A Novel Mathematical Model to Evaluate the Impact of Packet Retransmissions in LoRaWAN

Internet of Things (IoT) network enabling technologies are expected to support a broad range of applications with significant variations in application requirements. LoRaWAN is intended to address delay tolerant, low-data-rate applications such as agriculture, some of the smart city and smart environment applications, and inventory management systems while facilitating long range transmission of application nodes. Network lifetime, which can be broadly defined as the network operational duration fulfilling application objectives, is one of the most important key network performance indicators for IoT applications. In many real world sensor network applications, a nonrenewable battery is used as the energy source of the nodes and death of essential nodes to perform the meaningful network operation, which is marked by either the complete depletion of energy or unusable residual energy, determines the lifetime of a network. Typically, the furthest nodes in LoRa networks determine network lifetime as single-hop communication remains the trend. While packet retransmission helps recover lost packets and LoRa networks allow up to eight retransmissions, network lifetime can suffer significantly as nodes are required to transmit more packets. In order to strike a deal between expected network lifetime and number of packet retransmissions, it would be important to foresee the impacts of each retransmission in a network.