Towards Next-Generation Global IoT: Empowering Massive Connectivity with Harmonious Multi-Network Coexistence

 @inproceedings{Zhang2025AlphaWAN,
 title={Towards Next-Generation Global IoT: Empowering Massive Connectivity with Harmonious Multi-Network Coexistence},
 author={Ziyue Zhang, Xianjin Xia, Ruonan Li, and Yuanqing Zheng},
 booktitle={Proceedings of the ACM SIGCOMM 2025 Conference},
 year={2025},
 doi = {https://doi.org/10.1145/3718958.3750504}
 }
    

LoRaWAN offers a compelling solution for delivering cost-effective network access to millions of IoT devices worldwide. However, operators face challenges in scaling their services to meet the growing demands of IoT connections. Moreover, current LoRaWANs foster competition rather than cooperation among coexisting networks, resulting in substantial capacity degradation as network density increases. To identify the root causes limiting LoRaWAN scalability and to enable harmonious coexistence among network operators, this paper conducts an in-depth investigation of operational LoRaWANs. For the first time, our study reveals that the capacity degradation in LoRaWAN is not due to the traditionally believed issues (such as interference or packet collisions in the wireless medium) but rather a newly identified issue termed \textit{decoder contention problem}. This problem cannot be effectively resolved using conventional approaches and will significantly hinder the scaled deployment of LoRaWANs as a next-generation global infrastructure. Based on our new findings, we propose design principles that guide our exploration for effective strategies to address this emerging practical problem. We develop concrete deployable solutions to mitigate contention, optimize spectrum utilization, and promote spectrum sharing among network operators. Extensive evaluations demonstrate that our strategies effectively boost network capacity close to the theoretical bound, and support the coexistence of up to six networks with significant improvement in spectrum efficiency.

Our empirical studies on real-world operational LoRaWANs reveal substantial capacity degradation in two aspects:
(1) Limited network capacity with significant performance gaps. The achievable capacity of operational LoRaWANs is much lower than the theoretical maximum (e.g., less than 35%). Moreover, our investigations reveal that extra LoRaWAN gateways may not necessarily yield capacity improvements in practice.
(2) Capacity contention across networks. Since all LoRaWAN deployments operate in the same unlicensed spectrum, they must compete for shared communication resources. Our experiments with operational LoRaWANs reveal that the capacity limit (\ie, 16 concurrent users) not only constrains individual network concurrency but also caps the total concurrency across all coexisting networks. Consequently, each network acquires only a fraction of the limited capacity, which diminishes further as more networks coexist.

Existing studies primarily attribute such capacity limitations to packet collisions and interference in the shared wireless medium. However, our study indicates that even in the absence of packet collisions, LoRaWANs still experience significant capacity gaps. Our in-depth investigation identifies four primary factors that contribute to the practical capacity gaps:

1) Limited capability per gateway: A single gateway monitors multiple channels but can only receive a small portion of packets limited by decoder resources (e.g., 16 decoders per gateway). As incoming packets are processed in a First-Come-First-Served (FCFS) way, later-arriving packets are discarded when decoders are fully occupied.

Figure: The radio architecture and workflow of COTS gateway

2) Inefficiency with homogeneous reception: Operators typically configure their gateways with standard LoRaWAN channels. As a result, these gateways operating within the same spectrum observe identical packets in the same order, leading all gateways to receive the early packets while none captures the later ones.

Figure: Channel Plans of Standard LoRaWAN

3) Non-optimal operational strategy: LoRaWAN does not associate users with dedicated gateways. Instead, all gateways within range receive and forward a user's packets. While this strategy enhances coverage and redundancy, it can also introduce inefficiencies — some users may unnecessarily occupy decoder resources across multiple gateways, while others may be left without any decoders.

Figure: LoRaWAN architecture

4) Inefficient spectrum sharing: Current LoRaWANs lack mechanisms for spectrum sharing among network operators. In multi-network coexistence scenarios, a gateway operating in the shared spectrum detects packets not only from its own network but also from other coexisting networks. Although LoRaWAN packets include network identifiers, these identifiers cannot be accessed until the packets are successfully decoded. As a result, packets from all coexisting networks compete for and consume decoder resources at each gateway prior to packet reception or rejection decisions.

Figure: Packet reception pipeline

In summary, the capacity gaps observed in operational LoRaWANs are fundamentally caused by the Decoder Contention Problem as follows:
The practical capacity of a LoRaWAN is limited by the decoder resources of gateways. Suboptimal LoRaWAN operating strategies, such as homogeneous channel configurations across gateways and coexisting networks, lead to intensive contentions for gateway decoder resources among both intra- and inter-network users. These contentions are further intensified by LoRaWAN's unique characteristics, including operation in unlicensed ISM bands, long communication range, and the absence of user-gateway association. Specifically, because LoRaWAN allows any gateway within range to receive and forward packets from any user, some users unnecessarily consume decoder resources at multiple gateways, while others fail to access any available decoders. These factors collectively contribute to the practical capacity gaps of LoRaWANs.

Overview of our solutions

Strategy: Heterogeneous channel planning

Figure: Heterogeneous channel adoption

Strategy: Spectrum sharing with misaligned channel plans

Figure: Frequency misalignment

Figure: Spectrum sharing in AlphaWAN

AlphaWAN implementation

Figure: System architecture of AlphaWAN

Real-world performance

Figure: Capacity gain (single network)

Figure: Spectrum gain (Multi-network coexistence)

Template for this Website