<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Academic Papers on Clément Joly – Open-Source, Rust &amp; SQLite</title><link>https://cj.rs/tags/academic-papers/</link><description>Recent content in Academic Papers on Clément Joly – Open-Source, Rust &amp; SQLite</description><image><title>Clément Joly – Open-Source, Rust &amp; SQLite</title><url>https://cj.rs/images/open-graph-pages.jpg</url><link>https://cj.rs/images/open-graph-pages.jpg</link></image><generator>Hugo</generator><language>en</language><copyright>Clément Joly</copyright><lastBuildDate>Mon, 04 May 2026 19:41:10 +0100</lastBuildDate><atom:link href="https://cj.rs/tags/academic-papers/index.xml" rel="self" type="application/rss+xml"/><item><title>Academic Papers 2021</title><link>https://cj.rs/blog/academic-papers-2021/</link><pubDate>Sat, 21 Aug 2021 20:18:46 +0000</pubDate><guid>https://cj.rs/blog/academic-papers-2021/</guid><description>&lt;p&gt;Here is some interesting papers I’ve read this year.&lt;/p&gt;
&lt;h2 id="weaponizing-middleboxes-for-tcp-reflected-amplification"&gt;Weaponizing Middleboxes for TCP Reflected Amplification&lt;/h2&gt;
&lt;p&gt;&lt;a href="https://www.usenix.org/system/files/sec21-bock.pdf"&gt;https://www.usenix.org/system/files/sec21-bock.pdf&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;This paper uses the fact that some middleboxes are non-compliant TCP stacks. In particular, middleboxes sometimes see only one side of a connection and as a result, it’s possible that they answer spoofed packets. After an initial training (with a genetic algorithm) on a subset of known censoring networks to optimize amplification factors, the authors present the results of applying this to the whole IPv4 internet.&lt;/p&gt;</description><content:encoded><![CDATA[<p>Here is some interesting papers I’ve read this year.</p>
<h2 id="weaponizing-middleboxes-for-tcp-reflected-amplification">Weaponizing Middleboxes for TCP Reflected Amplification</h2>
<p><a href="https://www.usenix.org/system/files/sec21-bock.pdf">https://www.usenix.org/system/files/sec21-bock.pdf</a></p>
<p>This paper uses the fact that some middleboxes are non-compliant TCP stacks. In particular, middleboxes sometimes see only one side of a connection and as a result, it’s possible that they answer spoofed packets. After an initial training (with a genetic algorithm) on a subset of known censoring networks to optimize amplification factors, the authors present the results of applying this to the whole IPv4 internet.</p>
<h2 id="xor-filters-faster-and-smaller-than-bloom-and-cuckoo-filters">Xor Filters: Faster and Smaller Than Bloom and Cuckoo Filters</h2>
<p><a href="https://arxiv.org/abs/1912.08258">https://arxiv.org/abs/1912.08258</a></p>
<p>A bit like bloom filters, but read-only, more compact and faster. The companion blog post is a good introduction: <a href="https://lemire.me/blog/2019/12/19/xor-filters-faster-and-smaller-than-bloom-filters/">https://lemire.me/blog/2019/12/19/xor-filters-faster-and-smaller-than-bloom-filters/</a>.</p>
<h2 id="to-be-continued">To be continued</h2>
<p>I’ll update this page through 2021 if I want to share other papers. In the meantime, here are other posts listing papers:</p>
<ul>
<li><a href="https://ordep.dev/posts/my-favorite-papers">https://ordep.dev/posts/my-favorite-papers</a></li>
<li><a href="https://github.com/papers-we-love/papers-we-love">https://github.com/papers-we-love/papers-we-love</a></li>
<li><a href="https://github.com/facundoolano/software-papers">https://github.com/facundoolano/software-papers</a></li>
</ul>
]]></content:encoded></item><item><title>Linux Plumbers Conference 2020 (talk &amp; paper)</title><link>https://cj.rs/paper/lpc-2020/</link><pubDate>Fri, 14 Aug 2020 13:26:11 +0200</pubDate><guid>https://cj.rs/paper/lpc-2020/</guid><description>Evaluation of Tail Call Costs in eBPF</description><content:encoded><![CDATA[<ul>
<li>Co-authored a <a href="./evaluation-of-tail-call-costs-in-ebpf.pdf">paper</a> and gave a <a href="https://youtu.be/qBpPB3pnh1A?t=7125">talk</a> (<a href="./slides-evaluation-of-tail-call-costs-in-ebpf.pdf">slides</a>) to present the use of existing and custom benchmarking tools</li>
<li><a href="https://www.linuxplumbersconf.org/event/7/contributions/676/">https://www.linuxplumbersconf.org/event/7/contributions/676/</a></li>
</ul>
<div style="position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;">
      <iframe allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" loading="eager" referrerpolicy="strict-origin-when-cross-origin" src="https://www.youtube-nocookie.com/embed/qBpPB3pnh1A?start=7125?autoplay=0&amp;controls=1&amp;end=0&amp;loop=0&amp;mute=0&amp;start=0" style="position: absolute; top: 0; left: 0; width: 100%; height: 100%; border:0;" title="LPC 2020 - Networking and BPF Summit - Day 1"></iframe>
    </div>

]]></content:encoded></item><item><title>NOMS 2020 Conference (co-author)</title><link>https://cj.rs/paper/noms-2020/</link><pubDate>Thu, 01 Aug 2019 13:26:11 +0200</pubDate><guid>https://cj.rs/paper/noms-2020/</guid><description>The Rise of eBPF for Non-Intrusive Performance Monitoring</description><content:encoded><![CDATA[<ul>
<li><a href="./the-rise-of-eBPF-for-non-intrusive-performance-monitoring.pdf">Paper</a> submited to Network Operations and Management Symposium (NOMS) 2020</li>
<li><a href="https://www.researchgate.net/publication/342016524_The_rise_of_eBPF_for_non-intrusive_performance_monitoring">https://www.researchgate.net/publication/342016524_The_rise_of_eBPF_for_non-intrusive_performance_monitoring</a></li>
</ul>
]]></content:encoded></item></channel></rss>