Security-Aware Reinforcement Learning under Linear Temporal Logic Specifications | IEEE Conference Publication | IEEE Xplore