Multipath TCP (MPTCP) improves reliability and performance by enabling multiple network paths for a single connection, but its security remains a concern. In this paper we present MPTCPcap, a real-time and reusable analysis tool for monitoring MPTCP connections and simulating attacks. The tool tracks subflows, endpoints, and keys in real time and supports packet injection, modification, and dropping, enabling emulation of off-path, partial on-path, and full on-path adversaries. Using MPTCPcap, we reassess well-known MPTCPv0 attacks against the Linux MPTCPv1 implementation and identify two previously undocumented vulnerabilities: (i) an off-path attack that closes subflows in Linux’s MPTCP implementation, and (ii) a protocol-level weakness allowing an attacker with connection keys to prevent the creation of new subflows. Our results highlight how adversarial testing of MPTCP implementations can expose residual security issues and provide a practical basis for future security analysis and mitigation efforts. © 2026 Copyright for this paper by its authors.
MPTCPcap: a practical framework for security testing of multipath TCP / Mancini, R., Bianchi, F., Spognardi, A.. - 4198:(2026). (2026 Joint National Conference on Cybersecurity, ITASEC and SERICS 2026 Cagliari; Italy ).
MPTCPcap: a practical framework for security testing of multipath TCP
Riccardo Mancini;Federica Bianchi
;Angelo Spognardi
2026
Abstract
Multipath TCP (MPTCP) improves reliability and performance by enabling multiple network paths for a single connection, but its security remains a concern. In this paper we present MPTCPcap, a real-time and reusable analysis tool for monitoring MPTCP connections and simulating attacks. The tool tracks subflows, endpoints, and keys in real time and supports packet injection, modification, and dropping, enabling emulation of off-path, partial on-path, and full on-path adversaries. Using MPTCPcap, we reassess well-known MPTCPv0 attacks against the Linux MPTCPv1 implementation and identify two previously undocumented vulnerabilities: (i) an off-path attack that closes subflows in Linux’s MPTCP implementation, and (ii) a protocol-level weakness allowing an attacker with connection keys to prevent the creation of new subflows. Our results highlight how adversarial testing of MPTCP implementations can expose residual security issues and provide a practical basis for future security analysis and mitigation efforts. © 2026 Copyright for this paper by its authors.| File | Dimensione | Formato | |
|---|---|---|---|
|
Mancini_MPTCPcap_2026.pdf
accesso aperto
Tipologia:
Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza:
Creative commons
Dimensione
628.58 kB
Formato
Adobe PDF
|
628.58 kB | Adobe PDF |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


