Resource Allocation in UAV-Assisted Wireless Powered Communication Networks for Urban Monitoring

Algorithm 1

MJDDPG-based resource allocation algorithm.

   Initialize weights for main network and target networks;

   Initialize experience replay buffer, exploration variance and action exploration probability;

   for to

   for to

     Update the environment state and observe the current state and record and ;

     Based on selection action;

     Execute action to update state and calculate reward value ;

     Store the experience tuple into the experience buffer;

     if buffer is full then

       Randomly sampled mini-batch samples from the experience buffer;

       Computing the target network values;

       Updating the critic network by minimizing the loss of the critic network;

       Updating the actor network by maximizing actor network losses;

       Updating target network parameters;

       Updating the action random parameters;

     end if

   end for

 end for