On Spatial Diversity for LoRaWAN: Experimental Evaluation of Performance of a Dual-Gateway Network With and Without Downlink

Abstract

The Low Power Wide Area Networks (LPWANs) is today seen as one of the key connectivity enablers for the massive Internet-of-Things (IoT) applications. In this paper, we experimentally study the performance of a LoRaWAN network composed of multiple end devices and two independentlyoperating gateways (GW) for the case of uplink-only and uplinkdownlink traffic. Our results demonstrate that even though the spatial diversity due to the use of multiple GWs improves the performance of the network for the packet error rate, it still does not guarantee low packet loss under heavy uplink network traffic. Nonetheless, when the downlink traffic is present, a second GW operating in the uplink-only mode can dramatically increase the uplink packets delivery rate in the network. Therefore, we consider that spatial decoupling of receiver and transmitters, or addition of uplink-only GWs may become an efficient way to improve the performance of LoRaWAN networks with downlink traffic. Even though the limited laboratory environment might have affected the results of our measurements, we expect that many of the observed trends will also stand for the real-life LoRaWAN deployments. For this reason, the presented results might be interesting both for analysts and practitioners working in the field.

Performance of Narrow Band Wide Area Networks with Gateway Diversity

“…The study in [19] illustrates the benefit of GW diversity in terms of improved PER with only two gateways measured in an indoor anechoic chamber. This study with realistic system parameters such as outdoor NLOS conditions with shadow fading shows that a coverage area of 1.44 km 2 can be achieved with a TP-99 PSR ≥ 86%.…”

LoRaWAN for Smart Campus: Deployment and Long-Term Operation Analysis

“… The period between sequential transmissions exceeds the duration of transmission back-off due to duty cycle restrictions, as this is required by the EU frequency regulations in force (see Table 2 ) [ 49 ]. No acknowledgments, adaptive data rate, or regular downlink transmissions are present, since these are neither obligatory nor required for our application, thus implying “best-effort” delivery on the one hand, and may compromise the uplink delivery probability due to half-duplex architecture of the gateway [ 59 ]. The start-up time for each node is random, as no specific procedure for powering them up or connecting them to the network is implied.…”

“…No acknowledgments, adaptive data rate, or regular downlink transmissions are present, since these are neither obligatory nor required for our application, thus implying “best-effort” delivery on the one hand, and may compromise the uplink delivery probability due to half-duplex architecture of the gateway [ 59 ].…”