Mobile Networks are subject to "signaling storms" launched by malware or apps, which overload the the bandwidth at the cell, the backbone signaling servers, and Cloud servers, and may also deplete the battery power of mobile devices. This paper reviews the subject and discusses a novel technique to detect and mitigate such signaling storms. Through a mathematical analysis we introduce a technique based on tracking time-out transitions in the signaling system that can substantially reduce both the number of misbehaving mobiles and the signaling overload in the backbone.