Collaborative Research: SWIFT: Dynamic Spectrum Sharing via Stochastic Optimization

点击翻译按钮获取中文摘要

The next generation of communication networks must support extremely complex systems and challenging applications such as smart city/home, smart manufacturing, autonomous vehicles, and virtual and augmented reality. These new applications should be more accessible worldwide and lead to wider harmonization, lower broadband costs, and reduce the digital divide. At the same time, radar sensing is becoming more pervasive in areas with increasing demands for the 5G spectrum. These ever-increasing demands on spectral resources require the use of intelligent spectrum scheduling techniques. Meanwhile, several new technologies like reconfigurable intelligent surface (RIS) and dual-function unmanned aerial vehicles (UAVs) provide additional design degrees of freedom. This project develops the general mathematical framework that is needed to integrate and optimize these new technologies for resilient coexistence over a shared spectrum. Current spectrum allocation between communication and radar users, designed by regulatory bodies, aims to avoid interference between users at all times. This conservative approach is not suited in the wake of increasing demand for throughput and dual communication and sensing functionality and results in a reduced capacity. This inefficiency is exacerbated by the introduction of additional design degrees of freedom and the corresponding constraints related to the characteristics and dynamics of new technologies like RIS modules and UAVs. This project develops a new paradigm in which spectrum sharing moves from hard deterministic constraints to stochastic schemes with a desired low probability of harmful interference. This enables leveraging recent advances in stochastic programming to derive resilient solutions. In this approach, the constraints of optimization are random variables that have to satisfy the bounds given by interference limits with a desired high probability. These stochastic constraints may be time-varying to account for uncertainty in the dynamic environment. A key component to enable this development is an accurate characterization of the mutual interference and performance tradeoffs among coexisting radar and communication nodes. his project develops the general mathematical framework that is needed to integrate and optimize these new technologies for resilient coexistence over a shared spectrum.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.