Table 2 Attacks against CR environments and applicable countermeasures

PU emulation

CR networks based on non-cooperative schemes. Note: in such environments an attack against a specific SU may only affect that SU.

False alarms due to fake signals.

Sensing techniques that consider a priori known characteristics of the legitimate PU signals.

The affected SUs are denied access to the affected spectrum holes due to greedy or malicious motivations, and, therefore, their performances are likely to decrease.

Solutions based on capabilities such as location determination techniques and access to geo-location information about a priori known PUs.

Spectrum sensing data manipulation/falsification

CR networks based on cooperative schemes. Note: in such environments, attacks against a single SU may affect several SUs or the entire network.

Cooperative spectrum sensing accuracy decreases due to the propagation of false alarms and/or missed detections that are forged.

Solutions for providing characteristics such as mutual authentication, data integrity, and data encryption.

If learning is considered, the behavior of the SUs is likely to suffer a negative impact on the long-term basis due to the usage of manipulated data in the learning process.

Outlier detection techniques.

Malicious attacks may impact on PUs by inducing missed detections.

Approaches based on the exploration of spectrum spatial correlation and location techniques.

Malicious and greedy attacks may impact the performance of the SUs by inducing false alarms.

Schemes that enable determining the trustiness of the SUs and, therefore, dropping reports from untrustworthy sources.

Deployment of dedicated trusty sensors.

Usage of mechanisms to selectively forget past information in order to make beliefs and learning outputs temporary.