基于滑动窗的时域非相关信誉值机制

A Time Unrelated Reputation Mechanism based on Sliding Windows

协作频谱感知是认知无线电通信的重要环节,频谱感知数据篡改(Spectrum Sensing Data Falsification,SSDF)是协作频谱感知面临的严重安全威胁。SSDF攻击方式逐渐呈现智能化的趋势。变策略SSDF攻击行为是一种新的智能攻击方式,它通过改变自己的攻击参数,时而伪装成诚实用户,时而发起SSDF攻击,对协作频谱感知会造成更严重的破坏。而目前针对SSDF攻击方式的传统检测算法,都假定其攻击参数始终保持不变。为了应对变策略SSDF攻击方式,提出了一种基于滑动窗的时域非相关信誉值机制,能够在攻击时刻开始之后,迅速准确检测出恶意用户。

Collaborative spectrum sensing, as an important part of cognitive radio, is faced with fairly serious security threat-SSDF （Spectrum Sensing Data Falsification） attack. Nowadays SSDF gradually exhibits an intellectualization trend, and SSDF mode, with variable strategies, is a novel intelligent attack, and by changing its parameters, the attackers pretend to be honest users and lauches SSDF attacks, thus causing serious damage on cognitive radio network. However, most traditional detection algorithms assume that the attack parameters of SSDF keep invariable. To handle the SSDF attcack with variable parameters, a time unrelated reputation mechanism based on sliding window is proposed, which could detect malicious users quickly and accurately when the attackers start their malicious behaviors.