NSF Award #1547291 “EARS: Adaptive Behavioral Responses for Dynamic Spectrum Access-Based Connected Vehicle Networks”

bumblebeeforage

The memory-based channel selection algorithm from bee (top) and vehicle (bottom) perspectives. Similar to each bumblebee, each vehicle is equipped with memory to store channel (floral) reward information, which is then used to select the channel (floral species) with the highest reward quality out of those available in sampling interval. During the transmission interval, the vehicles (bumblebees) use their current channels for communication (forage on current species) while simultaneously tracking the change in the reward level. The vehicles switch to a better channel (floral species) based on their memory if the current channel level drops to a lower value. Vehicles alternate between sampling ${(T_{Sample})}$ and transmission ${(T_{Transmit})}$ periods to track changes in a time-varying noisy resource environment. (From https://ieeexplore.ieee.org/abstract/document/8347136)

Connected vehicle technology has the ability to provide drivers with a significantly higher level of environmental awareness relative to the present day. Thus, enabling reliable, seamless, and efficient wireless access to support vehicular connectivity is core to this safety technology. This project studies an approach that combines vehicular wireless networking with foraging theory, a concept that is extensively employed to describe the behavior of bumblebees. Specifically, the research will draw parallels between vehicular networks and bumblebees foraging for nectar in order to establish a novel framework for enhancing the performance of connected vehicles. This interdisciplinary project will make an educational contribution via the mentorship and training of graduate students from both Electrical & Computer Engineering and the Biological Sciences, with an emphasis on identifying qualified students from underrepresented groups.

In order to achieve a reliable and efficient connected vehicle networking architecture, this project studies how dynamic spectrum access (DSA)-based vehicular networks can be combined with foraging theory concepts. Although wireless networking research has previously looked to the insect world for insights on real-time decision-making across multiple communication nodes within a network in order to achieve some level of distributed optimization, e.g., ant colony optimization, honeybee swarm techniques, all of these approaches significantly depend on the high level of social dependency and information exchange found in these species in order to perform these operations. Conversely, bumblebees have been characterized as socially sharing past and present information with other bumblebees, but are still capable of making independent decisions, which is very similar to nodes within a vehicular networking environment. The application of mathematical tools used to temporally weigh the information shared between vehicular networking nodes as well as predict conditions in the near-future, such as autoregressive moving average (ARMA) filters and Kalman filters, have never been employed in models used to describe bumblebee behavior. Consequently, this effort could make an impact on biological sciences by providing mathematical tools that can be employed during the information weighing process of bumblebees.

Click here for the official NSF award abstract.

Personnel:

Goals:

  • Understand how information about available wireless spectrum would be collected by individual vehicles and shared with other vehicles via a VANET architecture.
  • Determine how an individual vehicular node would act upon shared wireless spectrum information from other vehicles in terms of optimizing its access to wireless spectrum.
  • Calculate the amount of time that each vehicular node within the VANET would need to store information regarding itse transmission environment.
  • Assess the impact of the response time versus the response performance as a function of the amount of present and past environmental information stored by each individual vehicular node.
  • Determine how we optimize vehicle responses in variable channel environments based on information gained through individual and social experiences.

Activities & Outputs:

  • A. M. Wyglinski, R. J. Gegear, E. F. Ryder, K. S. Gill, K. N. Heath (2018). Understanding Vehicular Wireless Access Using Bumblebee Behavioral Models. 3rd IEEE VTS Connected & Autonomous Vehicles Summer School @ WPI. Boston, MA, USA.
  • Kuldeep Gill, Bengi Aygun, Kevin Heath, Rob Gegear, Liz Ryder, Alexander M. Wyglinski (2018). Memory Matters: Bumblebee Behavioral Models for Vehicle-to-Vehicle Communications. 6. IEEE Access, 6. 25437 – 25447. DOI. 10.1109/ACCESS.2018.2830313.
  • Kuldeep Gill, Kevin N. Heath, Robert J. Gegear, Elizabeth F. Ryder, and Alexander M. Wyglinski (2018). On The Capacity Bounds For Bumblebee-Inspired Connected Vehicle Networks Via Queuing Theory. 2018 IEEE 87th Vehicular Technology Conference. DOI. 10.1109/VTCSpring.2018.8417762.
  • Kuldeep S. Gill, Kyle McClintick, Nivetha Kanthasamy, Jefferey Tolbert, Duong Nguyen, Son Nguyen, Galahad Wernsing, Valerie Moore, Ian Gelman, Alexander O’Neil, Nicholas Schubert, Corey Coogan, Krysta Murdy, Brian Mahan, Sylvester Halama, Kevin N. Heath, Elizabeth F. Ryder, Robert J. Gegear, and Alexander M. Wyglinski (2018). Experimental Test-Bed for Bumblebee-Inspired Channel Selection in an Ad-Hoc Network. 2018 IEEE 88th Vehicular Technology Conference (VTC-Fall). DOI. 10.1109/VTCFall.2018.8690978.
  • Alexander M. Wyglinski. “Bumblebees and Vehicular Networking: Intelligent Connectivity on the Road”, Boston University CISE Seminar, October 5, 2018. URL: http://www.bu.edu/systems/cise-seminar-october-5-2018-alexander-wyglinski-worcester-polytechnic-institute-wpi/
  • Alexander M. Wyglinski. “Intelligently Connected Vehicles: When The Information Highway Meets The Road”, University of Massachusetts Boston Seminar, November 2, 2018. URL: https://www.umb.edu/academics/csm/engineering/engineering_seminar_alex_wyglinski
  • Alexander M. Wyglinski. “Biology Meets Cognitive Radio – Using Nature to Make Good Decisions”, International Symposium on Wireless Communication Systems 2016 Keynote Presentation, September 23, 2016. URL: http://iswcs2016.radiokomunikacja.edu.pl/welcome/menu/keynotes
  • Alexander M. Wyglinski. “Bumblebees and Beamforming — Enabling the Vehicular Internet-of-Things”, WiFiUS Summer School on Wireless Challenges in the Internet of Things, June 13, 2018. URL: http://209.140.21.224/~jwifiusa/summer-school-on-iot-2018

In The Media:

Intellectual Merit:

Devising a connected vehicle networking architecture that can operate in rapidly time-varying environments, such as highway conditions, is a technically challenging problem. Although solutions do exist for supporting some level of wireless access in these environments, additional investigation is needed to make these types of wireless networks more reliable and enable them to support a larger number of vehicles. Researchers initially turned to highly social forms of simple organisms, such as ants and honeybees, in order to understand their collective behavior and model their vehicular networks after them with respect to resource optimization, such as distributed access to wireless spectrum. However, as this proposed effort will demonstrate, vehicular nodes that form part of a connected vehicle network behave more like organisms such as bumblebees, where each individual node shares information socially yet make decisions independently of one another. Consequently, there is significant potential for the proposed network architecture to achieve superior performance outcomes in terms of spectral efficiency and reliable, rapid wireless access that would otherwise not be considered possible given conventional networking techniques. Thus, this research will yield transformative advances in DSA and connected vehicle networks. Furthermore, it is strongly believed that the proposed activities possess substantial impact that will help shape the direction of this research field.

Broader Impact:

Connected vehicle technology has the ability to provide drivers with a significantly higher level of environmental awareness and safety relative to the present day. Thus, enabling reliable, seamless, and efficient access to wireless spectrum in order to support this vehicular connectivity is core to this safety technology. The interdiscplinary project will make an educational contribution via the mentorship and training of two graduate students (one Electrical & Computer Engineering Ph.D. student and one Biology & Biotechnology Ph.D. student) with an emphasis on identifying qualified students from underrepresented groups. It is expected the research activities from this proposed effort will be extensively collaborative, with the outcomes forming the dissertations of both students. WPI’s “Two Towers” paradigm for engineering education, defined by the institutions motto of Lehr und Kunst (“Learning and Skilled Art”), will be leveraged in order to introduce undergraduate students to this proposed effort via several project-oriented experiences emphasizing intensive learning and the direct application of knowledge. In particular, PI Wyglinski, Co-PI Gegear, and Co-PI Ryder will co-advise during the proposed effort several interdisciplinary senior-level undergraduate design experiences, referred to as Major Qualifying Projects (MQPs), which are a quintessential graduation requirement of every WPI  undergraduate student. These MQPs will enable undergraduate students from Electrical & Computer Engineering and Biology & Biotechnology to work together on projects related to the proposed effort, synthesizing fundamental concepts from their specific major while simultaneously learning concepts from other disciplines during their project experience. Finally, project outcomes will be disseminated via conference presentations, tutorials, and archival journal
publications.

 

nsf