IoT protocols and algorithms: foundations for optimal choices in connected projects

The Internet of Things (IoT) is growing rapidly and transforming many areas such as healthcare, home automation, automotive and industry. This expansion is based on the interconnection of heterogeneous devices with limited capabilities (energy, memory, computing) via adapted protocols and algorithms. This paper highlights the state of the art in IoT protocols and algorithms used in IoT gateways, in order to better understand the challenges associated with network heterogeneity and varied application requirements. IoT protocols can be classified into three main families: (i) communication protocols (RFID, NFC, BLE, Wi-Fi, ZigBee, LoRa, Sigfox, NB-IoT, LTE-M, EC-GSM, etc.) according to their range and energy consumption; (ii) application protocols (MQTT, XMPP, CoAP, WebSocket) developed specifically for constrained environments; and (iii) other essential protocols, including security protocols (TLS, DTLS) and service discovery protocols (mDNS, DNS-SD). The choice of protocol depends heavily on the type of application, the required throughput and energy constraints. On the algorithmic side, research focuses mainly on optimal resource allocation and energy optimisation in IoT gateways. Various studies have proposed: · user-centric network selection algorithms (Abdellatif et al., Liu et al.), · mechanisms for assigning services to interfaces (Angelakis et al.), · and energy optimisation solutions with adaptive selection of wireless technologies (Di Mu et al.). The aim of our article is to provide an in-depth overview of the state of the art in protocols and algorithms used in connected objects, using summary tables. The growing integration of the IoT into our daily lives requires a better understanding of these technologies in order to use them effectively and improve them in line with real needs.