Table 2.

Key points of the analyzed articles in the data replication and placement techniques in IoT

Source	Main idea	Technique	Results
Data replication and placement techniques in IoT
He et al. [71]	Supplying load-balancing and low-cost cloud data replica distribution strategy in IoT setting	A method for selecting low access cost storage server was presented based on access positions and frequencies	When achieving the minimum number of replicas, the replicas are adaptively distributed to the storage servers according to the data access status with low access cost
Cho et al. [37]	Investigating replica recognition in mobile applications	Efficient distributed detection methods of replicas (EDDRs)	The suggested approaches detect replicas with a high degree of accuracy and almost no detection mistakes. Moreover, the techniques' distributed and cooperative strategy decreases the amount of energy needed to identify replicas while also delivering faster replica recognition than previous systems
Zhang et al. [72]	Suggesting a data replica placement scheme for cloud storage under healthcare IoT environment	The suggested approach employs the MOX (Mosquitoes Oviposition Mating) algorithm to determine the best data replica placement answer and the SA (Simulated Annealing) technique to determine the best user access request allocation to the appropriate data replica	The suggested technique can increase system scalability and reliability while also reducing user access time and promoting load balancing
Qaim and Ozkasap [74]	Offering a totally distributed hop-by-hop data replication method for IoT-based wireless sensor systems	A data replication technique for improved data availability in IoT-based sensor systems (DRAW)	DRAW increases data availability and average replicas produced in the network by a maximum of 15% and18%, respectively, compared to a modern method. Besides, DRAW has a greater replica spread, which influences the network's data distribution quality
Shao et al. [20]	Suggesting a new data replica placement method for coordinated processing data-intensive IoT workflows in a collaborative edge and cloud computing environment	Suggesting an IoT-based method for data replica placement in collaborative edge and cloud settings	In comparison to existing algorithms, the suggested technique can not only discover a higher quality solution for data replica placement, but it also requires a smaller computational budget
Wakil et al. [70]	Decreasing the waiting time of IoT applications to access the replicated things utilizing a hybrid algorithm Decreasing the replica selection cost in IoT settings utilizing a hybrid algorithm	A hybrid optimization algorithm (Ant Colony Optimization (ACO) and Genetic Algorithm (GA))	Compared to High-QoS (Quality of Service) First-Replication (HQFR) and the dynamic cost-aware re-replication and re-balancing technique, the hybrid method performs better
Wang et al. [75]	Investigating active data replica recovery for quality-assurance big data analysis in information-centric IoT	Rarity-aware data replica recovery (RADR) algorithm	The RADR technique outperforms a conventional direct data recovery approach by a substantial margin
Wang and Batiha [9]	Enhancing the load balancing among the IoT centers utilizing repellent pheromone	Simulation using the cloudsim in Improved version of ACO (IACO)	In terms of waiting time and load balancing, the suggested solution excelled the ACO, HQFR, and Response Time-oriented Replica Management (RTRM) methods
Liu et al. [76]	Suggesting edge node data replica management method for distribution IoT	The data processing architecture of distribution IoT based on edge computing The edge computing data copy management approach is intended to achieve adequate distributed IoT data backup	The suggested replica management technique, which combines local and cluster management, has a faster average processing time than the conventional data replica management technique and may significantly enhance IoT data processing performance with a large dispersion