Repository navigation
Expand file tree
/
Copy pathindex.html
More file actions
71 lines (65 loc) · 8.28 KB
/
Copy pathindex.html
File metadata and controls
71 lines (65 loc) · 8.28 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width,initial-scale=1">
<title>MeshBench reports</title>
<meta name="description" content="Studies of MeshCore firmware behaviour, run on real firmware against a simulated radio and channel.">
<link rel="icon" href="brand/favicon.svg" type="image/svg+xml">
<link rel="stylesheet" href="styles.css">
</head>
<body>
<div class="brandbar"><a href="https://meshbench.github.io/" aria-label="MeshBench"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 324.6 49.2" fill="none" stroke-linecap="round" stroke-linejoin="round" role="img" aria-label="MeshBench"><g transform="translate(-6.25,-11.31) scale(1.3300)">
<path d="M 9.50 34.50 L 18.00 43.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 18.00 43.00 L 32.00 30.50" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 18.00 43.00 L 20.50 23.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 20.50 23.00 L 29.50 11.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 29.50 11.00 L 40.50 19.50" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 32.00 30.50 L 39.00 40.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 40.50 19.50 L 39.00 40.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 9.50 34.50 L 29.50 11.00" stroke="currentColor" stroke-width="1.8" opacity="0.26"/>
<path d="M 9.50 34.50 L 20.50 23.00" stroke="currentColor" stroke-width="3.2"/>
<path d="M 20.50 23.00 L 32.00 30.50" stroke="currentColor" stroke-width="3.2"/>
<path d="M 32.00 30.50 L 40.50 19.50" stroke="currentColor" stroke-width="3.2"/>
<circle cx="18.00" cy="43.00" r="2.50" fill="currentColor" opacity="0.4"/>
<circle cx="29.50" cy="11.00" r="2.50" fill="currentColor" opacity="0.4"/>
<circle cx="39.00" cy="40.00" r="2.50" fill="currentColor" opacity="0.4"/>
<circle cx="20.50" cy="23.00" r="3.80" fill="currentColor"/>
<circle cx="32.00" cy="30.50" r="3.80" fill="currentColor"/>
<circle cx="9.50" cy="34.50" r="4.80" fill="currentColor"/>
<circle cx="40.50" cy="19.50" r="4.80" fill="currentColor"/>
</g><g transform="translate(70.0,40.7)" fill="currentColor"><path d="M3.04 0V-32.2H14.31L19.87 -4.14H20.7L26.27 -32.2H37.54V0H31.65V-27.74H30.82L25.3 0H15.27L9.75 -27.74H8.92V0Z M53.36 0.64Q49.96 0.64 47.36 -0.8Q44.76 -2.25 43.31 -4.9Q41.86 -7.54 41.86 -11.13V-11.68Q41.86 -15.27 43.29 -17.92Q44.71 -20.56 47.29 -22.01Q49.86 -23.46 53.27 -23.46Q56.63 -23.46 59.11 -21.96Q61.59 -20.47 62.97 -17.82Q64.35 -15.18 64.35 -11.68V-9.71H47.75Q47.84 -7.36 49.5 -5.89Q51.15 -4.42 53.54 -4.42Q55.98 -4.42 57.13 -5.47Q58.28 -6.53 58.88 -7.82L63.62 -5.34Q62.97 -4.14 61.76 -2.74Q60.54 -1.33 58.51 -0.35Q56.49 0.64 53.36 0.64ZM47.79 -14.03H58.47Q58.28 -16.01 56.88 -17.2Q55.48 -18.4 53.22 -18.4Q50.88 -18.4 49.5 -17.2Q48.12 -16.01 47.79 -14.03Z M78.15 0.64Q73.69 0.64 70.84 -1.29Q67.99 -3.22 67.39 -6.81L72.73 -8.19Q73.05 -6.58 73.81 -5.66Q74.57 -4.74 75.69 -4.35Q76.82 -3.96 78.15 -3.96Q80.18 -3.96 81.14 -4.67Q82.11 -5.38 82.11 -6.44Q82.11 -7.5 81.19 -8.07Q80.27 -8.65 78.25 -9.02L76.96 -9.25Q74.57 -9.71 72.59 -10.51Q70.61 -11.32 69.41 -12.74Q68.22 -14.17 68.22 -16.42Q68.22 -19.83 70.7 -21.64Q73.19 -23.46 77.23 -23.46Q81.05 -23.46 83.58 -21.76Q86.11 -20.06 86.89 -17.3L81.51 -15.64Q81.14 -17.39 80.02 -18.12Q78.89 -18.86 77.23 -18.86Q75.58 -18.86 74.7 -18.29Q73.83 -17.71 73.83 -16.7Q73.83 -15.59 74.75 -15.06Q75.67 -14.54 77.23 -14.26L78.52 -14.03Q81.1 -13.57 83.19 -12.81Q85.28 -12.05 86.5 -10.65Q87.72 -9.25 87.72 -6.85Q87.72 -3.27 85.12 -1.31Q82.52 0.64 78.15 0.64Z M91.95 0V-32.2H97.75V-20.01H98.58Q98.95 -20.75 99.73 -21.48Q100.51 -22.22 101.82 -22.7Q103.13 -23.18 105.16 -23.18Q107.82 -23.18 109.82 -21.96Q111.83 -20.75 112.93 -18.61Q114.03 -16.47 114.03 -13.62V0H108.24V-13.16Q108.24 -15.73 106.97 -17.02Q105.71 -18.31 103.36 -18.31Q100.69 -18.31 99.22 -16.54Q97.75 -14.77 97.75 -11.59V0Z M118.36 0V-5.34H122.59V-26.86H118.36V-32.2H134.92Q137.86 -32.2 140.05 -31.21Q142.23 -30.22 143.45 -28.41Q144.67 -26.59 144.67 -24.06V-23.6Q144.67 -21.39 143.84 -19.99Q143.01 -18.58 141.89 -17.82Q140.76 -17.07 139.75 -16.74V-15.92Q140.76 -15.64 141.96 -14.88Q143.15 -14.12 144 -12.7Q144.85 -11.27 144.85 -8.97V-8.51Q144.85 -5.84 143.61 -3.93Q142.37 -2.02 140.19 -1.01Q138 0 135.1 0ZM128.66 -5.52H134.37Q136.34 -5.52 137.56 -6.49Q138.78 -7.45 138.78 -9.25V-9.71Q138.78 -11.5 137.59 -12.47Q136.39 -13.43 134.37 -13.43H128.66ZM128.66 -18.95H134.27Q136.16 -18.95 137.38 -19.92Q138.6 -20.88 138.6 -22.59V-23.05Q138.6 -24.79 137.4 -25.74Q136.21 -26.68 134.27 -26.68H128.66Z M159.57 0.64Q156.17 0.64 153.57 -0.8Q150.97 -2.25 149.52 -4.9Q148.07 -7.54 148.07 -11.13V-11.68Q148.07 -15.27 149.5 -17.92Q150.93 -20.56 153.5 -22.01Q156.08 -23.46 159.48 -23.46Q162.84 -23.46 165.32 -21.96Q167.81 -20.47 169.19 -17.82Q170.57 -15.18 170.57 -11.68V-9.71H153.96Q154.05 -7.36 155.71 -5.89Q157.37 -4.42 159.76 -4.42Q162.2 -4.42 163.35 -5.47Q164.5 -6.53 165.09 -7.82L169.83 -5.34Q169.19 -4.14 167.97 -2.74Q166.75 -1.33 164.73 -0.35Q162.7 0.64 159.57 0.64ZM154.01 -14.03H164.68Q164.5 -16.01 163.09 -17.2Q161.69 -18.4 159.44 -18.4Q157.09 -18.4 155.71 -17.2Q154.33 -16.01 154.01 -14.03Z M174.89 0V-22.82H180.6V-19.83H181.42Q181.98 -21.02 183.49 -22.1Q185.01 -23.18 188.09 -23.18Q190.76 -23.18 192.76 -21.96Q194.76 -20.75 195.87 -18.61Q196.97 -16.47 196.97 -13.62V0H191.18V-13.16Q191.18 -15.73 189.91 -17.02Q188.65 -18.31 186.3 -18.31Q183.63 -18.31 182.16 -16.54Q180.69 -14.77 180.69 -11.59V0Z M213.12 0.64Q209.81 0.64 207.12 -0.74Q204.42 -2.12 202.86 -4.74Q201.3 -7.36 201.3 -11.09V-11.73Q201.3 -15.46 202.86 -18.08Q204.42 -20.7 207.12 -22.08Q209.81 -23.46 213.12 -23.46Q216.38 -23.46 218.73 -22.31Q221.08 -21.16 222.53 -19.16Q223.97 -17.16 224.43 -14.63L218.82 -13.43Q218.64 -14.81 217.99 -15.92Q217.35 -17.02 216.18 -17.66Q215 -18.31 213.26 -18.31Q211.51 -18.31 210.11 -17.55Q208.7 -16.79 207.9 -15.29Q207.09 -13.8 207.09 -11.64V-11.18Q207.09 -9.02 207.9 -7.52Q208.7 -6.03 210.11 -5.27Q211.51 -4.51 213.26 -4.51Q215.88 -4.51 217.24 -5.87Q218.59 -7.22 218.96 -9.43L224.57 -8.1Q223.97 -5.66 222.53 -3.66Q221.08 -1.66 218.73 -0.51Q216.38 0.64 213.12 0.64Z M228.53 0V-32.2H234.32V-20.01H235.15Q235.52 -20.75 236.3 -21.48Q237.08 -22.22 238.39 -22.7Q239.71 -23.18 241.73 -23.18Q244.4 -23.18 246.4 -21.96Q248.4 -20.75 249.5 -18.61Q250.61 -16.47 250.61 -13.62V0H244.81V-13.16Q244.81 -15.73 243.55 -17.02Q242.28 -18.31 239.94 -18.31Q237.27 -18.31 235.8 -16.54Q234.32 -14.77 234.32 -11.59V0Z"/></g></svg></a></div>
<div class="wrap">
<header>
<div class="eyebrow">Field studies</div>
<h1>Reports</h1>
<p class="stand">Studies of how MeshCore actually behaves, run on real firmware against a
simulated radio and a simulated channel.</p>
</header>
<h2>What this is</h2>
<p>MeshBench compiles the real MeshCore application natively, one operating-system process per node,
and runs it against a model of an SX1262 and a model of the air. The firmware and its radio driver
are unmodified, so the decisions in these reports are the same decisions the code makes on hardware.
What is simulated is everything the radio would have told it.</p>
<p>That combination makes a particular kind of question answerable: what happens to this firmware
when the hardware misbehaves in precisely this way, or when a setting changes across a whole network
at once. Neither is something a fleet of real radios will do on request.</p>
<h2>Reports</h2>
<a class="report" href="1-17-listen-before-talk/">
<h3>Testing 1.17's listen-before-talk on a radio that lies</h3>
<p>MeshCore 1.17 rewrote listen-before-talk to survive LoRa interrupt flags that get stuck. On a
well-behaved radio the change is invisible. So we built a radio that sticks, and the two versions
separated by a factor of 43.</p>
<div class="meta"><span>MeshCore 1.16.0 vs 1.17.0</span><span>ScotMesh, 154 nodes</span>
<span>14 runs</span></div>
</a>
<footer>
Every figure is measured output from MeshBench: real MeshCore firmware making real decisions
against a simulated radio and channel. The interpretation and the prose are Claude Opus 5's.<br>
Results are a best case throughout. There is no multipath, the terrain is bare earth, and there is
no body loss, oscillator error, or interference beyond the mesh's own traffic. Every one of those
omissions makes real links worse.
</footer>
</div>
</body>
</html>