LAUSR

Efficient utilization of both licensed and unlicensed spectrum provides a fundamental solution to address the spectrum crisis. In this article, we propose a multi-scale dynamic spectrum allocation framework for software-defined heterogeneous networks in which both licensed and unlicensed spectra are used on a flexible and on-demand basis. On the micro scale, the small-timescale spectrum allocation based on real-time user demands, time-varying channel states, and dynamic interference conditions is formulated as a two-sided matching problem between resources and users, which is solved by using a pricing-based matching algorithm. On the macro scale, the gap between the achieved and desired average transmission rates is utilized as a feedback signal to adapt intercell spectrum coordination with large-timescale variations of traffic demands. Next, four typical application scenarios are elaborated to illustrate the potential benefits of the proposed framework, and its feasibility is validated via a case study. Finally, we identify several research opportunities and conclude this article.