-
Notifications
You must be signed in to change notification settings - Fork 0
/
the-discredited-levine-2017-approach-is-still-used.html
186 lines (173 loc) · 10.1 KB
/
the-discredited-levine-2017-approach-is-still-used.html
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
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
<!DOCTYPE html>
<html lang="en">
<head>
<link rel="stylesheet" href="/theme/style/base.min.css?2189187c">
<title>Hyphanet</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1, maximum-scale=1" />
<link href="https://www.hyphanet.org/feeds/all.atom.xml" type="application/atom+xml" rel="alternate" title="Hyphanet Full Atom Feed" />
<link rel="alternate" hreflang="en" href="https://www.hyphanet.org/the-discredited-levine-2017-approach-is-still-used.html" />
<link rel="alternate" hreflang="ru" href="https://www.hyphanet.org/ru/the-discredited-levine-2017-approach-is-still-used.html" />
<link rel="alternate" hreflang="fr" href="https://www.hyphanet.org/fr/the-discredited-levine-2017-approach-is-still-used.html" />
<link rel="alternate" hreflang="x-default" href="https://www.hyphanet.org /the-discredited-levine-2017-approach-is-still-used.html" />
<link rel="canonical" href="https://www.hyphanet.org/the-discredited-levine-2017-approach-is-still-used.html" />
<meta property="og:title" content="Hyphanet" />
<meta property="og:type" content="article" />
<meta property="og:url" content="https://www.hyphanet.org" />
<meta property="og:image" content="https://www.hyphanet.org/" />
<meta property="og:image:secure_url" content="https://www.hyphanet.org/theme/images/logo-blue.png" />
<meta property="og:description" content="Hyphanet is a peer-to-peer platform for censorship-resistant communication and publishing." />
<meta name="twitter:card" content="summary">
<meta name="twitter:title" content="Hyphanet" />
<meta name="twitter:description" content="Hyphanet is a peer-to-peer platform for censorship-resistant communication and publishing." />
<meta name="twitter:image" content="https://www.hyphanet.org/theme/images/logo-blue.png" />
</head>
<body id="index" class="home">
<div>
<nav id="menu">
<a href="https://www.hyphanet.org/">Hyphanet</a>
<a href="https://www.hyphanet.org/pages/about.html">About</a>
<a href="https://www.hyphanet.org/pages/volunteer.html">Volunteer</a>
<a href="https://www.hyphanet.org/pages/documentation.html">Documentation</a>
<a href="https://www.hyphanet.org/pages/download.html">Download</a>
<a href="https://www.hyphanet.org/pages/help.html">Help</a>
</nav><!-- /#menu -->
<aside class="social">
<a href="https://twitter.com/freenetproject">Twitter</a>
<a rel="me" title="Hyphanet News and Info in the Fediverse" href="https://floss.social/@Freenet">Mastodon</a>
</aside>
<nav id="language">
<span>Language</span>
<a href="https://www.hyphanet.org/ru/the-discredited-levine-2017-approach-is-still-used.html">ru</a>
<a href="https://www.hyphanet.org/fr/the-discredited-levine-2017-approach-is-still-used.html">fr</a>
</nav>
</div>
<main>
<header id="banner" class="body">
<h1>The discredited Levine 2017 approach is still used</h1>
</header><!-- /#banner -->
<section id="content" class="body">
<div class="post-info">
<time class="published" datetime="2022-08-28T00:00:00+02:00">
Sun 28 August 2022
</time>
<address class="vcard author">
By <a class="url fn" href="https://www.hyphanet.org/author/freenet-contributors.html">Freenet Contributors</a>
</address>
</div><!-- /.post-info -->
<div class="entry-content">
<p>In april 2020 the approach <a href="http://ceur-ws.org/Vol-1873/IWPE17_paper_12.pdf">by Levine et al.</a> to track downloaders was
proven to be wrong:
<em>Their false positives rate is wrong, their math is wrong, and their model is wrong.
Therefore results you get when using their method are false.</em> Yet this
approach is still used two years later to get warrants against Freenet users.</p>
<p>The Levine group themselves changed their method in response (they
posted this to conference proceedings, containing claims that are obviously
false), but they still use the 2017 method for arguments in court.</p>
<p>Therefore we are now providing the arguments that disprove it directly
on the Freenet Project website to ensure that they are found when
innocents are caught in that dragnet.</p>
<h1 id="wrong-false-positives-rate">Wrong false positives rate</h1>
<p>The core pillar of the detection they name is their claim of a 2.3%
<strong>false positives rate</strong>. But this claim is <strong>wrong</strong>, because they only
reach it through many <strong>false assumptions</strong>:</p>
<ul>
<li>
<p>They <strong>ignore</strong> that friend-of-a-friend routing breaks their metric when</p>
<ul>
<li>an intermediary node, or</li>
<li>the observing node <strong>has many connections</strong>.</li>
</ul>
<p>which is not the rare case but <strong>the normal case</strong>.</p>
</li>
<li>
<p>They assume that they only get a false positive, if a request for a
given file reached them with both HTL 18 and HTL 16 or with both HTL 17 and HTL 16.
But the
routing algorithm within Freenet causes them to <strong>almost always</strong>
receive requests from a given node <strong>over the same route</strong>. So they
will have the same HTL, regardless of the number of hops from the
source to an observer. Therefore Their 2.3% false positives rate contains a mixture
of</p>
<ul>
<li>
<p>the probability of <strong>two people</strong> requesting the file in the same
interval and</p>
</li>
<li>
<p>the rate of <strong>routing-changes</strong> within Freenet (for example because a
node on the path went offline). If a request from a given peer is
received both from HTL 17 and from HTL 16 (or 18 and 16) then routing changed,
otherwise this should not happen.</p>
</li>
</ul>
<p>Their <strong>false positives rate</strong> when measuring with only one node is
therefore <strong>meaningless</strong>.</p>
</li>
</ul>
<h1 id="wrong-math">Wrong math</h1>
<p>In addition <strong>their math is wrong</strong>:</p>
<blockquote>
<p>We construct a model by assuming that each request the
downloader makes is sent to exactly one of its peers, and that
the selection of that peer is made <strong>uniformly at random</strong>.</p>
</blockquote>
<p>This does not take friend of a friend routing into account. Therefore
their math is wrong: It does not match the actual selection of peers,
so the results are meaningless for the actual Freenet.</p>
<p>They expect even share peer selection, but Freenet does not use even share.</p>
<h1 id="wrong-model">Wrong model</h1>
<p>And their <strong>model</strong> of routing in Freenet, required to understand how
their measurement works and what it observes, <strong>is wrong</strong>:</p>
<blockquote>
<p>a simple expected fraction of 1/degree for the adjacent and
(1/degree)² for the two-hop case.</p>
</blockquote>
<p>This does not take the degree of the measuring node into account,
therefore it is not a model of routing in Freenet.</p>
<p>This is in addition to the wrong math, an error they repeat here:
Freenet does not use even share. Assuming scaling by inverse degree is wrong.</p>
<h1 id="summary-tldr">Summary (TLDR)</h1>
<p>Their false positives rate is wrong, their math is wrong, and their model is wrong. Therefore results you get when using their method are false.</p>
<h1 id="remarks">Remarks</h1>
<p>To somewhat save the grace for the Levine-team: They at least tried to
actually measure the false positives rate. They did it wrong and drew
false conclusions, and that they tried doesn’t make it right and it
doesn’t excuse persecuting people based on their flawed reasoning, but
at least they tried.</p>
<p>That the Levine-team <strong>did not contact Freenet developers</strong> prior to
publication is <strong>inexcusable</strong>, though. It’s like publishing a paper based
on evaluations of particle beams from CERN without ever talking to
someone from CERN. The minimum of due dilligence would be to write an email to press -at-
freenetproject.org saying
<em>“Hi, we found a method to track Freenet downloaders and drafted a paper based on that. Could you have a look to see whether we missed something?”</em></p>
<h1 id="final-words">Final words</h1>
<p>If you want to know the actual requirements for calculating a false
positives rate in Freenet, read the article <a href="https://freenetproject.org/statistical-results-without-false-positives-check-are-most-likely-wrong.html">Statistical results without false positives check are most likely wrong</a>.</p>
<p>While the Levine group has a false positives check, their check is
wrong. They measured the wrong metric. We have explained patiently
where the Levine group made mistakes. It is hard to understand when
assuming scientific integrity that they still claim in court that
their 2017 method is robust even after they changed their approach
themselves.</p>
</div><!-- /.entry-content -->
<a href="archives.html">News Archives</a>
</section>
</main>
<footer>
<header>
<h2>Hyphanet</h2>
<p>Navigate with Freedom</p>
</header>
<ul class="social">
<a href="https://twitter.com/freenetproject">Twitter</a>
<a rel="me" title="Hyphanet News and Info in the Fediverse" href="https://floss.social/@Freenet">Mastodon</a>
</ul>
<div id="contact">
<span style="display:inline-block; unicode-bidi:bidi-override; direction:rtl;" onmouseover="this.innerText=this.innerText.split('').reverse().join(''); this.style.unicodeBidi='';this.style.direction=''; this.removeAttribute('onmouseover');">gro.tcejorpteneerf@sserp</span></br>
<span style="display:inline-block; unicode-bidi:bidi-override; direction:rtl;" onmouseover="this.innerText=this.innerText.split('').reverse().join(''); this.style.unicodeBidi='';this.style.direction=''; this.removeAttribute('onmouseover');">gro.tcejorpteneerf@troppus</span></br>
<span>IRC: <a href="https://web.libera.chat/?nick=FollowRabbit|?#freenet">#freenet on irc.libera.chat</a></span></br>
</div>
<p id="copyright">Licensed under the <a href="https://www.gnu.org/licenses/fdl-1.3.html">GDFL</a>. <a href="https://github.com/hyphanet/website">Website source repository</a>, <a href="/pages/download.html#privacy-policy">Privacy Policy</a></p>
</footer></body>
</html>