Wireless community networks are mesh networks built by volunteers which own, configure, and manage their wireless node. Such networks are prone to either unintentional (e.g. misconfiguration) or intentional node misbehavior. This paper proposes a fully distributed trust-based routing framework, tightly integrated with OLSR, which is the most exploited routing protocol in real world wireless community networks. The framework, designed to be modular for easy upgrade, relies on active probes, hidden in the normal data traffic through adaptation of steganography techniques. The combination of path-wide measurements into a distributed trust-framework, preliminary based upon the well known EigenTrust mechanism, permit to infer whether, and which, packet-droppers (i.e. nodes misbehaving at the data plane) affect the network forwarding operation. The resulting per-node trust values are then transformed into suitable "weights" provides as input to the OLSR protocol for mitigation through re-routing. A simulation-based performance evaluation shows that the proposed framework appears already effective in detecting and circumventing packet-droppers, despite the relative simplicity of the preliminarily considered algorithms.
Proto, F., Detti, A., Pisa, C., Bianchi, G. (2011). A Framework for Packet-Droppers Mitigation in OLSR Wireless Community Networks. In IEEE International Conference on Communications (ICC), 2011 (pp.1-6). IEEE [10.1109/icc.2011.5963001].
A Framework for Packet-Droppers Mitigation in OLSR Wireless Community Networks
DETTI, ANDREA;BIANCHI, GIUSEPPE
2011-01-01
Abstract
Wireless community networks are mesh networks built by volunteers which own, configure, and manage their wireless node. Such networks are prone to either unintentional (e.g. misconfiguration) or intentional node misbehavior. This paper proposes a fully distributed trust-based routing framework, tightly integrated with OLSR, which is the most exploited routing protocol in real world wireless community networks. The framework, designed to be modular for easy upgrade, relies on active probes, hidden in the normal data traffic through adaptation of steganography techniques. The combination of path-wide measurements into a distributed trust-framework, preliminary based upon the well known EigenTrust mechanism, permit to infer whether, and which, packet-droppers (i.e. nodes misbehaving at the data plane) affect the network forwarding operation. The resulting per-node trust values are then transformed into suitable "weights" provides as input to the OLSR protocol for mitigation through re-routing. A simulation-based performance evaluation shows that the proposed framework appears already effective in detecting and circumventing packet-droppers, despite the relative simplicity of the preliminarily considered algorithms.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.