A Semiopportunistic Task Allocation Framework for Mobile Crowdsensing with Deep Learning

Algorithm 1:

Participant selection based on DDQN.

1: Initialize -network, target network with

2: while In each episode do

3: Initialize

4: for step in episode do

5:  With probability , select a random action

6:  otherwise select

7:  

8:  store transition in replay_buffer

9:  if current system budget cannot afford any participant then

10:    is terminal state

11:  else

12:   

13:  end if

14:  sample random minibatch of transitions from replay_buffer

15:  perform minibatch gradient descent

16:  every updated period

17: end for

18: end while

Algorithm 1: