Computer Networks In Tactical And Disaster Recovery Environments

abstracteng

Tactical and Disaster Recovery Operations (TDR) are often conducted in degraded environments and supported by ad hoc mobile communication mechanisms based on radio and satellite technologies. These networks are much less stable and reliable than what's commonly behind enterprise and commercial internet applications. These custom networks are required because TDR operations are conducted in remote and unforgiving locations and must continue regardless of meteorological conditions. Furthermore, TDR operations are very dynamic, with nodes entering and exiting radio ranges further complicating the communication environment. These aspects create a communication environment afflicted by high latencies, low bandwidth, and non-negligible packet loss that violates several assumptions behind message dissemination mechanisms commonly used in civilian applications. Even so, the critical nature of TDR operations requires fast, predictable, and reliable communication mechanisms capable of exchanging messages. This fact calls for a case-by-case evaluation of available solutions to reckon if they can operate effectively or if it's necessary to implement custom solutions tailored for this specific use case. Another problematic characteristic of Networks used in TDR operations is that they present variable and unpredictable performance. This characteristic calls for the development of network-aware solutions capable of continuously monitoring the network, evaluating its status, and reconfiguring parameters to adapt them to new conditions and improve the use of network resources. This thesis presents research on three major topics. The first two analyze the evolution of communication protocols for unicast and group communication. In particular, this thesis presents research on these systems to evaluate their applicability to degraded environments. The chapter on group communication also presents the ANGLOVA scenario, a sophisticated emulated network environment created in collaboration with the NATO Science and Technology Organization which can be used to perform tests in a realistic environment. The two chapters also present several experiments in which many modern protocols developed for tactical and consumer applications are evaluated and compared. Finally, this thesis presents the Smart Estimation of Network StatE Information (SENSEI) framework, a distributed system for passive monitoring and adaptation capable of harvesting network information, inferring the network state, and adapting the behavior of other components to improve the use of network resources. SENSEI implements several algorithms to control the information aggregation and distribution processes to minimize network footprint and increase the value of what is transmitted. The research presented in this thesis was carried out in collaboration with several international groups and research institutes such as the Florida Institute for Human and Machine Cognition (IHMC), FL, USA, the NATO Science and Technology Organization (STO) IST-124 Research Task Group on "Heterogeneous Tactical Networks - Improving Connectivity and Network Efficiency", and the United States Army Research Laboratory (ARL), Adelphi, MD, USA.