Cognitive Long-range: Towards efficient public communication infrastructure for Internet of Things

The most common Internet of Things (IoT) scenarios entail devices with limited energy resources and need to be connected to the Internet via wireless networks. This has driven the recent development of low-power wide-area networks (LPWANs) and the rise of the Long Range (LoRa) technology. The LoRa protocol has a simple modulation scheme that ensures low power consumption, high convergence, and resistance against interference. In most LPWAN technologies, several physical layer challenges arise, such as low data rates, spectral inefficiency, and increased interference. As a physical layer solution, the cognitive radio (CR) offers a possible way of resolving these challenges. CR allows wireless networks to operate without the need for a dedicated spectrum. Regarding the variety of end-user requirements, developing a public communication network that can support such diverse and heterogeneous applications is necessary to reduce the implementation costs than developing a dedicated communication network for each application. This paper proposes a Cognitive LoRa (C-LoRa) protocol that utilizes unlicensed and licensed frequencies as well as interference mitigation to improve the QoS of LoRa. To extract the priority list of traffic patterns, C-LoRa incorporates the Analytic Hierarchy Process (AHP) algorithm. The priority list enables real-time applications to receive optimal spectrum allocation. C-LoRa can be efficiently implemented as a public communication infrastructure for heterogeneous IoT devices. The addition of licensed channels improves the overall QoS and decreases the average waiting time in queues. The platform layer of C-LoRa consists of a cognitive engine that sends traffic priority lists to cognitive spectrum allocators. The IoT application servers are connected to the cloud platform layer via SNMP, HTTP, and other desired protocols. Access gateways equipped with a cognitive spectrum allocator are always connected to a power supply and serve as a transparent bridge to the cognitive engine at the platform layer, converting RF packets to IP packets and vice versa.