Towards End-to-End QoE Guarantees for Timely Video Delivery in Adversarial Multi-hop Battlefield Networks
Abstract
The research objective is to establish a new framework for developing network algorithms that are for live surveillance video streaming in multi-hop battlefield networks. These algorithms need to be both provably optimal in terms of the achievable end-to-end quality of experience (QoE), and implementable under software-defined network (SDN). Our proposed framework will be built upon two critical components: One is an analytical model that determines important QoE metrics, such as the frequency and duration of video interruptions, based on the network traces of packet deliveries. The other is a network model that addresses important features of battlefield networks, including unreliable wireless transmissions, end-to-end delay bounds, and the presence of event-driven IoT applications. Our research plan can be summarized in the following four thrusts: Thrust 1. QoE optimization for one-hop wireless networks: This thrust focuses on a one-hop wireless network where a wireless gateway serves live video streams for multiple wireless users. It will define new QoE metrics that realistically capture multiple important factors that impact user experience. It will also develop policies that optimize these new QoE metrics. Thrust 2. End-to-end QoE for multi-hop networks: This thrust addresses optimizing end-to-end QoE for multi-hop networks subject to end-to-end delay requirements. It will specifically focus on designing implementable policies that are tractable and distributed through a variety of decomposition techniques. Thrust 3. Coexistence with event-driven IoT applications under adversarial environments: This thrust considers the practical challenge that live surveillance video flows need to coexist with other mission-critical applications, such as the emerging IoT applications. It will develop online algorithms that are provably robust for both video flows and IoT applications even under an adversarial sequence of events. Thrust 4. Implementation and integration with software-defined networking: This thrust studies the implementation issues of policies produced from this project. It will focus on two important topics: First, the design of new policies that are consistent with SDN principles Second, implementing and testing the policies on an actual wireless testbed. The success of this project will significantly impact Army capabilities. By achieving stringent end-to-end QoE guarantees for live surveillance video flows without jeopardizing the robustness of other mission-critical applications, this project will enhance the situational awareness, and, as a result, greatly improve the effectiveness and safety of soldiers at the tactical edge.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 14, 2019
- Source ID
- W911NF1810331
Entities
People
- I-Hong Hou
Organizations
- Army Contracting Command
- Texas Engineering Experiment Station
- United States Army