The invention discloses a
cognitive model adaptive frequency hopping method of an industrial
wireless WIA-PA network, and belongs to the technical field of industrial
wireless communication. The invention aims to solve the core problems of weak
interference resistance, low spectrum
utilization rate, poor
time sequence adaptation and lack of environmental self-learning of the existing WIA-PA network frequency hopping scheme. According to the technical scheme, an environment
perception-cognitive decision-frequency hopping control-WIA-PA
time sequence cooperation architecture is constructed, a
reinforcement learning + channel quality evaluation two-dimensional
cognitive model is adopted, parameters such as industrial environment interference intensity, a channel
signal-to-
noise ratio (SNR) and a
frequency spectrum occupancy rate are collected in real time, a frequency hopping pattern adaptive to a WIA-PA
superframe time sequence is dynamically generated, and the WIA-PA time sequence cooperation is realized. And completing channel switching in a CAP / CFP gap (the
switching time is less than 10ms). According to the invention, the anti-interference intensity in the industrial temperature range of-40 DEG C to 85 DEG C reaches-70
dBm, the spectrum
utilization rate is improved by more than 40%, the
data transmission success rate is greater than or equal to 99.5%, the
delay is reduced to be within 5ms, and the method can be widely applied to WIA-PA network extreme interference scenes such as intelligent manufacturing,
process industry, intelligent power grids and the like.