Abstract
Decentralized storage has gained popularity due to its superior fault tolerance, scalability, privacy, and security compared to traditional cloud storage solutions. However, the untrusted and potentially unreliable nature of storage nodes in Internet of Things (IoT) and edge cloud settings necessitates careful selection of nodes for file storage in open and heterogeneous distributed solutions. We introduce a new trust management system for distributed storage networks that evaluates the performance of nodes in serving files to improve file placement decisions. We consider capacity and reliability as the most important parameters influencing a node's trustworthiness. Capacity refers to the available resources, while reliability is assessed based on node performance in responding to file requests and maintaining file integrity. We present a methodology to combine these two parameters into a joint trust score for node selection. Our experiments show that this trust model can enhance file placement decisions in distributed storage networks by providing more accurate estimations of the nodes' trustworthiness than the existing file placement algorithms.