Matrix Channel Measurement System (MCMS)



Overview: The Matrix Channel Measurement System (MCMS) is a joint project of Virginia Tech and Aeroflex Corp., and is supported through an NSF Major Research Instrumentation (MRI) grant.

Maintainer: Steve Ellingson (Virginia Tech)
This page is http://www.ece.vt.edu/swe/mcms/
Last Update: Feb 08, 2008

Click here for some example MIMO channels measured!
Click here to see the MCR in action at the 2005 MPRG Symposium!
Click here to see MCT arb testing!
Click here for HDTV capture!
Click here for Wi-Fi capture!
MCT, MCR, 4-Channel Transmitter 16-Channel Receiver 16-Channel Receiver
and battery (MCT) (MCR), Front (MCR), Back


Project Documents

In reverse chronological order.


Some Basic MIMO Channel Measurements using MCMS

A (Synthetic) 4x4 Full-Rank Channel
Center Frequency: 2450 MHz

Shown are the eigenvalues of the normalized matrix channel H (right multiplied by it's conjugate transpose) when each of the four channels of the MCT is connected one-to-one (using cables) to MCR channels 1 through 4. This creates the ideal MIMO channel: 4 independent subchannels with equal SNR.
A (Synthetic) 4x4 "Keyhole" (Rank-1) Channel
Center Frequency: 2450 MHz

Shown are the eigenvalues of the normalized matrix channel H (right multiplied by it's conjugate transpose) when the four channels of the MCT are summed through a passive combiner, split through a passive divider, and then connected to MCR channels 1 through 4. The result is every MIMO researcher's worst nightmare: 1 strongly-dominant subchannel (meaning a lot of wasted transmit hardware!).
A 4x4 Close-Range LOS Channel
Center Frequency: 2450 MHz

Here, channels 1-4 of the MCT are connected to a 4-element monopole array, channels 1-4 of the MCR are connected to a similar 4-element array, and the two arrays are separated by about 1 meter in a small room with line of sight (LOS). Shown are the eigenvalues of the normalized matrix channel H (right multiplied by it's conjugate transpose). The result is not quite as bad as a keyhole, but still it's hard to get too excited about a second subchannel that is more than 10 dB lower in SNR most of the time...
A 4x4 Close-Range Non-LOS Channel
Center Frequency: 2450 MHz

This is the exact same experiment as above, except in this case a metal sheet 1-m wide by 2-m high has been placed between the two arrays, blocking the LOS. Note that the second subchannel is now within 5 dB of the dominant subchannel much of the time, which is much more interesting from a MIMO perspective.


MCR Demonstration at the 2005 MPRG Wireless Personal Communications Symposium

Here the MCR is set up to monitor the 2.4 GHz ISM band inside one of conference rooms in the hotel, using a 4 element monopole array (shown fastened to the door). The MCT was not used (we didn't want to jam anyone!); instead the system simply captured the ambient IEEE 802.11b wireless LAN signals in the building and analyzed them in real time.
This is a screen shot of the display while the system was in operation. The spectral and temporal displays ("1" and "2") in the upper left were updated continuously in real time (40 MHz span updated about 10 times/second) while the rest of the display was only updated when a signal was detected. Shown in this case is a relatively good example of a signal which is incident with a nearly-planar wavefront, and thus is a good candidate for beamforming. This doesn't happen all the time, but is surprisingly common given the cluttered environment. Generally, however, this demo is pretty effective in showing that beam-steering (in the conventional sense) is usually not so good -- see plot no. 6. Click here for a higher-resolution image. (Note: Some browsers automatically reduce resolution -- You may have to download this and view it in a proper JPEG viewer to actually see it in the intended resolution)


MCT Arb Testing

Here's a spectrum analyzer measurement of the MCT output when the digital baseband section is generating a random I-Q modulation, pulse shaping it, and upconverting to 1250 MHz, where it is measured. In this case the synthesized bandwidth is greater than 6 MHz. Bandwidths up to 45 MHz can be synthesized.


Some Signals Observed

HDTV
Center Frequency: 497 MHz

This HDTV signal was observed using a scanner whip antenna with a 20 dB amplifier attached directly to the "Low Band" input of RFDC Ch. 1. This is 512 samples of buffered 52 MSPS complex baseband output from the digital IF board, routed through the corner turner and captured by the MCR's PC.
IEEE 802.11b ("Wi-Fi")
(11 Mb/s DSSS/CCK)
Center Frequency: 2410 MHz

This signal was observed using a scanner whip antenna attached directly to the "Mid Band" input of RFDC Ch. 1. This is 512 samples of buffered 52 MSPS complex baseband output from the digital IF board, routed through the corner turner and captured by the MCR's PC.


NSF Acknowledgment and Disclaimer

This material is based upon work supported by the National Science Foundation under Grant No. ECS-0215990. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.