An AI-Based Incumbent Protection System for Collaborative Intelligent Radio Networks

Abstract

Traditionally, the radio spectrum has been allocated statically. However, this process has become obsolescence as most of the allocated spectrum is underutilized, and the part of the spectrum that is mainly used by the technologies that we use for daily communication is over-utilized. As a result, there is a shortage of available spectrum to deploy emerging technologies like 5G that require high demands on data. Several global efforts are addressing this problem, i.e., the Citizens Broadband Radio Service (CBRS) and Licensed Shared Access (LSA) bands, to increase the spectrum reuse by providing multi-tiers spectrum sharing frameworks in the re-allocated radio spectrum. However, these approaches suffer from two main problems. First, this is a slow process that may take years before authorities can reassign the spectrum to new uses. Second, they do not scale fast since it requires a centralized infrastructure to protect the legacy technology and coordinate and grant access to the shared spectrum. As a solution, the Spectrum Collaboration Challenge (SC2) challenge has shown that Collaborative Intelligent Radio Network (CIRN), i.e., Artificial Intelligence (AI)based autonomous wireless radio technologies that collaborate, can share and reuse spectrum efficiently without any coordination and with the guarantee of incumbent protection. In this paper, we present the architectural design and the experimental validation of an incumbent protection system for the next generation of spectrum sharing frameworks. The proposed system is a twostep AI-based algorithm that recognizes, learns, and proactively predicts the transmission pattern of the incumbent in near real-time, less than 300 ms to perform a prediction, with an accuracy above 95% to correctly predict where the incumbent is transmitting in the future. The proposed algorithm was validated in Colosseum, the RF channel emulator built for the SC2 competition, using up to two incumbents simultaneously, which have different transmission patterns, and sharing spectrum with up to 5 additional networks.

Collaborative Flow Control in the DARPA Spectrum Collaboration Challenge

“…None of these published works discussed the collaborative flow control mechanisms of the SCATTER system. Overall architecture [25], predicted slot selection [26], physical radio details [27], technology-aware incumbent avoidance [28], [29], and the software architecture details used by the proposed algorithm [21] are discussed in previous work. Note that the discussed architecture could easily be implemented on top of modern wireless radios such as 5G new radio, 4G or Wi-Fi as we discuss later in this section.…”

“…• Incumbent Detection Stage: As with the passive incumbent, information about the presence of the incumbent is received and the policy informs the other modules. • Learning Stage: By combining signal recognition and pattern learning techniques using samples of the spectrum, SCATTER radios recognize and learn the transmission patterns of the active incumbent [29]. Here information from possible violations is also used to enhance the capabilities of the recognition system.…”

“…Other optimization modules are responsible for fulfilling the requirements as well as possible. All details about the behavior and the algorithm of the Active incumbent policy are described in previous work [28], [29].…”

“…No flows are blocked by the Active incumbent policy itself. Note that the active performancedriven policy still optimizes the policies based on the score of the neighboring teams, and unstable IMs are blocked first if necessary [28], [29].…”

A Survey on Machine Learning-Based Performance Improvement of Wireless Networks: PHY, MAC and Network Layer

“…), non-adaptive modulation scheme, static non-application cognizant MAC, etc. • Wireless interference identification [128,169,[177][178][179][180][181][182][183][184][185][186][187][188][189][190][191][192][193] MAC analysis…”