Today, Intel announced that it has joined a National Science Foundation (NSF)-led, multi-sector partnership to form RINGS -- the Resilient and Intelligent Next-Generation Systems program -- a new initiative that seeks to accelerate research in areas with potentially significant impact on Next-Generation networking and computing systems. Key federal agencies such as the Department of Defense (DOD) and the National Institute of Standards and Technology (NIST) as well as several other leading technology partners are also part of the RINGS initiative.
This program seeks to accelerate research in areas that can significantly impact emerging next-generation (NextG, or “5G and beyond”) wireless and mobile communication, networking, sensing, and computing systems, along with global-scale services. The primary focus is on improving the resiliency of networked systems as well as boosting network performance metrics such as throughput, latency, and connection density.
Proposed projects must develop and leverage advanced radio technologies, spectrum allocation, edge/cloud computing, cognitive networks, microservices, and artificial intelligence (AI)/machine learning (ML). Proposals are due July 15, 2021.
“Intel and NSF have a long-term commitment to driving network transformation. A big part of that is increasing network and system resiliency through automation, adaptability, and security. The RINGS program hopes to accelerate research so wireless networks can better support critical and essential services for 5G and beyond.”
Why It’s Important
Communication networks are a critical part of societal infrastructure—just like power, water, and transportation. But today’s communication networks are not resilient or performant enough to support emerging NextG critical and essential services, which range from communication to computation to intelligent decision making. Traditional network design has been narrowly focused. In contrast, the networks of the future require a system-level mindset that addresses all layers of the protocol stack.
NextG networks and systems (which include the next evolutions of cellular, Wi-Fi, and satellite networks) must meet high expectations:
- Connect billions of heterogeneous Internet of Things (IoT) devices along with billions of people;
- Enable machine-to-machine (M2M) communications; and
- Provide low-latency computational and storage resources on-demand at the edge and in the cloud.
Using a combination of intelligence, autonomy, and a microservices architecture, NextG networks and systems can be used to support a variety of critical and personalized services. These services include data processing, dissemination, and storage; real-time computing and learning; and large-scale content distribution. NextG networks and systems will often span a distributed user-to-edge-to-cloud continuum.
These networks and systems can be deployed in multiple application domains, including education, defense, critical infrastructure, transportation, healthcare, public safety, logistics, smart agriculture, finance, and entertainment. As the economy becomes ever more dependent on the high availability, security, and reliability of such network systems, any failure, tampering, or degradation in network service can be highly disruptive, or even catastrophic.
Therefore, it is essential that these NextG network systems be highly resilient, with strong expectations for both performance and service assurance, despite any combination of natural or human-induced disruptions. All these features should be able to scale as network and service complexity grows.
Resiliency Isn’t Easy to Achieve
Providing resiliency guarantees while ensuring cutting-edge performance is an under-explored topic in networking and systems research. The challenges are well-recognized: geographical spread, distributed attack surfaces, multi-modal failures, and unforeseen dependencies across subsystems leading to cascading failures. Here are some examples of potential failure points:
- Base system components, such as towers, antennas, fiber-optic cables, computing, storage devices, and software systems and services are susceptible to failure during extreme weather events.
- NextG networks and systems are expected to be complex, with many different components that have limited and diverse security capabilities. Some elements of the system may be highly vulnerable to failures and attacks.
- As software-defined networking (SDN) replaces many hardware-based functions, additional failure modes might be introduced.
- Increasing reliance on ML/AI techniques within the network may expose additional attack vulnerabilities.
- Relying on there being a human available to help recover quickly from disruptions might be impractical, given the scale of network systems and the impact of disruptive events.
The bottom line is that providing resilient communication and computation services across a heterogenous and dynamic networking system is a significant challenge.
What Resiliency Really Means
Resilience must apply across the entire ecosystem of network, service, and computational resources, characterized by the following:
- Resistance to or high tolerance of attacks, failures, and service disruptions, with rapid root cause identification
- Graceful degradation of services and rapid adaptability when resource availability is impacted by disruptive events
- Distributed computational capabilities spread across heterogeneous and disaggregated resources
Resiliency involves three foundational attributes, which must be provided at scale while ensuring cutting-edge performance. These attributes are the primary research vectors (Group A RVs) that the RINGS program focuses on:
- Full-stack security. A secure-by-design approach enables network designers and architects to eliminate entire categories of threats and address security requirements at the earliest stages of the design process.
- Intelligence and adaptability. Architectures, protocols, and network-system management solutions provide a quick response to performance issues and threat.
- Autonomy. The network must operate at a highly functional level without human intervention even during disruptive events.
The RINGS program also enumerates several enabling technologies (Group B RVs) that can enhance the performance of the network system in terms of throughput, latency, connection density, application support, and service composability. Research proposals must include at least one of these Group B RVs, in addition to one or more of the Group A RVs:
- RF and Mixed Signal Circuits, Antennas and Components
- Novel spectrum management technologies
- Scalable device-to-edge-to-cloud continuum
- Merging digital/physical/virtual worlds
Intel believes the RINGS program will extend and accelerate the network transformation that began with SDN and network functions virtualization (NFV) and continues with advances in programmable accelerators and AI/ML solutions. As network architectures shift to cloud-computing platforms, dynamic orchestration, and multi-access edge computing (MEC)—all deployed on industry-standard servers—more research is necessary to increase the resiliency of the network and optimize network performance to prepare for 5G and beyond.
The RINGS program anticipates funding 36–48 awards, up to three years in duration. In addition to Intel, other technology enterprises supporting the RINGS program include Apple, Ericsson, Google, IBM, Microsoft, Nokia, Qualcomm, and VMware. For more information, including full eligibility details and proposal preparation and submission instructions, read the full NSF press release.