IEEE 802.15.4 subsystem
Sign in or create your account | Project List | Help
IEEE 802.15.4 subsystem Git Source Tree
Root/
Source at commit e610a9ec2059c74d155e82536a68b2a3214ea839 created 12 years 8 months ago. By Werner Almesberger, TODO: removed EPERM, added usb_read_bulk vs. signal, toolchain, DFU on U1010 | |
---|---|
1 | Antenna measurements |
2 | ==================== |
3 | |
4 | The objective of antenna measurements is to determine how much energy the |
5 | antenna transfers at different frequencies. For this, we set up a sender, |
6 | a receiver, connect one to the antennas being tested, and the other to an |
7 | arbitrarily chosen lab antenna. |
8 | |
9 | Since none of the items (sender, receiver, lab antenna) are calibrated, |
10 | we can only compare antennas but we cannot determine any absolute |
11 | characteristics. |
12 | |
13 | |
14 | Preparing a measurement run |
15 | --------------------------- |
16 | |
17 | Before measuring the characteristics of an antenna, we need to set up the |
18 | test environment and obtain a number of filtering parameters. The filters |
19 | are used to reduce the effect of noise on the measurements and to suppress |
20 | contamination from other sources. |
21 | |
22 | 1) Install transmitter and receiver. The transmitter is an atusb or atusd |
23 | board, the receiver an USRP2+XCVR2450 with the antenna to test. |
24 | |
25 | (The same setup may also work with a USRP1 or UN210, and a RFX2400 |
26 | board.) |
27 | |
28 | Both should be spaced at least twenty times the wavelength (12.5 cm at |
29 | 2405 MHz), or 2.5 m apart. For test runs that can be compared with each |
30 | other, antenna placement and orientation have to be exactly the same. |
31 | |
32 | The sender runs tools/atrf-txrx/atrf-txrx, the receiver runs utilities |
33 | from gnuradio. |
34 | |
35 | 2) Obtain baseline performance values. For example, activate the sender |
36 | with |
37 | |
38 | atrf-txrx -f 2455 -p 0.5 -T +0.5 |
39 | |
40 | Emit a constant wave at 2455+0.5 MHz with a power of 0.5 dBm or 1.1 mW. |
41 | |
42 | Monitor the received signal with |
43 | |
44 | usrp2_fft.py -f 2455.5M -d 16 |
45 | |
46 | Record the range in which the frequency peak falls. Variations of a few |
47 | dB are to be expected. |
48 | |
49 | 3) Generate a series of sample for a specific setting. |
50 | |
51 | Example: |
52 | |
53 | The following script sets up the transmitter, lets it "warm up" for ten |
54 | seconds, then takes 100 measurements, stored in files tmp00 through |
55 | tmp99 in a directory $PWD/100/. |
56 | |
57 | In this setup, the receiver's gnuradio runs on a different host than |
58 | the sender. Therefore we use ssh and pass the directory from $PWD. |
59 | |
60 | atrf-txrx -f 2455 -p 2.6 -T +0.5 \ |
61 | 'sleep 10; |
62 | for a in 0 1 2 3 4 5 6 7 8 9; do |
63 | for b in 0 1 2 3 4 5 6 7 8 9; do |
64 | ssh ws usrp2_rx_cfile.py -d 16 -f 2455.5M -g 46 -N 1124 \ |
65 | '$PWD'/100/tmp$a$b |
66 | done |
67 | done' |
68 | |
69 | Each measurement obtains 1124 samples, 1024 samples for the FFT and |
70 | 100 samples to cut off (see below). |
71 | |
72 | 4) Determine the shape of the captured waves in the time domain, e.g., |
73 | with |
74 | |
75 | gnuplot |
76 | gnuplot> plot "<./avg 1 <100/tmp00" with lines |
77 | |
78 | "avg" outputs the magnitude of the recorded wave, averaging over the |
79 | specified number of sample. |
80 | |
81 | Some waves will probably show a peak in the first few samples. We need |
82 | to cut off these peaks in the later processing steps. In this example, |
83 | we will skip the first 100 samples. |
84 | |
85 | Besides the initial peak, the waves should be of comparable amplitude. |
86 | |
87 | 5) Verify the distribution in the frequency domain and determine the noise |
88 | floor. |
89 | |
90 | gnuplot> plot "<./fft -s 100 -d <100/tmp00" with lines |
91 | ^ |
92 | skip initial peak |
93 | |
94 | The spectrum should be U-shaped, with narrow peaks tens of dB above |
95 | the noise floor near the beginning and the end. Note that the noise |
96 | floor is curved and not perfectly flat. |
97 | |
98 | From this, we pick level of the noise floor. The value should be at or |
99 | slightly below the highest peaks of the noise between the large peaks |
100 | at the end of the spectrum. |
101 | |
102 | This noise floor value is used to filter uninteresting samples later |
103 | on, removing a constant bias from the results. |
104 | |
105 | In this example, we'll use a noise floor value of -60 dB. |
106 | |
107 | 6) Determine the "interesting" frequency range. For this, we consider all |
108 | the spectra of the measurements: |
109 | |
110 | gnuplot> plot "<for n in 100/tmp*; do ./fft -s 100 -d <$n;echo;done" \ |
111 | with lines |
112 | |
113 | There should be a thick noise band in the middle, with pronounced |
114 | narrow peaks at the edges. If there are one or two signals on top of |
115 | the noise band, some measurements have been compromised and need to be |
116 | removed or redone. We will do this in the next step. |
117 | |
118 | When zooming into the left peak, the "bins" which contribute to the |
119 | peaks can be identified. The range should be chosen with some |
120 | tolerance, since the frequency may shift a bit during the measurement |
121 | process. |
122 | |
123 | By not considering bins far from the peak, less noise is included in |
124 | the final result, complementing the filtering by noise threshold from |
125 | step 4). Restricting the bins also eliminates the second peak at the |
126 | end of the spectrum. |
127 | |
128 | In this example, we'll use a range from 0 to 20. |
129 | |
130 | 7) Obtain the peaks from all measurements |
131 | |
132 | gnuplot> plot "<for n in 100/tmp*; do ./fft -s 100 0 20 60 <$n;done" \ |
133 | with lines |
134 | ^ ^ ^ ^ |
135 | | | | | |
136 | skip, from step 4 | | threshold, 5) |
137 | lowest bin highest bin |
138 | |
139 | This should yield a jagged more or less horizontal line with values |
140 | differing by not more than 1-2 dB. If there are any large outliers, |
141 | they have been contaminated and should be dropped. |
142 | |
143 | 8) The final result for one measurement run can be obtained as follows: |
144 | |
145 | for n in 100/tmp*; do ./fft -s 100 0 20 60 <$n;done | ./range -v 2 |
146 | |
147 | In this example, "range" eliminates all outliers more than 2 dB from |
148 | the average and reports this. |
149 | |
150 | The output are three numbers: the average (after eliminating |
151 | outliers), the minimum, and the maximum. |
152 | |
153 | |
154 | Performing a measurement run |
155 | ---------------------------- |
156 | |
157 | The script "fscan" performs 100 scans of the 26 channels used by IEEE |
158 | 802.15.4. The frequency scan is the inner loop, so that slow changes |
159 | in environmental parameters (background noise, temperature, etc.) will |
160 | affect the spread of the results over the entire frequency range instead |
161 | of causing seemingly frequency-dependent distortions. |
162 | |
163 | The script is written for a setup that uses a pair of hosts, both |
164 | sharing the same file system. |
165 | |
166 | Usage: fscan out-dir [tx-power] |
167 | |
168 | The output directory must not exist yet. The transmit power is in dBm and |
169 | defaults to 2.6 dBm. |
170 | |
171 | Example: ./fscan testant |
172 | |
173 | The full run takes approximately half an hour. |
174 | |
175 | The results are filtered and averaged by the script "evscan". This script |
176 | contains the filtering parameters obtained in the preparation, described |
177 | above. |
178 | |
179 | Example: ./evscan testant >testant.out |
180 | |
181 | The output is a graph with frequency, average signal strength, minimum |
182 | and maximum. The format is compatible with gnuplot's "with errorbars" |
183 | (or "with errorlines") plot style. |
184 | |
185 | Finally, the results can be plotted with the script "plscan", which uses |
186 | gnuplot to output either in a window or to a PNG file. |
187 | |
188 | Usage: plscan [-o pngfile] file ... |
189 | |
190 | More than one graph can be plotted in the same run. The file name is used |
191 | as the title for each graph. Titles are truncated at the last dot. |
192 | |
193 | Example: ./plscan testant.out |
194 |