Jamming attacks can severely affect the performance of wireless networks due to the broadcast nature. The most reliable solution to reduce the impact of such attacks is to
detect and localize the jammer. In this paper, we propose our research into participatory sensing based scheme, named as CrowdLoc, for the collection of measurements to collaboratively localize a jammer in wireless ubiquitous environments which are suffering from jamming attacks. CrowdLoc mainly contains three phases: 1) Crowds as Sensor. The sensor nodes at the boundary of jammer region are weakly impacted by the jamming attack, and conduct the sensing functions to record the information related to the jammer, such as received signal strength (RSS); 2) Crowds as Network. These boundary nodes cooperate with each other to share the recorded measurements of the jammer; and 3) Crowds as Estimator. Based on the crowdsourcing measurements of the jammer, we propose a
novel localization scheme to estimate the position of the jammer: Range-based Jammer Localization (RJL). As opposed to existing solutions, RJL is independent of the propagation parameters, which are difficult to obtain in hostile jamming circumstance. The experimental results indicate that the localization accuracy of RJL is close to the Cramer-Rao Bound (CRB) for the RSS-based Localization in most area.
