Dynamic Q-Learning-Based Optimized Load Balancing Technique in Cloud

Table 1

Comparison techniques of different framework algorithms.

Author	Technique	Metric	Experiment environment	Focus	Comparison technique	Con
Ullah [1]	BAT algorithm	Energy consumption	Data centre	VM, T, M	ABC	Contributes to bottlenecks, because it needs larger operation coordination
Panda et al., [2]	Hungarian algorithm	Layover time	Windows machine	T, time, schedule	FCFS, HA-LT, HA-CLT	There is still a limited possibility of having the wrong responses
Rjoub et al., [3]	BigTrustScheduling	Monetary cost, makespan, cost	Hadoop, VM	VM	SJF, RR, PSO	If all remote procedures are overrun, all procedures would be manually assigned
Ko and Cho [4]	Auto distributed speed scaling + load balancing algorithm + G/G/1/PS queue	Speed, price	Server based	Response time, workload	M/M/1 queue	The tasks are planned slowly because it does not determine the node’s current execution period
Xu et al., [5]	Resource scheduling + NSGA-II	Average service latency, average stability	Windows machine	Average service latency, average stability	FIRMM, RSS-IN	Complexity high, homogenous resources
Gawali and Shinde [6]	MAHP + BATS + BAR + LEPT + DaC	Bandwidth and memory utilization	Data centre	Turnaround time, response time	BATS, IDEA, heuristic approach	Completion and run time of node tasks are not considered in MAHP. That is why it takes a lot of time to complete the tasks
Gopinath and Vasudevan [7]	Min-min algorithm and max-min algorithm LBA	Time make span	Cloud sim	Task, resources	MCT, MET	Leads to starvation
Armstrong et al., [10]	Energy efficient architecture	Component performance	Data centre	Time, power	DVFS, RAPL	Suffers from starvation
Asghari et al. [11]	Mobile-edge computing (MEC)	Reduce global latency	Cloud sim	Response time	Deep Q-network (DQN) and markov game (MG)	There is still a limited in quality of services
Asghari and Sohrabi [12]	Energy efficient architecture	Processors’ energy consumption	Cloud sim	Reducing the voltage and frequency of processors	DVFS technique	There is still a limited in quality of services