Source-linked AI summary
The Internet of Battle Things
Alexander Kott, Ananthram Swami, Bruce J West
TL;DR
IoBT's unprecedented scale, heterogeneity, information complexity, and adversarial environment create major challenges for effective military use. The paper identifies research needs including adaptive communications, information-brokering methods, and approaches to counter deception.
Problem
IoBT will combine unprecedented scale and heterogeneity with severe information-processing and adversarial challenges that existing approaches may not adequately address.
Method
The paper examines IoBT's communications, information-use, and adversarial challenges and outlines research directions for adaptive networking and mission-relevant information management.
Results
The paper identifies major IoBT research needs, including scalable adaptive communications, formal representations of information needs, and methods for handling non-ergodic information processes and deception.
Takeaways & Limitations
Effective IoBT operation will require technologies that dynamically manage heterogeneous networks and deliver concise, relevant information while accounting for deception and adversarial attacks.
Takeaways & Limitations
Learning normal patterns and detecting anomalies can fail against well-designed deception, making learning a potentially dangerous double-edged sword.
Abstract
from arXiv · showhide
The battlefield of the future will be densely populated by a variety of entities ("things") -- some intelligent and some only marginally so -- performing a broad range of tasks: sensing, communicating, acting, and collaborating with each other and human warfighters. We call this the Internet of Battle Things, IoBT. In some ways, IoBT is already becoming a reality, but 20-30 years from now it is likely to become a dominant presence in warfare. To become a reality, however, this bold vision will have to overcome a number of major challenges. As one example of such a challenge, the communications among things will have to be flexible and adaptive to rapidly changing situations and military missions. In this paper, we explore this and several other major challenges of IoBT, and outline key research directions and approaches towards solving these challenges.
Managing and Adapting the IoBT
IoBT’s exceptional scale, heterogeneity, and dynamism require new ways to discover, model, communicate across, and adapt the network. That same scale may also improve reliability by making many sensing resources available.
- Managing and Adapting the IoBT: IoBT may reach about a million things per square kilometer, requiring new theories, models, and technical approaches beyond current practice.The challenge is especially acute when military forces use civilian IoT infrastructure and devices they do not own.
- Managing and Adapting the IoBT: Its massive scale can improve availability because densely positioned sensors reduce concerns about any single sensor being unavailable.The paper calls for theoretical results on how very large ensembles of things and data produce determinism.
- Managing and Adapting the IoBT: Extreme heterogeneity and absent commercial standards require rapid adaptation across diverse things, protocols, and communication technologies from multiple manufacturers.Future military systems may combine local commercial IoT with commercially sourced warfighter equipment.
- Managing and Adapting the IoBT: Military forces will need dynamic discovery, characterization, tracking, and automatic updating of device behaviors and performance during operations.Human behaviors and intents—including those of warfighters, adversaries, and civilians—must also be detected, characterized, and projected.
- Managing and Adapting the IoBT: Communication management will require intelligent automation to allocate and reconfigure resources, dynamically modify information-sharing policies, and support scalable architectures and protocols.The network must accommodate heterogeneous, competing, unpredictable things and diverse timing constraints.
- Managing and Adapting the IoBT: Rapid, automatic situational awareness should infer sufficiently complete system information from measurements of relatively few variables.The paper also identifies new IoBT uses, including possible support for position, navigation, and timing needs.
Making IoBT Information Useful
Making IoBT information useful requires reducing unprecedented volume and complexity while preserving mission-relevant meaning, trustworthiness, and timeliness. Because adversaries can attack both information and cognition, the paper calls for models, fusion, anomaly detection, and counter-deception approaches.
- Making IoBT Information Useful: Human cognition bandwidth is the most severe constraint: warfighters need concise, relevant information rather than IoBT’s enormous data flows.Inappropriate information can mislead humans or intelligent things into worse actions than they would take without it.
- Making IoBT Information Useful: IoBT information will have unprecedented volume and complexity, with large differences in abstraction, trustworthiness, and value across sources.The paper argues that existing information-theoretic foundations may need reconsideration for nonlinear, non-ergodic processes and interactions between dissimilar networks.
- Making IoBT Information Useful: A proposed target is compressing and fusing IoBT data by a factor of 10E15 before delivery to humans and intelligent things.Layered information brokers or “concierges” are proposed to aggregate information into manageable, meaningful content, though they may restrict access to underlying data.
- Making IoBT Information Useful: Information brokers need machine-interpretable representations of mission information needs, supplemented by self-learning that integrates machine learning with semantic knowledge techniques.Mission planning and rehearsal could help identify required and likely available information.
- Making IoBT Information Useful: Executable models, large-scale simulation, distributed self-modeling, self-calibration, and self-validation are proposed for validating, interpreting, fusing, and assessing information trustworthiness.The paper describes automatic creation and maintenance of such models as an early-stage research area.
- Making IoBT Information Useful: IoBT’s adversarial environment threatens its physical infrastructure, information confidentiality, integrity, availability, and human cognition.Attacks include kinetic, directed-energy, electronic, jamming, eavesdropping, malware, and deception operations.
- Making IoBT Information Useful: Very large, heterogeneous IoBT deployments may support additional information-protection measures, including discarding devices suspected of compromise.The paper identifies minimizing enemy opportunities to acquire information as a top priority.
- Making IoBT Information Useful: Anomaly detection and learned normal patterns are insufficient against well-designed deception, motivating dedicated counter-deception research.The paper suggests IoBT’s scale and heterogeneity may help because maintaining consistent deception across many diverse sources is difficult.
Disclaimer
The article disclaims that it represents the positions or views of the authors’ employers.
- Disclaimer: The article does not reflect the positions or views of the authors’ employers.