You are not logged in.
two warp v3 board. warplab 7.4 installed under matlab 2013a, 7.6 installed under matlab 2017a. 7.4 works well. As I ran example with 7.6, the problem appears:
>> wl_example_mimo_ofdm_txrx
Error in node 2 with ID = 1:
ME =
MException with properties:
identifier: 'wl_transport_eth_java:send:noReply'
message: 'maximum number of retransmissions met without reply from node'
cause: {0×1 cell}
stack: [7×1 struct]
Error message follows:
maximum number of retransmissions met without reply from node
Error in node 1 with ID = 0:
ME =
MException with properties:
identifier: ''
message: 'Node 0 reports WARPLab version 7.4.0 while this PC is configured with 7.6.0'
cause: {0×1 cell}
stack: [5×1 struct]
Error message follows:
Node 0 reports WARPLab version 7.4.0 while this PC is configured with 7.6.0
Error using wl_initNodes (line 162)
The following nodes with IDs = [ 1 0 ] are not responding. Please ensure that the nodes have been configured with the WARPLab
bitstream.
Error in wl_example_mimo_ofdm_txrx (line 45)
nodes = wl_initNodes(NUMNODES);
I think the reason is in the warp board:
>> clear
N = 2;
nodes = wl_initNodes(N);
wl_nodeCmd(nodes,'identify');
disp(nodes)
-------------------------------------------------------------------------------
WLVER are both 7.4.0.
Should I follow: http://warpproject.org/trac/wiki/howto/SD_Config? In that way, can this problem be solved?
Or anything I can do to update boards to 7.6.0?
Appreciate!
Offline
By the way, https://warpproject.org/trac/wiki/howto/SD_Config
Are first two steps in the link necessary?
Offline
That error message means that you need to program the boards with the 7.6.0 bitstream in order to use the 7.6.0 MATLAB code. The bitstreams are provided in the release zip. You can program the ".bit" file directly with iMPACT or you can program the ".bin" file onto an SD card via the link you provided in your reply.
myxyuan0617 wrote:
By the way, https://warpproject.org/trac/wiki/howto/SD_Config
Are first two steps in the link necessary?
What two steps are you referring to? You don't need anything in the "Generating the .bit file" section -- the .bit files are already generated and in the released zip file. The steps you need are in the Copying .bin Files to SD Cards section. Note: you need to be careful using a tool like dd and follow the instructions exactly. dd is very powerful and can write arbitrary bytes anywhere in your file system, including critical areas of your computer's hard drive.
Offline
Thank you so much!!! Problem solved.
Offline
Also, *please* use the latest version of WARPLab (currently 7.7.1), especially if you're upgrading from an old release anyway. Every release has bug fixes and/or new features. There have been many improvements since version 7.6.
Offline
If I want to extend "wl_example_mimo_ofdm_txrx"(https://warpproject.org/trac/wiki/WARPL … /MIMO_OFDM) to 4*4 model, is there any guidance on setting the legacy preamble? The 3rd and 4th antenna also transmit during the first burst of Short Training Symbols (STS), symbols with cyclically shifted?
Offline
In general you need the early part of the preamble to accurately represent the power of the rest of the waveform for all receivers. The cyclic shift avoids any beamforming patterns that would occur when transmitting the same waveform from multiple antennas. For the channel training symbols you need some way to orthogonalize the stream from each antenna so the receiver can calculate individual per-antenna-pair channel estimates.
Our WARPLab MIMO OFDM example uses a circular shift on the STF for AGC and orthogonal-in-time (i.e. zeros on one antenna, signal on the other) for the channel training symbols. This is a simple scheme but has the drawback of large power changes from each antenna during the preamble.
The preamble design for the 802.11n mixed format waveform would be a good starting point too. For multi-antenna Tx the HTMF waveform uses cyclically shifted L-STF and L-LTF, followed by orthogonalized HT-LTF symbols per spatial stream. The 11n spec orthogonalizes the HT-LTF symbols in space and time (similar to Alamouti coding, except for the receiver the Tx data is known so it can "decode" for the channel estimates).
Offline
Thank you so much, Murphpo! Right now I can run the 4*4 MIMO program in sim mode, however,once run in warp mode, two warp v3 board each equipped with four antennas, the problem shows out:
Error using wl_baseband_buffers>writeIQ (line 1490)
write_iq: Length of buffer selection vector smaller than the number of columns in the sample matrix. If trying to
write to multiple interfaces, use vector notation vs bitwise addition: [RFA, RFB] vs RFA + RFB
Error in wl_baseband_buffers/procCmd (line 664)
writeIQ(obj, node, transport, buffSel, cmdStr, varargin{:});
Error in wl_baseband_buffers/subsref (line 142)
varargout{:} = builtin('subsref', obj, S);
Error in wl_node/wl_basebandCmd (line 298)
out(n) = currNode.baseband.procCmd(n, currNode, varargin{:});
Error in mimo (line 346)
wl_basebandCmd(node_tx, TX_RF_VEC, 'write_IQ', tx_mat_air);
what's the reason?
Offline
I know the reason, it seems even when it has four antennas on the board, only two antennas works(I just left two antennas on each board, the error information is the same). So I can't just put the new antennas on the board, is there anything else I need to do?
Offline
Did you use the 4-antenna WARPLab bitstream with your 4-antenna (WARP v3 + FMC-RF-2X245) nodes?
Offline
Yes, SD card is written with w3_WARPLab_7.7.1_4RF.bin
Offline
Problem solved!!!!! I ignore to change the number of operating antennas line in 2*2 MIMO code:
TX_RF_VEC = ifc_ids_TX.RF_ON_BOARD_VEC;(same at receiver) only talks about two interface!
Offline
Since we can achieve 4*4 MIMO, is there any chance to know where can we buy the antennas in the setup figure of link http://warpproject.org/trac/wiki/WARPLa … /8x2Array?(the antennas can be arranged linearly). Appreciate!
Offline
You can use any antenna with an SMA plug. The WARP v3 and FMC-RF-2X245 boards use standard SMA jacks (standard (not reverse) polarity, standard (not reverse) gender). The antennas in that demo are the L-COM RE05U-SM. Always double check that you're using standard SMA connections - many "Wi-Fi antennas" have reverse polarity SMA plugs which will not work with the WARP hardware SMA jacks.
Offline
I have another question, how can we change the code " wl_example_mimo_ofdm_txrx.m" in continuous transmission mode?
Offline
How can I keep the transmitting speed in a relatively high level, because I would like to observe the received signal power by a signal analyzer. However, nothing shows out on the signal analyzer, I guess I need a continuous transmission in a high speed.
Offline
To use the OFDM example in continuous mode you will need to set the Tx length and enable the WARPLab node's Continuous Tx mode. Refer to wl_example_continuous_tx.m for an example of using the continuous mode in WARPLab.
Offline
I can run wl_example_continuous_tx.m, however, I can't figure out the reason there's nothing shows on the signal analyzer. I would like to observe the power harvested at the signal analyzer. What's the reason for that, is that related to the transmission speed?
Offline
I would suggest verifying your analyzer's frequency, span and reference amplitude are configured correctly. You should also verify your WARPLab script is defining a sufficiently-long waveform, the 'tx_length' is set correctly and the node is actually in Tx mode. You can use a second WARPLab node to capture the transmitter's output to verify its output (as in the continuous_tx example script).
Offline
I am able to find the power spectrum on the signal analyzer running with wl_example_continuous_tx.m, however, even when I terminate the matlab, it still exists on the signal analyzer. The only way to make it disappear is to turn off the warp board. What's that for? Why the matlab is not taking control of the warp board when the warp board starts to operate?
Offline
When in continuous Tx mode the WARP node will continue transmitting until you send the 'txrx_dis' and 'tx_rx_buff_dis' commands. If you exit MATLAB without sending these commands the WARP node will continue transmitting. MATLAB only sends Ethernet packets when your script sends commands - the node has no way to know that MATLAB has exited.
Offline
I do really appreciate!
Offline
By the way, when I set up interface and select 5GHz channel 1, the peak of the power spectrum is around 5.3GHz, why it's not exactly at 5GHz, is there any explaination for this?
Offline
Refer to the WARPLab docs for the mapping of channel indexes to center frequencies.
Offline
What's the transmission rate, can I adjust it in the program?
Offline