Fragmented Task Scheduling for Load-Balanced Fog Computing Based on Q-Learning

Table 1

Sets, parameters, and variables.

	Cloud computing server
	Set of fog nodes
	Set of IoT tasks
	State space of system at time slot
	Action space of system at time slot
	Possible states of available resources
	Binary variable. One indicates IoT task is real-time task; zero, vice versa;
	Binary variable. One indicates there is a delay violation while serving task at fog node ;
	Binary variable. One indicates there is a security violation while serving task at fog node ;
	Required resources of IoT task ;
	Security requirement of IoT task ;
	Computing required of IoT task ;
	Number of fragments of IoT task ;
	Maximum delay threshold of IoT task ;
	Available resources of fog node ;
	Security level of fog node ;
	Processing power of fog node ;
	Response time of fog node for IoT task ;
	Response time of cloud
	Reward based on response time of fog node ;
	Reward based on security adherence ;
	Transmission delay;
	Propagation delay;
	Processing delay;
	Task size;
	Link bandwidth;
	Distance between fog node and source of IoT task ;
	Speed of light
	Arrival rate of tasks at fog node ;
	Service rate of tasks at fog node ;
	Reward received by agent after taking an action in state ;
	Cumulative future reward;
	Optimal fog node ;
	Learning rate
	Discount factor