The cognitive ability of equipment embodies the fundamental attributes of cognitive electronic warfareand it is also the key to effective jamming on complex and intelligent electronic equipment.As a hot technique in the field of artificial intelligencereinforcement learning has the ability of self-learning not relying on priori datawhich is an important approach to multifunctional radar jamming.Based on the review on traditional algorithms of radar jamming decision-makingas for reinforcement learning based radar jamming decision-makingthe principles and status quo of the technology are analyzedand its performance is verified through simulations.Finallya summary is givenand the outlook for the technology is predicted.