Robust Decentralized Signal Processing and Distributed Control of Autonomous Sensor Networks

Ryan Goldhahn | 17-ERD-101


Intelligent autonomous sensor networks, often comprised of large numbers of sensors, must be capable of jointly exploiting data collected at each agent in the network and using that data to optimize future actions towards multiple mission objectives. Centralized signal processing and optimization solutions process all data and determine all future actions at a single agent, and the resulting information and commands are disseminated back to the network. The communications bandwidth required and the single point of failure the central agent represents often make these solutions untenable for national security applications.

We developed several fundamental algorithms for providing both decentralized signal processing and network optimization as well as simulation software to validate the results of these algorithms at scale. Specifically, novel algorithms for Bayesian decentralized estimation and decentralized detection and optimization based on the alternating direction method of multipliers were developed for autonomous sensor networks and published in the literature. The first large scale simulation of autonomous sensor networks (1,000 agents) was conducted on this project, validating the performance of the developed algorithms. The algorithms and simulation tools are critical components of any decentralized autonomous network and have current and future national security applications, including distributed sensor networks for detection, estimation, and tracking problems as well as large decentralized cyber-physical infrastructure such as the power grid.

Impact on Mission

This work advances Lawrence Livermore National Laboratory's core competencies in cybersecurity and in high-performance computing, simulation and data sciences. The research represents a new research direction for the Laboratory with direct application to multiple program areas at the Department of Energy, National Nuclear Security Administration, Department of Homeland Security, and the Department of Defense for decentralized power-grid management, detection/estimation, and intelligence, surveillance and reconnaissance applications. This is an important capability for future missions in intelligent autonomous sensor networks, and to test, calibrate and benchmark their hardware implementations.

Publications, Presentations, Etc.

Ho, N., et al. 2018. "Collaborative Autonomy: Evaluating Feature Extraction on the Nvidia Jetson." Lawrence Livermore National Laboratory Poster Showcase, Livermore, CA, 2018. LLNL-POST-755801.

Hogan, T., et al. 2019. "Universal Decision-Based Black-Box Perturbations: Breaking Security-Through-Obscurity Defenses." KDD 2019 Workshop on Adversarial Learning Methods for Machine Learning and Data Mining, 2019. LLNL-CONF-761205.

Kailkhura, B., et al. 2017. "Byzantine-Resilient Detection Using Collaborative Autonomous Swarms." IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), Curacao, Dutch Antilles, December 2017. LLNL-POST-742699.

––– . "Byzantine-Resilient Locally Optimum Detection Using Collaborative Autonomous Networks." 2017 IEEE 7th International Workshop on Computational Advances in Multi-Sensor Adaptive Processing (CAMSAP), pp. 1-5. IEEE, 2017. LLNL-CONF-731964.

Schmidt, K., et al. 2018. "Optimal Positioning of Mobile Sensors Using Mutual Information." Conference on Data Analysis (CODA). LLNL-POST-746706.

––– . 2019. "Sequential Optimal Positioning of Mobile Sensors Using Mutual Information." Statistical Analysis and Data Mining: The ASA Data Science Journal , July 2019. LLNL-JRNL-753008.

Wapman, J., et al. 2018. "Chemical Plume Detection with Collaborative Autonomous Sensor Networks." Lawrence Livermore National Laboratory Signal and Image Sciences (CASIS) Workshop, 2018. LLNL-POST-749008.

Yen, A., et al. 2018. "Large-Scale Parallel Simulations of Distributed Detection Algorithms for Collaborative Autonomous Networks." International Sensor Society for Optics and Photonics Disruptive Technologies in Information Sciences, Orlando, FL, April 2018. 10652: 106520G. LLNL-CONF-749406.