week of November 6 through 11 with Eric Blossom (K7GNU) and Matt Ettus (N2MJI) of GnuRadio and USRP. Please see
http://www.ettus.com
GnuRadio
We managed to get some exciting work done. We did a complete packet radio system that ran at 768 kbps using GMSK on the computers we had available with no tuning whatsoever. The packets were 1500 byte packets (go ethernet). We managed to get resamplers, pll based fm detectors, lms_dfe equalizers, and myriad other modules working. Matt is about to introduce his FlexRF boards. We used the FlexRF boards for 70 cm and got a functional 70 cm FM transceiver, including squelch and PL tone decoding and generation going. (Remember, this was five days work!). Matt and Eric were extremely patient with me while I learned the Gnu Radio system and I feel together we made some very real contributions.
My best friend and work partner, AB2KT, Frank Brickle, was supposed to go as well representing AMSAT. We all missed him tremendously but his father was having surgery done. We are grateful that has gone well and look forward to our next meeting with Frank in attendance.
I managed to get the code at the heart of the Flex Radio SDR-1000 to run using SSE and the SSE version of FFTW.
Flex Radio
This should provide for increased efficiency and faster code at the heart of the SDR.
This coming weekend, the Saturday after Thanksgiving, AMSAT-NA will hold an RF meeting in the Princeton, NJ area for the Eagle Spacecraft project.
AMSAT Eagle
As AMSAT's new Vice President for Engineering, I have some fun technical duties and some managerial duties but for the most part it is like returning home. It has been a very six weeks since I became the AMSAT VPE, and I expect the next few years to be very interesting indeed.
On Dec. 2-4 I will be operating at the world class 160 meter station in Tuckerton, NJ assembled by the W2GD crew. John Crovelli has gathered a world class crew and for some reason allows me to play with them. It has been a lot of fun! Expect us to be LOUD. I was able to make some changes to the antenna farm and expect to build my receive array ideas there during the next year.
From Dec. 5 - Dec. 16 I will be in Marburg Germany. I am helping to write the software for the flight control computer for AMSAT-DL's Phase 3 Express program. Written in a language designed by the technical lead for the project (IPS by Karl Meinzer, go grab an old Byte magazine and read about it), it will be a challenging task. Since we will be using the same computer for the AMSAT Eagle program, this is not interference but synergy between the two projects.
AMSAT-DL Phase 3 Express
Please visit my photo pages to see pictures related to this blog
N4HY Pictures
More next month!
Monday, November 21, 2005
Sunday, October 16, 2005
Wow, what a month: AMSAT, OFDM, etc.
The month since my last post has seen a whirlwind of activity. I have been working on several things and you can expect to hear more about all of this as time passes.
In August and September, I joined a conversation with John Stephensen, KD6OZH about doing some really nice coding for OFDM based modems for communications systems for amateur radio and even outside of amateur radio. Frank Brickle and I are building our Near Vertical Incidence Skywave antenna systems to use the 7 MHz (40 meter) amateur radio band to some experimentation on this system.
We will be doing our first experiments with the SDR-1000 because of the ease of using it for wideband HF communications. We will be implementing a concatenated code system (two forward error correcting codes back-to-back) in conjunction with a permutation system for the data and then encoding this data for tranmission in a system called Orthogonal Frequency Division Multiplexing (OFDM). The basic reasons for such a system is that in makes use of diversity in time and frequency to spread out the pain of transmitting data over bad paths. It has been demonstrated in many places how effective this type system can be on the horrible paths of HF (shortwave).
In addition to the work briefly described above, I have begun a serious study on the tranmission protocol for such a system. There has been tremendous progress in systems that transmit data and then retransmit data if the receiving misses all or part of the data. We believe we can implement an automatic request for retransmission (ARQ) system that can achieve rates that approach 90% efficiency and waste a minimum amount of time doing the transmission protocol. Lin/Costello's new edition of Error Control Coding has just been released and their is an absolute spectacular discussion in an entire chapter with a huge bibliography on this topic.
At the annual meeting of AMSAT, NA Software Defined Radios were a major topic. Tom Clark and I have been working on a system we jointly proposed with Phil Karn two years ago and we have made steady progress in studying the system (see http://www.gpstime.com for the CC Rider papers). In the two years we have been pursuing this scheme, we have done a fair (C+) job of selling the system. Matt Ettus of GnuRadio and USRP fame
http://www.ettus.com
did a better job in 45 minutes than we have done in two years of selling the entire concept. He proposed a hardware system for doing the communications and really blew the audience out of their socks. With the system proposed by Matt, we would be able to build a fairly complex text messaging system that would need a "paper clip" for an antenna!
I was first alternate to the board of directors for AMSAT this year. It was interesting to do be in on these conversations this year as there is a lot of new blood in the AMSAT board. I was a member of the board of directors a decade ago and more and it brought back pleasant and unpleasant memories of the wrangling that goes on in these meetings. Anyway, the organization asked me to be its Vice President of Engineering. This is one of the corporation official offices (in the by-laws) and will require me to change modes a bit. Thankfully Jan King and Karl Meinzer (google for many pages) will be available for technical consulation. Please visit:
AMSAT
More later . .
In August and September, I joined a conversation with John Stephensen, KD6OZH about doing some really nice coding for OFDM based modems for communications systems for amateur radio and even outside of amateur radio. Frank Brickle and I are building our Near Vertical Incidence Skywave antenna systems to use the 7 MHz (40 meter) amateur radio band to some experimentation on this system.
We will be doing our first experiments with the SDR-1000 because of the ease of using it for wideband HF communications. We will be implementing a concatenated code system (two forward error correcting codes back-to-back) in conjunction with a permutation system for the data and then encoding this data for tranmission in a system called Orthogonal Frequency Division Multiplexing (OFDM). The basic reasons for such a system is that in makes use of diversity in time and frequency to spread out the pain of transmitting data over bad paths. It has been demonstrated in many places how effective this type system can be on the horrible paths of HF (shortwave).
In addition to the work briefly described above, I have begun a serious study on the tranmission protocol for such a system. There has been tremendous progress in systems that transmit data and then retransmit data if the receiving misses all or part of the data. We believe we can implement an automatic request for retransmission (ARQ) system that can achieve rates that approach 90% efficiency and waste a minimum amount of time doing the transmission protocol. Lin/Costello's new edition of Error Control Coding has just been released and their is an absolute spectacular discussion in an entire chapter with a huge bibliography on this topic.
At the annual meeting of AMSAT, NA Software Defined Radios were a major topic. Tom Clark and I have been working on a system we jointly proposed with Phil Karn two years ago and we have made steady progress in studying the system (see http://www.gpstime.com for the CC Rider papers). In the two years we have been pursuing this scheme, we have done a fair (C+) job of selling the system. Matt Ettus of GnuRadio and USRP fame
http://www.ettus.com
did a better job in 45 minutes than we have done in two years of selling the entire concept. He proposed a hardware system for doing the communications and really blew the audience out of their socks. With the system proposed by Matt, we would be able to build a fairly complex text messaging system that would need a "paper clip" for an antenna!
I was first alternate to the board of directors for AMSAT this year. It was interesting to do be in on these conversations this year as there is a lot of new blood in the AMSAT board. I was a member of the board of directors a decade ago and more and it brought back pleasant and unpleasant memories of the wrangling that goes on in these meetings. Anyway, the organization asked me to be its Vice President of Engineering. This is one of the corporation official offices (in the by-laws) and will require me to change modes a bit. Thankfully Jan King and Karl Meinzer (google for many pages) will be available for technical consulation. Please visit:
AMSAT
More later . .
Wednesday, September 14, 2005
Been a very busy month
I have IPS (see IPS ) running on the AMSAT IHU-3. The IHU is the internal housekeeping unit of the spacecraft. It receives and processes commands, it maintains the state of the spacecraft and attempts to mitigate faults and generally manage the health of the spacecraft. I have pictures of our meetings in Germany and Washington state in these gallery entries:
Germany: Germany IHU-3 pix
and
Washington: Washington at KK7P IHU-3 pix
We can talk to the bus of the spacecraft and programming of the IHU flight code has commenced.
The American Radio Relay League sponsors technical working groups. I am chairperson for the Software Defined Radio working group and a member of the High-Speed Digital Networks and Multimedia Working Group. We are working on a class of modems ideally suited for both HF and VHF and Microwave communications. The modem class is called OFDM and this is orthogonal frequency division multiplexing. This is a fancy name for sending a bunch of data down parallel channels (parallel in frequency) and done in a way that the frequencies used provide an very nice mathematical property which aids considerably in the modulation and demodulation of these signals.
AMSAT will be holding a board of directors meeting and AMSAT Eagle spacecraft design meeting in Pittsburgh Oct. 6-9. I will be demonstrating with my partners a working software defined transponder SDX. You can hear the first QSO by visiting
GPS Time page with SDX pix and audio
Please visit the last bulleting and you can download and listen to the audio.
The ARRL review of the work Frank Brickle (AB2KT) and I have done, in cahoots with the folks at Flex Radio is available here:
ARRL Review of SDR-1000
By the time of my next entry, I will know if I am on the board of directors of AMSAT. I was talked into it, I have done it before and chose not to run back when AEA failed and amateur radio lost its luster. I am glad it has regained it.
Germany: Germany IHU-3 pix
and
Washington: Washington at KK7P IHU-3 pix
We can talk to the bus of the spacecraft and programming of the IHU flight code has commenced.
The American Radio Relay League sponsors technical working groups. I am chairperson for the Software Defined Radio working group and a member of the High-Speed Digital Networks and Multimedia Working Group. We are working on a class of modems ideally suited for both HF and VHF and Microwave communications. The modem class is called OFDM and this is orthogonal frequency division multiplexing. This is a fancy name for sending a bunch of data down parallel channels (parallel in frequency) and done in a way that the frequencies used provide an very nice mathematical property which aids considerably in the modulation and demodulation of these signals.
AMSAT will be holding a board of directors meeting and AMSAT Eagle spacecraft design meeting in Pittsburgh Oct. 6-9. I will be demonstrating with my partners a working software defined transponder SDX. You can hear the first QSO by visiting
GPS Time page with SDX pix and audio
Please visit the last bulleting and you can download and listen to the audio.
The ARRL review of the work Frank Brickle (AB2KT) and I have done, in cahoots with the folks at Flex Radio is available here:
ARRL Review of SDR-1000
By the time of my next entry, I will know if I am on the board of directors of AMSAT. I was talked into it, I have done it before and chose not to run back when AEA failed and amateur radio lost its luster. I am glad it has regained it.
Tuesday, August 02, 2005
August 2, 2005
I have managed to get the IPS-E0 image (the basic IPS operating system) into the bootloader for the IHU-3 (integrated housekeeping unit for Phase 3E and Eagle). It will be easily loaded via the bootloader by just striking the E key. This enables us to upload, using the bootloader, any type of IO, interrupt handler, etc. that we wish to upload. This should greatly speed the development process for the flight code for Phase 3E.
I have added some CAN controller initialization code to the bootstrap IPS routine as well as an asynchronous communications mechanism. It will use the flag facility already built into IPS to send and receive 512 byte blocks but it will be almost completely limited to printable ascii characters.
On the SDR-1000 front, I modified some of the settings for the "microphone leveler". It appears that we have too much early attack which appears to limit the average power too much. It may be that we never wish to allow for negative gain at this stage.
It is clear that serious investigation of this process is required. I only wish I could force myself to be interested in doing it. Sigh.
The SDR-1000 official numbers are in. 99 dB IMD-DR and +27 dBm IP3 on the low bands at 2 Khz with the preamp in medium setting (giving <-120 dBm MDS). OUTSTANDING.
Bob
I have added some CAN controller initialization code to the bootstrap IPS routine as well as an asynchronous communications mechanism. It will use the flag facility already built into IPS to send and receive 512 byte blocks but it will be almost completely limited to printable ascii characters.
On the SDR-1000 front, I modified some of the settings for the "microphone leveler". It appears that we have too much early attack which appears to limit the average power too much. It may be that we never wish to allow for negative gain at this stage.
It is clear that serious investigation of this process is required. I only wish I could force myself to be interested in doing it. Sigh.
The SDR-1000 official numbers are in. 99 dB IMD-DR and +27 dBm IP3 on the low bands at 2 Khz with the preamp in medium setting (giving <-120 dBm MDS). OUTSTANDING.
Bob
Subscribe to:
Posts (Atom)