The invention discloses a
channel access method and apparatus based on intelligent time slot selection. The problem that the dynamic topology of the network in a
wireless sensor network and the efficient utilization of a channel cannot exist at the same time is mainly solved. The method comprises the following steps: establishing a model of the
wireless sensor network, and combining a
wireless sensor network environment with a
reinforcement learning method; determining a customized frame structure and the number of time slots; initializing a state value vector V(s) and a time slot selection probability Pi; selecting a time slot according to the time slot selection probability Pi for transmission, and updating a
reward value of each time slot according to the feedback condition of the channel access; then updating a state value v of the corresponding time slot and a qualification trace
delta according to the reward of each node; after the update is completed, evaluating the state of each node, when Vi is greater than Vth, deeming that the node is in a
stable state at the ith time slot, finally adjusting a temperature factor parameter of a Boltzmann strategy according to the above state evaluation state and the state value, and selecting an optimal time slot of the node for sending. By adoption of the
channel access method and apparatus disclosed by the embodiment of the invention, the channel
utilization rate of the
wireless sensor network is not only ensured, the reliability of the
wireless sensor network is also improved, and the application value is wide.