CNS: CNS Core (CNS Core)

The CNS Core program deals with all aspects of computer and network systems. Society’s reliance on such systems has grown dramatically in the last decade. At the same time, both the resources from which those systems are built—computing, storage, communication networks, and software—and the way those resources are organized and distributed have continued to evolve rapidly. Therefore, the science of understanding and designing networked computing systems is of critical importance.

Current and future systems need to satisfy various common and purpose-driven requirements. Common system requirements include security, reliability, manageability, usability, and sustainability, as well as cost-effectiveness and fitness for purpose. Depending on the context, other requirements may include performance, privacy preservation, scalability, responsiveness, and survivability.

The CNS Core program supports innovative research that considers technology trends and emerging challenges, while emphasizing a system focus and awareness of the types of requirements mentioned above. The CNS Core program recognizes the interdependency and blurring of boundaries among computing, storage, and networking (sub)systems and the research associated with them. As such, specific sub-programs are not called out. It is not the intent of the CNS Core program to reduce the scope of the research areas covered by the division. Rather, the intent is to encourage cross-fertilization among areas of CNS research.

Research of interest for this program:

• Explores fundamental principles and creates innovative technologies, protocols, and systems that define the future or—more realistically—harness current and emerging technologies, trends, and applications;

• Produces practical abstractions, techniques, tools, artifacts, or datasets that address/enhance both general and functional requirements such as those outlined above;

• Reflects a clear understanding of what each component does and how it interfaces with the rest of the system and the environment.

Although purely intellectual investigations are within scope, research that is anchored in current and future systems for societal needs is encouraged.

A hallmark of systems research is the investigation and understanding of design tradeoffs that must be navigated when designing and implementing systems against the requirements above. Proposals that expose underlying principles or tradeoffs having predictive value which extends across different domains are especially encouraged. Proposers should identify and describe the context of the proposed system(s), the objectives or capabilities envisioned, and their expected contribution to advance towards the goals. Three especially important example requirements are:

• Secure by design: How can one ensure integrity and confidentiality of networked systems and data? How can one enhance abstractions, delineate permissible actions, enforce compliance, and establish security defaults in design processes that anticipate vulnerabilities and provide defense against unforeseen attacks from adversaries?

• Robustness: How can systems—existing and future—be made more adaptable and resilient to natural and anthropogenic hazards (e.g., weather events, malware, sabotage), as well as other normal or expected events, such as component failures, misconfigurations, and overloads? What innovative approaches would enable one to ensure system robustness and to identify, communicate, and mitigate system anomalies in real-time for outages at both small and large scale?

• Manageability: What new architectures and protocols, measurement and monitoring capabilities are needed to support a growing set of diverse applications? How can these measurements and monitoring capabilities aid in overall system management? What are novel approaches to enable comprehensive, pervasive, accurate, and usable measurement capabilities, near real-time system analytics, and systems management when the systems are massive and at the scale of the Internet? What innovations are needed to enable truly autonomous systems, which are self-managing by definition?

In general, any topic having to do with augmenting, understanding, enhancing, or transforming computing and communication systems undertaken from a systems point of view is within scope.

CNS and Closely-Related Programs: Issues that reside primarily at the device or application level and that are highly context-specific will not be considered a good fit for this program. Projects that focus exclusively or primarily on cybersecurity threats and countermeasures may be a better fit for the Secure and Trustworthy Cyberspace (SaTC) program. Projects focused primarily on design or enhancement of sensing and control systems that interact with the physical world may be a better fit for the Cyber-Physical Systems (CPS) program.