Efficient resource management is a challenging task in distributed systems, such as the Internet of Things, fog, edge, and cloud computing. In this work, we present a broad overview of the Internet of Things ecosystem and of the challenges related to managing its resources. We also investigate the need for efficient resource management and the guidelines given/suggested by Standard Development Organizations. Additionally, this paper contains a comprehensive survey of the individual phases of resource management processes, focusing on resource modeling, resource discovery, resource estimation, and resource allocation approaches based on performance parameters or metrics, as well as on architecture types. This paper presents also the architecture of a generic resource management enabler. Furthermore, we present open issues concerning resource management, pointing out the directions of future research related to the Internet of Things.

The Internet of Things (IoT) made communication between people and objects easy. It helps to build smart cities, homes, manufacturing systems, health monitoring systems, etc., for mankind. The increased adoption of IoT applications enabled many smart devices on the Internet platform. These devices deployed across the globe may have varying computational and communication capabilities. It is a great challenge to manage IoT resources efficiently. Some of well-known protocols are defined to identify and access IoT resources locally in a real-world environment. Many authors have adopted the Distributed Hash Table (D.H.T.) based Peer to Peer (P2P) model for global and massive resource management. However, D.H.T. based solutions have many shortcomings and are not perfectly suitable for the IoT domain. In this paper, it has been proposed a novel Hybrid Centralized Peer to Peer (HCP2P) architecture for efficient resource discovery and access mechanism. The proposed solution builds a secure communication channel among trusted peer devices with the aid of an HCP2P server. The trusted devices can discover and access the required resource efficiently and securely with reduced load on the central server. The proposed HCP2P solutions are evaluated on both hardware prototypes and simulations. The proposed model gives almost constant resource registration, discovery, and access time. This evaluation showed that HCP2P architecture performance is superior to traditional DHT-based P2P architecture. Finally, the performance parameters of the proposed scheme are evaluated in terms of resource registration time, discovery time, and hop-count.