Anti-Jamming Communication Using Slotted Cross Q Learning

Most of the existing intelligent anti-jamming communication algorithms model sensing, learning, and transmission as a serial process, ideally assume that duration sensing learning timeslots is very short, almost negligible. However, when jamming environment changes rapidly, time can no longer be ignored, adaptability wireless system to time-varying will significantly reduced. To solve this problem, paper proposes parallel Q-learning (PQL) algorithm. In case long time, by modeling processes, transmitter remains silent during Aiming at situation PQL algorithm susceptible faster, an for based on Slot Cross (SCQL). rapid change channel, sense learn patterns in multiple successive periods same timeslot, use Q-tables different periods, so achieve effect reliable with jamming. The simulation results show collision rate proposed under condition blocking equivalent traditional (QL), but timeslot utilization higher. Compared PQL, has slot lower rate. random frequency hopping (RFH) anti-jamming, not only higher utilization, also