Optimization of LoRa SF Allocation Based on Deep Reinforcement Learning

Algorithm 1

Procedure of deep Q network.

Initialize LoRa node based on LoRaWAN

Initialize replay memory D to capacity N

Initialize action-value function with random weights

Initialize target action-value function with weights

For episode =1,M do do

  Initialize and choose state

   from LoRa network server(MAC)

  For do

   With probability select a random action otherwise select execute action in emulator

   Observe reward and next state

   Store experience

   If episode terminates at step j +1 then

   Else

   End if

   Perform a gradient descent step on with respect to the network parameters

   If batch size > = memory capacity then

    Update

   End if

  End for

End for