Millimeter-wave Networking in Transient Topologies
Abstract
The millimeter-wave spectrum has emerged as a new frontier in designing the next generation of high-speed wireless networks. With the availability of a large unlicensed spectrum at 60 GHz, WLANs and cellular networks can provide multi-gigabit per second data rates. Such gigabit wireless links can enable a range of military applications such as real-time data sharing and video streaming between the autonomous ground and aerial vehicles, facilitating platooning and coordinated emergy and combat response. However, millimeter-wave wireless links are susceptible to blockages and mobility due to their directionality and high frequency. This requires continuous beam searching and steering to maintain connectivity. These challenges are especially aggravated in military scenarios where the underlying topology is transient (for example, vehicle to vehicle mmWave communication in a platoon moving at a high speed). Conventional beamforming protocols are ineffective for such scenarios of mmWave V2V communication where contact and line-of-sight (LoS) durations are very short and connectivity is frequently hampered by transient blockages. The objective of this project is to systematically understand the impact of transience on mmWave links and to design beam adaptation and management techniques custom-tailored for transient topologies in military settings. The exploration will be divided into three tasks: (1) The first task will be to analytically study beam coherence time and LoS probability of mmWave links with respect to vehicular mobility, density, and presence of transient blockages. The study will be followed by an investigation on how transient reflections from other vehicles can be predicted and exploited in real-time with limited beamsearching. (2) While the vehicular reflections are helpful in low to moderate density and blockage, the second task will explore the use of ambient reflectors to maintain the link connectivity. Given that existing methods of beamsearching are not sufficient for short-lived, transient connection durations, this task will include exploring the camera imagery and use of machine learning techniques to identify, locate and leverage the reflectors for maintaining mmWave links. (3) The existing methods of interference estimation and medium access in presence of beamsearching are ineffective for mmWave V2V networks. To address this problem, interference estimation techniques with low measurement overhead will be designed. These techniques will be integrated into two types of medium access protocols. In the first protocol, the link interference information will be shared over a low-frequency backbone, while in the second protocol, a completely distributed, random access with spatial backoffs on different mmWave paths will be investigated, requiring no medium access coordination between vehicles. The three tasks will create a comprehensive solution space that can realize the true potential of mmWave wireless in transient network topologies. The proposed research will be evaluated through experimentation on mmWave testbeds at George Mason University and the University of California, Davis. The experimental evaluation will include a real-world deployment of 60 GHz mmWave commodity and software radio devices. A simulation framework based on Remcom ray-tracing propagation modeling, Sumo vehicle mobility modeling, and NS3 packet-level network simulation will be used to create large-scale evaluations of the proposed techniques and protocols.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 11, 2022
- Source ID
- W911NF2210216
Entities
People
- Prasant Mohapatra
Organizations
- Army Contracting Command
- United States Army
- University of California, Davis