Exploiting Spatio-temporal Locality Information for Optimal Physical-layer Security
Abstract
This STIR proposal aims at investigating the design of a system that optimizes the physical-level secrecy, when the geographical locations of blue and red forces are (even approximately) known. In most military deployment scenarios, in general, and army forces deployments, in particular, there is spatial and temporal locality of friendly and enemy resources, where the knowledge of such locality could be used to improve the communication secrecy. As a 2D example, there are areas where friendly forces are known to be located, areas where red forces are known to be located, and gray areas where there is certain likelihood of enemy/friendly forces. This knowledge can be exploited by the transmitter T, to improve its secrecy channel to the receiver R. In general, in practical theater operations, it would be required to integrate complex 3D locality information. We also note that due to mobility and changing radio propagation conditions, the reliance on the locality information needs to be frequently reevaluated. As discussed in this proposal, the rate of secrecy information that could be conveyed over a channel depends on the spatial locality of the blue/red forces; e.g., when the transmitter is located deeply within the blue force region, the T-R link can transmit secret information at higher rate, then when the transmitter is located closer to the red forces. Gray areas should be used according to their likelihood to contain a red receiver as to maximize the secrecy rate of blue links. The high-level operation of the system is as follows: The system assumes a Cooperative MIMO setting, where a MIMO link is formed by temporarily activating transmitting and receiving clusters, each of nearby trusted devices, and with each cluster being centrally coordinated by its corresponding cluster head. The transmitter system and the receiver system collect the spatial locality information (i.e., blue-red-gray spatial locality information); the locality information will, in general, include ground, air, and marine forces. A computation is then performed to determine, given the locality information, what is the maximum feasible secrecy rate. There are trade-offs between the processing of the algorithms as either centralized or distributed, and in this project will study the performance of trade-offs. A code is then chosen to encode the communication with this secrecy rate (i.e., to implement the secrecy rate). As the locality of the information at the transmitter and the receiver change, the computation is repeated. The re-computation is dependent on the size and shape of the localities, and on the speed of the assets in those localities. This project will study methods to optimize the performance as a function of the system parameters. In summary: This STIR project will attempt to evaluate feasibility and performance of Distributed MIMO system based on Transmitter and Receiver locality information, as to maximize the secrecy rate. The location information will be represented as a contour constraint. Evaluation and comparison will be made to optimize the system of practical-complexity systems.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 14, 2019
- Source ID
- W911NF1810406
Entities
People
- Zygmunt J. Haas
Organizations
- Army Contracting Command
- United States Army
- University of Texas at Dallas