Service provisioning over wireless networks in highly volatile environments
Abstract
Network connectivity among the different tactical units in the battlefield is a key capability for operational success. Yet it is quite difficult to ensure in the highly volatile and adversarial battlefield landscape. At the same time network technology is constantly evolving. On one hand, in the tactical networks, traffic demands are constantly increasing among different interoperating systems to support coalition operations and information-centric applications. On the other hand, tactical networks involve heterogeneous mobile devices, including unmanned vehicles and Internet of Things (IoT) sensors to execute various tasks, as well as heterogeneous network technologies such as WiFi, LTE and upcoming mmWave radio. New challenges are raised for efficient network management frameworks to finish configuration for the complex setup, and security measures towards the increased attack surface. Softwarization and virtualization are promising solutions to address these challenges. Software Defined Radio (SDR) is a programmable system enabling fast reconfiguration of the wireless physical layer to support diverse protocols. Software Defined Networking (SDN) introduces a centralized and programmable control plane to facilitate optimizing the network configuration by making decisions with a global view. Enabled by SDN, Network Function Virtualization (NFV) creates different virtual networks and services over the same infrastructure, allocating resources in a flexible way to adapt to the dynamic network environment. The purpose of this proposal is to build a testbed developing and evaluating SDN, SDR and NFV solutions for tactical networks. We plan to deploy a full-scale softwarized and programmable network architecture covering three levels, the cloud infrastructure, radio access networks and edge mobile nodes. The testbed will incorporate several 5G features, including mmWave radio, virtual and augmented reality (VR and AR) devices and unmanned vehicles. We also concern about the computation capability to adopt different machine learning methods in related tasks. The proposed testbed will enable a number of ongoing and new research thrusts. First, we will investigate Software Defined Coalition (SDC) architectures as an extension of current SDN approaches addressing the agile configuration and resilience requirement of the tactical networks under potential risks such as network fragmentation. Second, we will focus on distributed and decentralized machine learning algorithms that enable data analytics across multiple drones and sensors, taking the robustness and security into consideration. Last but not the least, we will elaborate on the notion of Utility of Information (UoI) to capture the impact of network latency in tactical tasks and develop policies filtering the enormous volume of information between multiple pairs of data sources/receivers based on the value of information. The testbed equipment and different research thrusts proposed have a high military relevance, involving other ongoing research projects such as DAIS ITA funded by ARL and MURI funded by ONR. Besides, this project will lead to educational benefits. Students and young researches will be trained with various cutting-edge network technologies across computer science and electrical engineering.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 25, 2021
- Source ID
- W911NF2110256
Entities
People
- Leandros Tassiulas
Organizations
- Army Contracting Command
- United States Army
- Yale University