Research Article
A Semiopportunistic Task Allocation Framework for Mobile Crowdsensing with Deep Learning
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: |