Duplicate-Free Gateway Forwarding for LoRaWAN Uplink Using Network Coding

The Internet of Things (IoT) is being hailed as the next phase in the evolution of the Internet. IoT promises to connect billions of devices to the Internet to enable many new applications. While ambitious, IoT brings about many new challenges. Scalability is considered one of the most demanding challenges. Many access technologies have been proposed to provide connectivity to IoT devices; with Long Range Wide Area Network (LoRaWAN) being a strong contender for the Low-Power Wide Area Network (LPWAN) applications. In a LoRaWAN network, multiple gateways are used to forward packets transmitted by end nodes. Since a transmission can be received and forwarded by multiple gateways, duplicate packets can be forwarded to the network server. This obviously increases the amount of traffic crossing the network and flowing to the network server. The network server is also required to implement a deduplication logic to filter redundant packets. In this study, we investigate the application of inter-flow random linear network coding to improve the efficiency and scalability of LoRaWAN systems. In particular, random linear coding is applied to traffic moving from LoRaWAN gateway to the network server to remove duplicate packets generated by gateways. We hypothesize that this should result in less traffic crossing the LoRaWAN backend and eliminate the need for deduplication at the network server. In turn, this is also expected to permit a LoRaWAN system to support more gateways and consequently more end nodes. The study will include the development of a distributed coding scheme where encoding takes place at the gateways while decoding takes place at the network server. The design will follow a mathematical approach utilizing results from both information and coding theories. A simulation environment will be developed to verify and tune the parameters of the proposed scheme. In addition, the proposed scheme will be implemented on a real test-bed to confirm its applicability and gain insights into any practical issues. The study will provide an insight into the issues resulting from incorporating network coding into LoRaWAN systems. In addition to the final report, the project is expected to yield several publications and - potentially - a patent. It would also result in a platform that can be used for further research into the LoRaWAN technology and IoT systems.