Field Test -- October 15th, 1999

Information
Date & Time October 15th, 1999, 1:30pm - 4:00pm
Location Nichols Hall Roof and Clinton Lake dam
Participants Dan DePardo
Otto Koudelka
Artur Leung
Brian Sparks
Robert Washburn
Test Goal
  • Evaluate performance of the radios and compare result against data collected in laboratory;
  • Run Microsoft NetMeeting across the wireless link and evaluate performance;
  • Run FTP and evaluate link speed and utilization.
Tests Performed (1) One-way (up or down link) bit-error-rate test
(2) Two-way (up and down link) loop-back bit-error-rate test
(3) Application test -- PING, Microsoft NetMeeting, and FTP
Test Result
This was a very successful field test. Transmit power (ERP) was ~20dBm(100mW) and the distance between the two nodes was ~4 miles (~6.5km). The radio achieved bit-error-rate(BER) of 3e-7 with Eb/No of around 16dB, which is very closed to the performance obtained in a "wired" link in laboratory. NetMeeting ran great with no major picture quality defect, except when there were cars passing by in front of the radio at Clinton Lake dam. FTP transfer rate was 66.49kBps with NetMeeting running. With a 1Mbps wireless link, FTP utilized almost 50% of the link capacity while NetMeeting was still running in the background with no noticable picture quality degradation (remember the link is not error-free). Bit-error-rate on loop-back test was 6e-7, which verified the bit-error-rate of a one-way link with half the bit errors.

Overview of the Field Test

The main purpose of this field test was to determine the radios' performance. Therefore, they were configured as wireless ethernet bridges -- the radio would forward any ethernet packet received from the computer. This configuration allowed us to test the radios without relying on the RDRN backbone network. Here is a block diagram of the field-test setup:

Block Diagram

At Nichols Hall, the radio was connected to a laptop computer via 10BASE-T. A spectrum analyzer was utilized to monitor the intermediate frequency (IF) spectrum, and to record the signal-to-noise-ratio (SNR). Bit error rate (BER) of the wireless link was monitored via a bit-error-rate tester (BERT) when the remote node transmitted pseudo-random (PN) sequence.

The setup at Clinton Lake Dam was similar to the one at Nichols Hall except that there was no BERT. However, when performing BER measurement, the radio could re-route the receiving data to the transmitter, resulting in a loop-back test.

Before presenting the field test results, let's take a look at a satellite view of the two test locations:

Lawrence Satellite View

Pictures

Please click on the thumbnail to have a detailed look at each picture with explanation. A new window with an image 700x469 pixels in size will be displayed.
 
This was the test setup at Nichols Hall, showing the radio (Skittle), laptop computer, spectrum analyzer, and bit-error-rate tester.
This picture shows the location of the remote node (Twix) relative to the radio on the roof of Nichols Hall.  The remote node was located at the Clinton Lake dam, approximately 4miles (6.5km).
Remote node transmitting the bit pattern 0x01111110b (0x7Eh) at 1Mbps.  This is the spectrum of the bit pattern at the receiver IF (f_IF = 70MHz).
Remote node transmitting a 63-bit long PN sequence at 1Mbps.  This is the spectrum of the bit pattern at the receiver IF (f_IF = 70MHz).  Signal-to-noise ratio [ (C+N)/N ] was roughly 16dB.
Bit error rate of a one-way link with SNR shown in the previous picture.
This shows a snap-shot of the laptop computer screen, with Microsoft NetMeeting running.  The DOS window shows some PING tests between the two nodes.
Video-conferencing in action.  Robert and Brian were talking to folks at Clinton Lake, "Can you hear me?"
Another picture showing Robert and Brian having video conferencing.
The AGC level of the receiver.  The level is quite high......
Artur talking to folks at Clinton Lake.
Our field-test fleet.  Hey Dan......
The setup of the remote node (Twix).  It was similar to the setup at Nichols Hall except that there was no bit-error-rate tester.
Location of Nichols Hall when viewed from the Clinton Lake dam.
Dan was talking to folks at Nichols Hall.
A snap-shot of the laptop on the Clinton Lake node.  The windows show Brian on the remote site, with Artur taking picture of the screen.  The DOS window shows an FTP session with transfer rate of 66.49kBytes/sec (530kbps).
Loop-back test was performed.  The node at Nichols Hall transmitted PN sequence to the Clinton Lake node.  The Clinton Lake node received the sequence and re-transmitted it back to Nichols Hall.


Last Update: November 01, 1999 at 01:20pm (GMT-0600)