University of Colorado
As new and more complex technologies become available to first responders, there are many open problems surrounding usability, reliability, and security of these networks that are particularly unique to the first responder context. Public safety systems will increasingly rely on distributed applications as well as voice communication, and our goal is to study the problems this new architecture brings, and develop and research practical solutions. In many cases, we believe this development can be built using combinations of existing open-source software, enabling an open ecosystem of applications to be created at reduced costs and greater flexibility. - July 2019
Principle Investigator: Dirk Grunwald, University of Colorado
Eric Wustrow, University of Colorado
Sangtae Ha, University of Colorado
Daniel Frank Massey, University of Colorado
New technologies such as LTE that become available to first responders will enable new applications such as mapping, localization, or even video streaming. However, ensuring that these applications remain resilient and robust on a complex new network architecture remains a challenge. For example, when groups of responders wander out of communication range, or an incident requires inter-agency coordination without a single centralized control structure in place, applications must be able to quickly and flexibly adapt. We plan to build on distributed system concepts (and existing systems) with the goal of enabling these new and advanced applications to operate seamlessly in temporarily disconnected or decentralized environments.
To accomplish this, we will modify and extend available software systems for group membership, ad hoc routing, multi-path routing and distributed data structures to simplify the construction of distributed applications that are then able to work in disconnected configurations. Key to this effort is a focus on usable security systems that will allow disparate groups to work together effectively in a secure networking environment.
We intend to provide a programming model and library to develop public safety critical applications to provide robust access to stored content (maps, photos, etc), replicate new content as it is acquired, maintain network connections over a plurality of links based on estimation of available link quality and also allow the construction of applications using a transactional distributed-object system. Distributed transactional systems simplify software development, but they are sensitive to the kinds of network partitions that may arise in the public safety context.
A core focus of the system design will be to insure that public safety applications, which involve time-varying membership, can be implemented easily while avoiding some of the security mis-designs common to earlier communication systems.
Our goal is to simplify the production of complex mobile applications; a key challenge will be designing a system that exploits the benefit of 3GPP networks as well as standard internet software design and that can handle disconnected operation and coordinate trust in inter-agency operations at a particular incident.