How to Connect to Edge Device Without using IP: Techniques and Applications in Modern Networks

IP addresses have always been the primary method for identifying and communicating with devices across networks. However, as computing shifts closer to the edge—where devices such as smart sensors, home appliances, and embedded systems operate in changing and often disconnected environments—this model begins to show its cracks. In many edge settings, devices may not have stable IP addresses, or any IP at all. This makes traditional, IP-based communication and discovery difficult, if not impossible. This paper explores how devices can connect and communicate without depending on IP. Instead of the conventional approach, we explore alternatives such as Bluetooth, hardware-based identifiers (like MAC addresses), and nameresolution techniques like Multicast DNS (mDNS) and Link-Local Multicast Name Resolution (LLMNR). These methods enable devices to communicate with each other, even in areas with low internet availability and where IP assignment is not possible. Each of these technologies has its own series of trade-offs. Some are optimized for short-range connections, while others are optimized for local networks. Some are platform-independent; others are more specialized. By comparing these methods side by side, we aim to offer a practical guide for engineers and researchers building edge systems. The goal is to design communication frameworks that are flexible, resilient, and better suited for the increasingly decentralized world of edge computing.