Dynamic Task Assignment Framework for Mobile Crowdsensing with Deep Reinforcement Learning

Algorithm 2

Maximum value maximum flow matching strategy (MaxflowQ).

Input: The platform worker set in the current period, the task set within the perception range of each worker, the set of workers .

Output: Worker task matching set in the current period.

 1: According to , , constructs the flow graph

 2: Initialize flow f to 0

 3: while there exists an augmenting path in the residual network do

 4: Select an augmenting path with the largest Q value

 5: 

 6: Augment flow f along with

 7: Update residual network

 8: Save worker task matches

 9: Update matching set

 10: end while