Best-effort Data Leakage Prevention in inter-organizational tactical MANETs

TLDR

Reconfigurable Radio Systems built on SDR and MANET provide military network survivability and interoperability, yet protecting these RRS‑based networks from attacks and data leaks—especially when tactical data crosses organizational boundaries—remains a key research challenge. This work proposes a best‑effort Data Leakage Prevention approach for inter‑organizational RRS networks. The architecture employs data‑mining and an efficient n‑dimensional clustering algorithm, previously validated for real‑time anomaly detection, and is implemented as an extension of the GTRS system, modeled and simulated in OPNET™ Modeler. Experiments demonstrate that the scheme can efficiently detect common data leaks while maintaining a very low false‑positive rate.

Abstract

Reconfigurable Radio Systems (RRS), based on Software Defined Radio (SDR) and Mobile Ad-hoc Network (MANET) technologies, offer considerable advantages for military operations, such as increased network survivability and interoperability. The RRS-based Common Tactical Radio System (GTRS), currently in development by the Swedish Armed Forces, is designed for use in diverse geographical settings and for purposes varying from international combat missions to national contingency operations. However, protecting these networks from attacks and safeguarding the carried information against leaks is an ongoing research challenge, especially in combined scenarios where tactical data may flow across organizational boundaries. This paper presents a best-effort approach to Data Leakage Prevention (DLP) for inter-organizational RRS-based networks. The proposed architecture makes use of data mining techniques and an efficient n-dimensional clustering algorithm which has previously been successfully used for real-time anomaly detection in critical infrastructure protection. The DLP architecture is developed as an extension to the GTRS system, modeled and simulated in OPNET™ Modeler. Our results show that common data leaks can be efficiently identified by the proposed scheme, while keeping the important false positive rate at a very low level.