A cognitive radio resource allocation method based on case-based reasoning and cooperative Q learning

The invention discloses a cognitive radio resource allocation method based on case-based reasoning and cooperative Q learning. According to the invention, a case-based reasoning technology, a cooperation algorithm and a Q learning algorithm are combined to realize joint distribution of channels and power. The method comprises the following steps: firstly, determining a reward function, learning aplurality of random resource allocation cases by utilizing a traditional Q learning algorithm, and constructing a case library to store characteristic values, Q values and utility values of the cases;Retrieving a case most similar to the current case through matching, extracting a Q value of the case, and normalizing the Q value to serve as an initial Q value of the new case; And then, accordingto the reward value, a cooperation algorithm is adopted to fuse the Q value, and learning is carried out by referring to experience of other users. Aiming at the problem that the traditional Q learning convergence speed is low,