Three-Tier Computing Platform Optimization: A Deep Reinforcement Learning Approach

Table 1

Summary of some of the key notations used.

Notation	Description
N	Set of IoT nodes
	Set of small base stations (high-performance processors)
	Set of macro base stations (collaborative edge devices)
	Set of resource edge devices
	Task edge device
	Set of subcarriers from user to small BS
	Set of subcarriers from small base station to macro BS
	Bandwidth
	Transmission rate
	Power allocated to
	Power allocated to
	Indicator variable for
	Local computing power
	CPU computational frequency of user
	Effective capacitance coefficient
	The available power
	Latency
	Number of bits in
	Computing capacity of IoT node
	The computational capacity of small base station
	Computing capacity of macro base station
	Propagation latency
	System latency
	Total APC
	Required processing capacity for user
	Energy consumed at the small base station
	Static power offset
	Total consumed power
	Reward at time slot
	Loss function
	Learning rate
	Temporal difference
	Policy
	Maximum communication capacity allocated by and
	Maximum computing capacity allocated by and