Security Issues in the Optimized Link State Routing Protocol Version 2 (Olsrv2)

Abstract

Mobile Ad hoc NETworks (MANETs) are leaving the confines of research laboratories, to find place in real-world deployments. Outside specialized domains (military, vehicular, etc.), city-wide community-networks are emerging, connecting regular Internet users with each other, and with the Internet, via MANETs. Growing to encompass more than a handful of "trusted participants", the question of preserving the MANET network connectivity, even when faced with careless or malicious participants, arises, and must be addressed.A first step towards protecting a MANET is to analyze the vulnerabilities of the routing protocol, managing the connectivity. By understanding how the algorithms of the routing protocol operate, and how these can be exploited by those with ill intent, countermeasures can be developed, readying MANETs for wider deployment and use.This memorandumtakes an abstract look at the algorithms that constitute the Optimized Link State Routing Protocol version 2 (OLSRv2), and identifies for each protocol element the possible vulnerabilities and attacks -in a certain way, provides a "cookbook" for how to best attack an operational OLSRv2 network, or for how to proceed with developing protective countermeasures against these attacks.

Improved flooding of broadcast messages using extended multipoint relaying

“…We present a series of simulations that show the validity of our approach, and that reveal that our method considerably improves the performance of previous existing proposals. Our proposal is, moreover, compatible with the forthcoming evolutions of the OLSR protocol (referred as OLSRv2 in the literature), by addressing the sketched extended quality criteria, such as the security aspects of the resulting new protocol [9,10].…”

Router and Link Admittance Control in the Optimized Link State Routing Protocol Version 2 (OLSRv2)

“…routers not able to correctly sign messages), but malicious routers can still record and replay messages (see [8] for details on such "replay attacks"). These replay attacks can be partly avoided by introducing "freshness" information, such as timestamps or nonces, in messages.…”

“…that the MPR flooding process operates correctly. [8] provides a detailed security analysis of OLSRv2, observing how, and with which consequences, a disruptive attack might be conducted against an OLSRv2 network. A common, and not surprising, observation from [8] is, that identity spoofing and link spoofing, i.e., that a router in its control traffic either pretends to have the identity of another router or pretends to have (non-existing) links to another router, are major vectors for disruptive attacks on an OLSRv2 network.…”

Vulnerability Analysis of the Simple Multicast Forwarding (SMF) Protocol for Mobile Ad Hoc Networks

“…In a neighbor discovery based mechanism, a malicious router can intentionally and frequently alter the information declared -including the RSSV -so as to cause generation of inordinate amounts of control traffic by legitimate routers. Such indirect jamming is discussed in [15] for neighborhood discovery and link state advertisement, and also applies for the RSSV, which a malicious router may alter and signal frequently, causing its neighborhood to launch (possibly computationally intensive) RSS recalculations and signal selected relay set (causing increased channel occupation) as well as the change of its neighbor's status.…”