Communication and Signal Processing in Adverse Scenarios (COMPASS)
The current proliferation of wireless devices, often with sensing, computation and actuation capabilities, making use of either ad-hoc or infrastructure-based (cellular, WLAN, satellite) communication networks, with ever-increasing demands on data transfer speed and reliability, makes it necessary to enhance the robustness of such systems in stressful adverse scenarios including, on one hand, wireless “incidental” interference resulting from the normal operation of both the network users (intra-system interference) as well as of other networks (inter-system interference); and, on the other hand, “intentional” malfunctioning of communication and/or sensing system components in situations where an attacking entity purposefully tries to disrupt network operation, or where legitimate system users attempt to exploit vulnerabilities in order to secure an unfair share of system resources. This project addresses several aspects of these complementary problems involving efficiency and security along a number of research lines, focusing on the sensing, physical communication, and medium access layers:
- Development of novel intra-system interference management techniques and the corresponding performance analyses for a number of settings including multiple access and broadcast scenarios, as well as full-duplex transceivers.
- Development of a hardware demonstrator for full-duplex links using software defined radio platforms.
- Investigation of advanced link adaptation schemes for adverse static and dynamic scenarios, including aspects such as how to dynamically estimate link performance, make efficient use of outdated or unreliable link state information, and how to optimize system performance in the long term.
- Focus on Dynamic Spectrum Access (DSA), also known as “Cognitive Radio”, as the paradigm of choice for efficient spectral usage and inter-system interference avoidance. In this framework, the key enabler for ensuring protection to other networks is Spectrum Sensing. Novel Nyquist-rate and compressed sensing techniques will be investigated in narrowband/wideband, single-/multi-antenna settings.
- Modeling the behavior of estimation, detection and learning algorithms based on distributed observations of the environment in the presence of acquisition and/or communication failures and/or intentional attacks, for both centralized (fusion center-based) and decentralized architectures.
- Development of robust schemes and countermeasures to combat adversarial threats in such settings, using statistical and game theoretic tools, and with special focus on standalone and cooperative Spectrum Sensing methods for DSA systems subject to primary user emulation and sensing data falsification attacks, as well as distributed sensor networks under data-injection adversarial attacks.
GPSC Principal Investigator
START / END DATE
2014-2016
Partners
- University of Vigo (Project coordinator)
- Technical University of Catalonia (UPC)
Funding type
Public Funding
Funded by
“Programa Estatal de Investigación, Desarrollo e Innovación orientada a los Retos de la Sociedad” Ministerio de Economía y Competitividad and European Regional Development Fund. Ref. TEC2013-47020-C2-1-R
