Comparative Analysis of Opp-IoT Routing Protocols Under Varying Node Density

The Internet of Things increasingly operates in environments where stable, infrastructure-based connectivity cannot be assumed: environmental sensing, vehicular and drone networks, industrial automation, healthcare wearables, and disaster-response scenarios among them. Opportunistic IoT networks (Opp-IoT) apply the store-carry-forward paradigm of opportunistic networking to such settings, accepting delay as a structural feature in exchange for delivery where conventional schemes fail. Despite a rich routing-protocol literature, the absence of standardised, density-aware comparison has produced inconsistent guidance for protocol selection — a particular problem for IoT, where deployments span densities from below one device per square kilometre in remote sensing to several hundred in industrial and urban contexts. This paper presents a controlled simulation study of nine routing protocols from the six dominant Opp-IoT families: flooding-, forwarding-, probability-, knowledge-, and social-relationship-based, plus hybrid. Using the ONE simulator on a 4500 × 3400 m² scenario over ten hours, node density was varied from 50 to 500 nodes and evaluated against eight Quality-of-Service metrics. Spray-and-Wait and FRESH dominate across the density spectrum with complementary strengths: Spray-and-Wait optimises delivery rate and overhead, FRESH minimises drops, latency, and buffer occupancy. For IoT deployments, where reliability, latency, and resource economy are paramount, FRESH is recommended as the default routing protocol.