Table 1.

Edge Computing for Healthcare Systems: Problems in Existing Systems and Techniques Used to Solve

Reference	Problem	Technique Used
Edge computing in IoT context: Horizontal and vertical Linux container migration42	The increasing IoT devices pose challenges to data privacy and network performance.	Cloud4IoT: A platform that performs vertical (offloading) and horizontal (roaming) movement of various IoT operations
A framework for edge-assisted healthcare data analytics using federated learning43	Existing machine learning (ML) algorithms that were developed and validated on a single central data store will no longer work.	The approach enables efficient and scalable integration of user-generated wellness and behavioral data. It proposes a conceptual framework for leveraging Edge computing in healthcare analytics. User-generated data is utilized to provide valuable insights into wellness and behavior. Distributed machine learning methods, such as Federated Learning, are employed to protect privacy. The framework ensures privacy while analyzing and processing the user-generated data. Edge computing enhances the efficiency and speed of healthcare analytics.
Edge computing-based secure health monitoring framework for electronic healthcare system44	The Quality of Services (QoS) parameter in the cloud-based healthcare system is lowered by high latency and response times.	Patient data security in the cloud is maintained using the ABE (Attribute-Based Encryption) approach and an access policy. ABE ensures that only authorized users with specific attributes can access and decrypt patient data. The access policy defines the rules and criteria for granting access to patient data. Adding more edge devices at the edge layer enhances the efficiency of the proposed system. Edge devices can offload processing tasks from the cloud, reducing latency and improving response times. Distributing computation and storage closer to the data source improves system performance and scalability.
Health-Fog: An ensemble deep learning-based Smart Healthcare System for Automatic Diagnosis of Heart Diseases in integrated IoT and fog computing environments45	The present edge models have significant drawbacks and are only concerned with either results precision or response time reduction, not both.	Health-edge system offers low latency and power-saving data processing solutions. This technology effectively manages the data of heart patients while also automatically diagnosing heart disorders.
A secured framework for SDN-based edge computing in IoT-enabled healthcare systems18	The associated data (patients’ personal, sensitive information) and low-powered technologies are vulnerable to several security risks.	A simple authentication mechanism is employed to verify the IoT devices before granting access. Patients’ data is transferred to the Edge server for processing once the authentication is successful. Multiple Edge servers cooperate to ensure scalability and handle the processing load efficiently. An SDN (Software-Defined Networking) controller is established to make intelligent decisions and manage the network effectively. The SDN controller facilitates dynamic routing, load balancing, and resource allocation in the Edge network. By leveraging the cooperation of Edge servers and utilizing an SDN controller, the system achieves improved scalability and optimized performance.
An effective training scheme for deep neural networks in edge computing enabled Internet of medical things (IoMT) systems46	The main problem in deploying an efficient latency-aware health monitoring system is that DL inference is embedded into an edge device, which has limited computational power.	The developed ETS-DNN (Deep Neural Network) model makes it easier to gather and process data quickly so that judgments may be made quickly using the patterns that are present in the data.
Analysis of Privacy-Preserving Edge Computing and Internet of Things Models in the Healthcare Domain47	Healthcare systems acquire and manage sensitive information, and the misuse of this data by malicious attackers can have disastrous effects.	The lightweight privacy-preserving data aggregation scheme for edge computing (LDPA-EC) is introduced. LDPA-EC aims to reduce computational costs while ensuring data integrity and confidentiality in edge computing scenarios. The scheme employs lightweight techniques to preserve privacy without compromising the quality and security of the aggregated data. LDPA-EC utilizes efficient cryptographic algorithms and protocols for secure data aggregation. The scheme ensures that sensitive data remains confidential and cannot be inferred from the aggregated results. By lowering computational costs and preserving data integrity and confidentiality, LDPA-EC provides an efficient solution for privacy-preserving data aggregation in edge computing environments.
An intelligent edge computing-based semantic gateway for healthcare systems interoperability and collaboration48	Due to a lack of cooperation and information sharing among healthcare systems (clinics, hospitals, and pharmacies), compatibility with these systems is still an issue.	Each healthcare system incorporates an intelligent edge semantic gateway within its infrastructure. The intelligent edge semantic gateway utilizes a web application and RESTful API. The gateway securely facilitates the sharing and disclosure of patient data from each healthcare system. The web application provides a user-friendly interface for accessing and managing patient data. The RESTful API allows seamless integration and interoperability between different healthcare systems. The intelligent edge semantic gateway ensures data security and privacy during the collaboration process. Through the gateway, healthcare systems can securely collaborate and exchange patient data, enabling improved coordination and decision-making in healthcare scenarios.
Edge computing health model using P2P-based deep neural networks.19	Deep learning algorithms are hampered by the over-fitting issue in a neural network and increased computing costs related to elevated levels of time complexity. Response delays are common in large data learning processes and deep neural network-based data extraction procedures because of these issues, which exponentially raise the cost of data extraction.	The reaction time of a deep learning framework has been improved by implementing P2P network principles and an edge computing architecture. Faster response times resulting from reduced user waiting time contribute to a more enjoyable user experience. The application of efficient network consumption prevents unnecessary growth in computing power within the server architecture. The optimized system facilitates practical use by ensuring effective resource utilization. By leveraging the fundamentals of P2P networks and edge computing, the deep learning framework achieves enhanced performance, decreased user waiting time, and practical network usage.