Search

One of the key parameters that plays a major role in enabling the data rate requirements is spectrum or bandwidth, which is scarce and expensive. Therefore, spectrum management policies are required to optimize its usage to meet all the requirements and the promised data rate growth. Another strategy to deal with spectrum scarcity is to move toward higher frequency bands (terahertz bands), which are expected in the next 6G communication standard. It is therefore important to develop not only new techniques that enable efficient dynamic spectrum access and sharing at such bands, but also suitable channel models for the terahertz bands. Meanwhile, offloading mechanisms are very promising for cellular networks where a plethora of options have been proposed in the research arena in terms of device-to-device, licensed assisted access, or WiFi offloading, among others; but their behavior, when operated at high frequencies (terahertz band) remains unclear. Therefore, this article will tackle two technologies that will shape future networks: terahertz channel modeling/communications and offloading mechanisms.

Statistical information in terms of spectrum occupancy is useful for the efficient and smart dynamic spectrum sharing, and it can be obtained by long-term, broadband, and wide-area spectrum measurements. In this paper, we investigate an energy detection (ED)-based spectrum measurements, in which the noise floor (NF) estimation is a key functionality for the appropriate ED threshold setting. Typically, the NF has the slowly time- varying property and frequency-dependency, and several NF estimation algorithms, including forward consecutive mean excision (FCME) algorithm-based method, have been proposed. However, these methods did not deeply consider the slowly time varying property of the NF and is computationally inefficient. Accordingly, we propose a computational complexity reduction algorithm based on NF level change detection. This algorithm is computationally efficient, since it skips the NF estimation process when the NF does not change. In numerical evaluations, we show the efficiency and the validity of the proposed algorithm.