<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Yulong Ming&apos;s Homepage</title><description>My profile, blog, and tools I built.</description><link>https://myl7.org/</link><item><title>用 OpenWrt 默认 LuCI 界面连接 eduroam（i.e., 802.1x EAP）作为出口</title><link>https://myl7.org/posts/connect-eduroam-using-openwrt-luci/</link><guid isPermaLink="true">https://myl7.org/posts/connect-eduroam-using-openwrt-luci/</guid><description>网上教程很多，但都是针对命令行或定制 UI 的。本文截图记录了用默认的 LuCI web 界面连接 eduroam 这类 802.1x EAP 安全性 Wi-Fi 节点的过程，不用开命令行 SSH，只用手机都行。额外探讨了一些容易配错的选项。</description><pubDate>Fri, 10 Jul 2026 16:00:00 GMT</pubDate></item><item><title>MHcast: Metadata-hiding Multicast with Resilience against Malicious Clients and Servers</title><link>https://myl7.org/posts/mhcast/</link><guid isPermaLink="true">https://myl7.org/posts/mhcast/</guid><description>In this paper, we present MHcast, a high-throughput metadata-hiding multicast system. MHcast uses two distributed comparison functions with a public mapping to construct a multicast scheme. MHcast enforces a mailbox access control scheme to prevent malicious clients from disrupting multicast through malformed or junk messages. Additionally, MHcast provides an identification scheme that allows honest clients to assist an honest server in identifying the malicious server. We implement and evaluate MHcast in comparison to other metadata-hiding systems. Experiments show that MHcast’s throughput is 2–10× higher than repeatedly using the unicast protocol when the group size is only 16. MHcast supports groups of any size, from 0 to N, where N is the total number of clients.</description><pubDate>Fri, 10 Jul 2026 16:00:00 GMT</pubDate></item><item><title>P²RAG: Efficient Privacy-Preserving RAG Service Supporting Arbitrary Top-k Retrieval</title><link>https://myl7.org/posts/p2rag/</link><guid isPermaLink="true">https://myl7.org/posts/p2rag/</guid><description>Unlike existing systems, P²RAG avoids sorting candidate documents. Instead, it uses an interactive bisection method to determine the set of top-k documents. For security, P²RAG uses secret sharing on two semi-honest non-colluding servers to protect the data owner&apos;s database and the user&apos;s prompt. It enforces restrictions and verification to defend against malicious users and tightly bounds the information leakage of the database. The experiments show that P²RAG is 3--300× faster than the state-of-the-art PRAG for k = 16--1024.</description><pubDate>Wed, 10 Jun 2026 16:00:00 GMT</pubDate></item></channel></rss>