Monitoring Infrastructure Affordances

Abstract: Computing has made its way into most of our lives as a key processor of vast quantities of information. This has happened directly in terms of gadgets and devices that assists us in everyday life, but also indirectly, through the critical infrastructures that enables these devices to function. A key issue with critical infrastructures such as transportation, communication, power-grids and finance, is increasingly circular interdependencies. Because of this issue, a disruption in either one can cascade and have a global effect on the others. To manage these complexities, we are depending on a number of monitoring systems that allow operators and other stakeholders to, within their respective expert domains, discover disruptions as early as possible and then take appropriate actions. These monitoring systems are not without challenges of their own. In addition to having evolved organically alongside their respective infrastructures, there is a considerable legacy to account for, with both hardware and software components spanning decades of computing history. This puts heavy restrictions on the kinds of interventions that can be performed safely, implying that these systems are ill fit for handling the software and software security landscapes of today, where updates and adjustments need to be applied on a daily basis in order to stand a fighting chance. The work presented herein address some of the major challenges in securing these monitoring systems against current and future threats posed by antagonistic actors, dormant software defects and changes imposed by technological advances and academic discoveries. This is approached on several fronts in parallel: by embedding resilience in order to allow for controlled experimentation and evaluation of new protection mechanisms in incrementally sensitive settings; by developing laboratory facilities for resilient smart power-grids; and by developing tools and training scenarios for operators of adaptive and reconfigurable monitoring systems.