EAGER: Malicious Behavior Detection in Hybrid Dynamic Spectrum Access

为了解决日益严重的频谱短缺问题，动态频谱接入被设想为一套有前途的新的频谱管理范式。尽管它使未充分利用的许可频段的机会接入成为可能，但各种实际因素，如环境动态、故意干扰和未经授权的传输，阻碍了它的广泛部署。最近发布的FCC规则表明，参与式实时频谱感知可以大大提高数据库驱动频谱共享的频谱利用效率，形成混合动态频谱共享的新范式。然而，二次用户与频谱数据库之间频繁的信息交换很容易被恶意用户拦截和操纵，这不仅降低了频谱效率，而且给混合动态频谱接入系统带来了严重的安全隐患。该项目将探索保护未来认知无线电系统的新范式，重点是不合规行为检测。该项目的成功将成为在不久的将来提供可靠无线通信的关键推动者。该项目将研究新定义的混合动态频谱接入中的几个基本安全挑战。第一个研究任务将确定新的攻击模型，损害频谱效率，然后提供适应未来无线系统的对策。由于数据库驱动频谱访问的固有特性，主用户仿真攻击者可以检索频谱可用性信息，在现有用户（IU）不存在时作为现有用户（IU）执行，或者试图增加辅助用户的传输功率以干扰现有的IU。利用存储在数据库中的传感结果，设计新的检测方案来减轻攻击带来的影响。第二个研究任务利用物理层方法来检测不同通道模型下的未经授权访问。为了解决这个问题，通道可用性信息将用于检测恶意的辅助用户。同时，将综合考虑实用性和效率，发展检测机制。此外，该项目包括强大的验证组件，它结合了模拟研究、原型设计和实验。因此，它将为包括无线网络、无线通信和网络安全在内的跨学科学科提供有效的培训基地，所有这些都对未来国家劳动力的多元化专业人员至关重要。

To tackle the ever-increasing spectrum scarcity issue, dynamic spectrum access is envisioned as a set of promising new spectrum management paradigms. Although it has enabled the opportunistic access of underutilized licensed bands, various practical factors, such as environmental dynamics, intentional interference, and unauthorized transmission, hinder it from wide deployment. The recently released FCC rules suggest participatory real-time spectrum sensing can greatly improve the spectrum utilization efficiency for database-driven spectrum sharing, which forms a new paradigm, hybrid dynamic spectrum sharing. However, the frequent information exchanged between secondary users and spectrum database can be easily intercepted and manipulated by malicious users, which not only downgrades the spectrum efficiency but also incurs severe security breaches to the hybrid dynamic spectrum access system. This project will explore new paradigms of safeguarding the future cognitive radio system with focus on non-compliance behavior detection. The success of this project will serve as a key enabler to provide reliable wireless communication in the near future.This project will investigate several fundamental security challenges in the newly defined hybrid dynamic spectrum access. This first research task will identify new attack models that compromise the spectrum efficiency and then provide countermeasures adapted to future wireless systems. Due to the inherent nature of database-driven spectrum access, primary user emulation (PUE) attackers can retrieve the spectrum availability information to either perform as the incumbent user (IU) when it is not present, or try to increase secondary users' transmission power to interfere with present IUs. Featuring the sensing results stored in the database, novel detection schemes will be designed to mitigate the influence brought by the attack. The second research task leverages physical-layer approaches to detect unauthorized access under different channel models. To address this issue, channel availability information will be used to detect malicious secondary users. Meanwhile, the detection mechanisms will be developed with joint consideration on practicality and efficiency. Additionally, the project includes strong validation component that combines simulation study, prototyping, and experimentation. It will thus provide an effective training ground for interdisciplinary subjects including wireless networks, wireless communication, and cybersecurity, all of which are critical to diversified professionals for future national work force.