LoRa Mesh Networks: Innovations in Self-Healing, Multi-Application, and Distributed Communication for IoT and Disaster Response
- Sharon Rajendra Manmothe
- 20 hours ago
- 17 min read
LoRa mesh networks are a way for Long-Range (LoRa) devices to communicate over multiple nodes to extend their coverage area. Unlike traditional LoRaWAN networks, which use a star shape where devices only talk directly to a central gateway, LoRa mesh networks allow devices to forward messages for each other.
What is LoRa? It's a low-power wireless technology that can send signals over long distances, often kilometers.
Why use mesh networks with LoRa?
They solve coverage problems where signals might be blocked, like in non-line-of-sight (NLoS) conditions.
They allow communication by routing packets through many nodes.
They can create self-healing networks that adapt to changes.
They are low-cost and can be set up in a distributed way.
They are especially useful for disaster response and various Internet of Things (IoT) applications like smart cities and digital agriculture where reliable, long-range communication is needed without relying on existing costly infrastructure.
How LoRa Mesh Networks Work (General Principles): To make a LoRa mesh network work well, several issues need to be handled:
Local Synchronization:Â Nodes synchronize with their neighbors locally by adjusting when they transmit, based on when they receive a special "beacon" signal. This avoids needing a central, global time.
Collision Avoidance:Â Nodes choose their transmission times ("slots") without a central coordinator to prevent messages from clashing.
Routing:Â Nodes can find the best path to send messages to a destination without needing to know the entire network map.
Self-Healing:Â The network can adapt to changes, like nodes moving or failing.
Challenges in Building LoRa Mesh Networks:
Building reliable, large-scale LoRa mesh networks with low-cost devices is not easy due to several challenges:
Collisions:Â Multiple devices trying to transmit at the same time can cause messages to be lost.
Lack of Time Synchronization:Â Devices often don't have a shared, accurate clock.
Asymmetrical Links:Â The connection quality between two devices might be different in each direction due to varying transmit powers or hardware issues.
Time-Variant Channels:Â Wireless communication conditions can change a lot over time.
Low-End Microprocessors:Â Many LoRa devices use simple processors to save battery, which can make it hard to manage complex network tasks.
Intermittent Nodes:Â Devices might not always be active or available.
Specific Protocols and Implementations:
Proposed Slot-Based Protocol (Åžahin & Arslan):
This protocol is slot-based and does not need global time synchronization.
Each node synchronizes with its neighbors locally by adjusting its transmission time based on received beacon signals.
Nodes choose their slots to avoid collisions.
It helps find forwarding nodes on the shortest path without knowing the whole network.
It effectively reduces collision events and resolves hidden node problems (where a node can't "hear" two other nodes that are interfering with each other).
It has been implemented and tested with LoRa devices.
LoRaMesher Library and LoRaChat Middleware:
LoRaMesher is a C++ library that enables routing in LoRa mesh networks using a proactive distance-vector protocol. It uses FreeRTOS and RadioLib.
It supports two types of messages: data messages (routed between nodes) and routing messages (broadcast to update neighbor information).
LoRaChat is a middleware built on top of LoRaMesher that allows multiple distributed applications to run concurrently on LoRa mesh network nodes.
It introduces "application ports" to identify which application a message belongs to, and "message ports" to identify networks (LoRa mesh, Bluetooth, WiFi) for cross-network applications.
A MessageManager component handles multiplexing and demultiplexing of messages to and from different applications.
It enables bidirectional communication between the LoRa mesh network and the Internet, especially via MQTT.
Gateway nodes (nodes connected to the Internet) can share their "gateway" role with other nodes through routing table exchanges, allowing remote nodes to send data to the Internet.
ClusterDuck Protocol (CDP):
This is a software for LoRa radios to send, receive, and relay messages to a central "sink" node called a PapaDuck.
It's not LoRaWAN compliant and uses a "managed flooding" technique to reduce network congestion.
It's designed for telemetry gathering and two-way emergency communications.
Different node types ("ducks") include DuckLink (for sensor info), MamaDuck (relays messages), and PapaDuck (message sink, can connect to Internet).
The current implementation has had high error rates and bugs, particularly with its Bloom filter (used to prevent messages from endlessly circulating). It also had issues with processing corrupt data, which affected battery life and caused missed messages.
Hybrid Low-Power Wide-Area Mesh Network (Jiang et al.):
This approach combines LoRa for wide-area coverage (Long-Range Mesh Network - LRMN) with ANT (a low-power, short-range protocol) for dense deployments (Short-Range Star Network - SRSN).
It uses a centralized TDMA (Time-Division Multiple Access) scheduling algorithm managed by a "hub" node to prevent collisions and save energy by putting nodes to sleep.
It includes Adaptive LoRa Link (ALL)Â to optimize energy by adjusting transmit power based on signal strength.
It also features Adaptive Mode Switching (AMS)Â for ANT networks to balance energy consumption among hub nodes by rotating the role based on battery levels.
Comparison to LoRaWAN
LoRaWAN uses a star topology, meaning devices communicate directly with a gateway. This limits coverage and can degrade in conditions where signals are obstructed.
LoRa mesh networks, in contrast, use multi-hop routing to significantly extend communication range beyond a single gateway's reach.
LoRaWAN is often focused on one-way data (sensors to cloud) with limited downlink, while LoRa mesh can enable bidirectional communication and distributed applications directly on the nodes.
LoRa mesh networks can be more energy-efficient for large ranges by distributing power consumption across multiple nodes and using lower spreading factors over multiple hops, compared to LoRaWAN which might need higher power for long single-hop connections.
When we delve into the world of Long-Range (LoRa) communication, two prominent architectural approaches often emerge: LoRaWANÂ and LoRa mesh networks. While both leverage the fundamental LoRa radio technology, their underlying design philosophies, operational characteristics, and ideal applications diverge significantly. Understanding these distinctions is crucial for anyone looking to deploy robust and efficient Internet of Things (IoT) solutions over vast distances with minimal power consumption.
The Centralized Authority of LoRaWAN
LoRaWANÂ is a standardized network protocol that primarily operates on a star topology. Imagine a central hub, like a bicycle wheel's center, to which all the spokes connect directly. In this analogy, the hub represents a gateway, and the spokes are the individual LoRa devices, or "nodes". These nodes communicate their data exclusively to one or more gateways, which then act as bridges, forwarding the collected information to central network servers, often hosted in the cloud.
One of the defining characteristics of LoRaWAN is its unidirectional data flow, largely optimized for uplink communication, meaning data typically flows from the sensor nodes to the cloud. While a downlink capacity exists, allowing the network to send messages back to the devices, it is quite limited and rarely utilized in practice. This design simplifies the end-device operation, making it highly power-efficient, as devices can enter deep sleep modes when not actively transmitting or awaiting a brief downlink window. LoRaWAN even categorizes devices into different classes—Class A, B, and C—each with varying levels of two-way communication capability and power consumption, with Class A being the most power-efficient and Class C offering continuous listening for constant power sources.
However, this star-based architecture comes with inherent coverage limitations. The reach of a LoRaWAN network is directly constrained by the transmission range of its gateways. In environments with obstructions like buildings or terrain, known as non-line-of-sight (NLoS) conditions, the narrowband LoRa signals can suffer significant degradation due to fading, leading to reduced coverage or communication failures. Furthermore, deploying a large-scale LoRaWAN network can be costly, as gateways can range from hundreds to thousands of dollars, and they require a reliable internet connection, which might necessitate expensive cellular services in remote areas. From a reliability standpoint, LoRaWAN's asynchronous, ALOHA-based protocol can limit its scalability, as simultaneous transmissions from many devices can lead to packet collisions and reduced packet delivery rates, especially in dense deployments. Despite these challenges, LoRaWAN offers robust security with two layers of 128-bit AES encryption, securing both network connection and application-level data.
The Collaborative Nature of LoRa Mesh Networks
In contrast to LoRaWAN's centralized star, LoRa mesh networks embrace a multi-hop topology. Here, devices are not restricted to communicating directly with a central gateway; instead, they can relay messages to one another, effectively acting as repeaters to extend the network's reach. This fundamental difference allows LoRa mesh networks to significantly extend communication range and overcome signal obstructions in NLoS conditions by finding alternative paths around blockages. If one path is blocked or a node fails, the network can often adapt and find a new route, demonstrating a self-healing capability.
LoRa mesh networks are also typically designed for bidirectional communication, allowing not only data transmission from devices to the internet but also commands and information flow back to the devices. This enables more dynamic and interactive IoT applications. The distributed nature of mesh networks often translates to a lower overall deployment cost because the individual nodes are generally inexpensive, and the network can be built out incrementally without heavy reliance on costly central infrastructure. Specialized gateway nodes within a mesh network can connect to the Internet (e.g., via Wi-Fi), allowing messages to flow between the mesh and external networks, and vice-versa.
A key technical challenge in LoRa mesh networks is managing collisions and synchronization without a central coordinator. Researchers have proposed innovative solutions to address these, such as local time synchronization, where nodes adjust their timing based on received beacon signals from neighbors, eliminating the need for global time sources like GPS. Collision avoidance mechanisms allow nodes to intelligently choose specific time slots for transmission, or use protocols like Time-Division Multiple Access (TDMA), often coordinated by a central "hub" node, to schedule transmissions and allow nodes to sleep and save power when not active.
Furthermore, smart routing protocols allow nodes to determine the shortest path to a reference node (like a gateway) through "hopping numbers" without needing a full map of the entire network. Techniques like Adaptive LoRa Link (ALL) enable nodes to adjust their transmit power for optimal energy efficiency based on signal strength. In hybrid networks, Adaptive Mode Switching (AMS) can dynamically shift the role of "hub" node within short-range clusters based on battery levels to distribute energy consumption.
For practical implementation, open-source libraries like LoRaMesher provide a C++ framework for proactive distance-vector routing in mesh networks. This library handles routing tables, packet sending, reception, and can even manage reliable transfers of large data payloads by splitting them into smaller chunks. Building upon this, LoRaChat is a middleware that enables a single LoRa mesh node to run multiple distributed applications concurrently. It introduces "application ports" to identify which application a message belongs to, and "message ports" to facilitate communication across different networks (e.g., LoRa mesh, Bluetooth, Wi-Fi, MQTT). This allows for complex distributed applications that span from the LoRa mesh to internet-based services, often leveraging MQTT for bidirectional communication.
Another notable example is the ClusterDuck Protocol (CDP), an open-source software project specifically designed for emergency communications and telemetry. It employs a "managed flooding" technique where nodes relay messages to a central "PapaDuck" sink node and includes built-in encryption. While simple to deploy and low-cost, early implementations of CDP have shown high error rates and bugs, leading to lost messages and faster battery drain due to processing corrupted data. Challenges in building large-scale mesh networks include issues like perfect clock synchronization, asymmetrical communication links, and the limited processing power of low-cost microprocessors in LoRa devices. Implementing security measures like encryption can also reduce the available payload size, potentially impacting data transfer efficiency.
A Comparative Perspective
At their core, the major difference lies in their network architecture and the resulting implications for coverage, reliability, and application complexity. LoRaWAN, with its star topology, offers simplicity and excellent power efficiency for devices that primarily need to send small bursts of data to a central cloud service over a single hop. It is well-suited for traditional sensor monitoring applications where direct connectivity to a gateway is feasible and sporadic data transmission is sufficient.
LoRa mesh networks, by contrast, utilize a multi-hop topology, allowing devices to collaborate and extend the network's geographical footprint far beyond the range of a single gateway. This inherent multi-hop capability makes mesh networks more resilient to signal blockages and node failures, as messages can route around obstacles or find alternative paths. They are particularly advantageous in challenging environments like underground infrastructure or vast agricultural fields where direct line-of-sight to a gateway is not guaranteed. The shift towards bidirectional communication and the ability to host multiple applications directly on mesh nodes, as exemplified by LoRaChat, opens up new possibilities for distributed intelligence and edge computing, where devices can collaborate, process data locally, and even participate in tasks like federated learning, rather than sending all raw data to the cloud.
While LoRaWAN thrives on its standardized simplicity and mature ecosystem for basic IoT data collection, LoRa mesh networks are emerging as a powerful, flexible alternative for more complex, distributed applications that require extended range, enhanced reliability, and on-device intelligence, pushing the boundaries of what low-power, long-range wireless communication can achieve in dynamic and challenging environments.
Alright, let's explore the intricacies of Long-Range (LoRa) mesh networks, delving into the challenges they face and the exciting opportunities they present. It's a field bustling with innovation, aiming to push the boundaries of what low-power, long-range communication can achieve for the Internet of Things (IoT).
One of the significant hurdles when establishing these networks is the inherent complexity of building a large-scale, reliable mesh network in a distributed manner. Imagine trying to orchestrate a vast symphony without a conductor; keeping everything in perfect harmony becomes a monumental task. Key among these complexities are collision problems, where multiple devices might try to transmit at the same time, leading to lost data. This is compounded by the lack of global time synchronization among nodes, meaning devices might not always be "listening" when another is "speaking," resulting in missed packets. Even if nodes use local time synchronization methods, the inherent drift of crystal oscillators can still lead to time slot mismatches and lost data over time.
Furthermore, the real-world environment adds layers of difficulty. We often encounter asymmetrical communication links, where a signal might travel well in one direction but struggle in the reverse, perhaps due to non-identical transmit powers or hardware differences. The wireless channels themselves can be time-variant, constantly changing due to environmental factors. And because LoRa devices frequently use low-end microprocessors to conserve battery, they might experience "blind periods" when the processor is busy with other tasks, unable to send or receive data. This can be particularly problematic for gateway nodes that handle significant data loads, as processing can delay radio access and lead to missed messages.
While mesh networks are designed to extend coverage, non-line-of-sight (NLoS) conditions can still significantly degrade narrowband LoRa signals due to fading in multi-path channels. And although individual LoRa nodes are often inexpensive, the overall deployment cost can still be substantial, especially if gateways require costly internet backhaul like cellular connections in remote areas.
Security is another critical consideration. While LoRaWAN mandates robust AES 128-bit encryption, many LoRa mesh implementations don't yet include this, and adding it can reduce the available payload size for data, potentially impacting efficiency. There's also the challenge of detecting malicious modifications of packets by rogue nodes and ensuring the encryption burden isn't too heavy for these low-power devices.
From a performance perspective, LoRa is fundamentally designed for low data rate communication and often faces duty cycle limitations (e.g., 1%). This means communication-intensive applications can strain the network and impact the number of applications that can run concurrently. The lack of inherent packet delivery guarantees in LoRa necessitates additional protocols for reliable messaging, which in turn introduces more control messages, increasing traffic and potentially leading to out-of-order packet reception or scalability issues, especially at gateway nodes with limited memory. Moreover, message delivery in mesh networks can incur delays, making them unsuitable for applications with strict real-time requirements.
Many existing LoRa applications have been developed with a monolithic design, meaning the application code is ad hoc and tightly coupled to the specific communication needs, lacking a general networking layer. This makes it challenging to run multiple distributed applications concurrently on a single node, which requires a more sophisticated middleware to handle application identification, message multiplexing, and cross-network communication. Even well-known open-source projects like Meshtastic, while effective for their specific purposes, might use simpler flooding techniques that can lead to network congestion at a larger scale.
On the other hand, the opportunities presented by LoRa mesh networks are equally compelling and often directly address the aforementioned challenges. One of the most significant advantages is their ability to significantly extend communication range far beyond what a single gateway can cover, especially in environments with obstructions or NLoS conditions, thanks to their multi-hop capability. This multi-hop approach can also be more energy-efficient for long ranges (over 10 km) compared to single-hop LoRaWAN setups that require high spreading factors.
LoRa mesh networks also boast a self-healing capability, meaning they can adapt to changes in the network, finding alternative paths around blockages or failed nodes. This resilience is crucial for deployments in dynamic or unpredictable environments. They naturally support bidirectional communication, which opens the door for more dynamic and interactive IoT applications, moving beyond simple one-way data reporting. The distributed nature of mesh networks often translates to a lower overall deployment cost as it reduces the heavy reliance on expensive central gateway infrastructure.
A burgeoning area of opportunity is distributed intelligence and edge AI, where LoRa nodes can perform machine learning tasks directly on the device, sending only small, processed classification results instead of large amounts of raw sensor data. This saves crucial bandwidth and can lead to more autonomous local decision-making systems. The potential for collaborative distributed learning, such as federated learning, across these tiny nodes is also being explored.
Researchers are actively developing sophisticated solutions to overcome the technical challenges. Local time synchronization methods, where nodes adjust their timing based on neighboring beacon signals, can effectively mitigate the need for global time sources like GPS. Intelligent collision mitigation techniques, such as slot-based protocols and random slot choices, are designed to reduce interference and address the "hidden node problem". Smart routing protocols, like those that determine the shortest path to a reference node based on "hopping numbers," allow efficient data forwarding without requiring a full map of the entire network. Techniques like Adaptive LoRa Link (ALL) enable nodes to adjust their transmit power based on signal strength to optimize energy efficiency.
The development of hybrid networks is another promising direction, combining the wide-area coverage of LoRa with ultra-low power short-range protocols like ANT for dense deployments. Within these hybrid setups, Adaptive Mode Switching (AMS) can dynamically select hub nodes based on battery levels to distribute energy consumption evenly, and heuristic timers can detect and recover from hub node failures.
The growing ecosystem of open-source solutions, such as the LoRaMesher library for routing and the LoRaChat middleware for supporting multiple concurrent applications, is fostering rapid development and innovation. LoRaMesher, for instance, offers reliable transfer of large data payloads by automatically splitting them into smaller chunks. LoRaChat, in particular, uses "application ports" and "message ports" to enable complex distributed applications that can span from the LoRa mesh to internet-based services, often leveraging MQTT for bidirectional communication.
Finally, LoRa mesh networks hold immense promise for emergency communications and disaster response due to their low-cost hardware, ease of deployment, and encouraging battery life even in challenging scenarios. They can be rapidly set up in areas where traditional infrastructure is scarce or damaged, providing vital information flow. The continuous operation of hardware platforms in harsh environmental conditions has also demonstrated their robustness and durability. Moreover, these mesh networks can provide significantly better quality of service (QoS) compared to traditional single-hop LoRaWAN in large-area deployments.
In essence, while LoRa mesh networks are still maturing and face their share of technical puzzles, the collaborative efforts in research and open-source development are steadily paving the way for them to become a flexible, resilient, and cost-effective communication backbone for a new generation of smart, distributed IoT applications, especially in environments where traditional connectivity falls short.
Here's a blog post addressing the identified research gaps in LoRa mesh networks:
Unlocking the Future of Connectivity: Addressing the Gaps in LoRa Mesh Networks
Long-Range (LoRa) technology has emerged as a game-changer for Internet of Things (IoT) applications, especially in scenarios where traditional connectivity is limited or absent, such as remote areas or disaster zones. While LoRaWAN, with its star topology, is widely used, the concept of LoRa mesh networks promises to extend coverage, enable node-to-node communication, and unlock new possibilities for distributed intelligence. However, despite their immense potential, several critical research gaps currently impede their widespread and reliable deployment.
Understanding these gaps is crucial for researchers and developers aiming to build more robust, scalable, and versatile LoRa mesh solutions. Let's delve into the key challenges:
Building Reliable and Scalable Large-Scale Deployments:
A significant challenge lies in creating large-scale, reliable mesh networks in a distributed manner. Current implementations struggle with fundamental issues like collision problems due to concurrent transmissions, a lack of global time synchronization among nodes leading to missed packets, and asymmetrical communication links caused by varying transmit powers or hardware impairments.
Existing projects, such as Meshtastic, often rely on plain flooding techniques, which can cause network congestion as the scale increases, especially in complex communication scenarios.
While slot-based protocols are being proposed to mitigate collisions and synchronize nodes locally, the inefficiency of flooding entire schedule tables for Time Division Multiple Access (TDMA) in large networks and minimizing setup overhead remains a concern.
The literature also points to a notable absence of robust systems specifically using LoRa technology in the disaster response space.
Enabling Concurrent Applications on Resource-Constrained Nodes:
A major gap identified is the lack of solutions that demonstrate multiple applications with diverse communication patterns running simultaneously on LoRa mesh network nodes. Many traditional LoRaWAN applications are designed monolithically, making it difficult to host several distributed applications concurrently on a single device.
This problem is compounded by the limited hardware resources of tiny IoT nodes, including microcontroller memory, CPU, and energy budgets. Running multiple communication-intensive applications simultaneously on low-end microprocessors presents a significant management challenge.
There's also a recognized need for a balanced selection mechanism for gateway nodes in routing protocols like LoRaMesher to distribute the load evenly, preventing scalability issues due to resource limitations.
Ensuring Robust Security and Data Integrity in Low-Power Networks:
The burden of encryption on low-power LoRa devices and its impact on payload size and network performance needs more thorough assessment. While LoRaWAN mandates AES 128-bit encryption, many LoRa mesh implementations don't, and adding it can reduce the available payload size.
Future research needs to determine the feasibility of detecting malicious modifications of packets by rogue nodes in the network.
Moreover, ensuring data integrity and minimizing corruption is crucial, especially given the high error rates observed in some current implementations like the ClusterDuck Protocol. Methods to collect telemetry on corrupt packets without overwhelming the Micro-Control Unit (MCU) are also needed.
Integrating Advanced Functionality like Edge AI:
LoRa mesh networks hold promise for transforming IoT nodes into "smart nodes" capable of machine learning tasks. However, there is a gap in fully enabling and demonstrating networked machine learning applications or Edge AI as a service over the low-data rate LoRa mesh links.
Further investigation is required to identify the optimal trade-off between local computation and communication resources for distributed intelligence applications within a LoRa mesh network.
Proving Long-Term and Real-World Robustness in Dynamic Environments:
Much academic work on LoRa mesh networks relies on simulations, with a dearth of usable prototypes that can be easily deployed in real environments.
Developing experimental designs to test for resilience against nodes becoming unusable or unavailable (e.g., due to damage or battery depletion) in dynamic and unpredictable environments, like disaster scenarios, is a critical area for future work.
Continued evaluation of hardware durability and long-term stability in harsh environmental conditions is necessary to truly prove the robustness of these systems.
FAQs:
What is LoRa Wi-Fi network?
Hybrid Structure and Purpose
This architecture combines the strengths of LoRa (Long-Range), which operates in sub-GHz industrial, scientific, and medical (ISM) bands and provides long-range, low-power communication over several kilometers, with Wi-Fi, which operates in the 2.4 GHz band and offers higher bandwidth and Internet connectivity.
The primary goal is to achieve wide-area communication coverage and low power consumption for IoT devices while supporting data collection in both densely and sparsely deployed device scenarios.
This hybrid setup is crucial for applications where traditional connectivity is limited or absent, such as remote areas or disaster zones, allowing LoRa nodes to access Internet services.
Role of Gateway Nodes
Gateway nodes are central to this integration, acting as bridges between the LoRa mesh network and the Internet.
These gateway nodes are regular LoRa mesh network nodes that are also equipped with a Wi-Fi interface and an Internet connection.
Upon receiving a LoRa packet, a gateway node determines whether the message should be delivered to a local application on that node or forwarded to the Internet.
The presence and role of gateway nodes are disseminated through routing table exchanges within the LoRa mesh network, allowing other LoRa nodes to discover and select the nearest gateway for Internet communication.
Communication Flow and Protocols
The system supports bidirectional traffic between the LoRa mesh network and the Internet.
For communication to Internet services, MQTT (Message Queuing Telemetry Transport) is a commonly used protocol. Gateway nodes can act as MQTT clients, publishing data from LoRa nodes to an MQTT broker on the Internet and subscribing to topics to receive messages from Internet hosts for LoRa nodes.
An HTTP client can also be implemented on the gateway for unidirectional data transfer from the gateway to an Internet host, particularly for monitoring applications. However, an HTTP server on a resource-constrained IoT node is considered a "heavy task".
Middleware solutions like LoRaChat, built on top of the LoRaMesher library, facilitate this multi-interface application communication, including LoRa, Wi-Fi, and Bluetooth.
These applications can either reside entirely within the LoRa mesh network or span across both the LoRa mesh network and the Internet.
Enabling Concurrent Applications
The integration allows for multiple distributed applications to run concurrently on resource-constrained LoRa mesh network nodes.
Application ports and message ports are introduced within the middleware to identify and route messages to specific applications or across different network interfaces (LoRa mesh, Wi-Fi, Bluetooth).
Challenges and Limitations
Resource constraints: Running multiple communication-intensive applications concurrently on low-end microprocessors (common in IoT nodes) poses significant challenges regarding microcontroller memory, CPU, and energy budgets.
Scalability: Reliable messaging and supporting multiple concurrent applications can increase bandwidth utilization and queue sizes, potentially leading to scalability issues at gateway nodes due to limited hardware resources.
Packet overhead: Extending LoRa packet headers for application identification slightly reduces the maximum payload size, which can increase the total number of packets for large data transfers.
Packet loss and retransmission: Collisions in the LoRa mesh network can lead to packet losses. While reliable messaging mechanisms re-send lost packets, this can result in messages being received out-of-order and increased response times.
Encryption: While LoRaWAN mandates AES 128-bit encryption, its integration into LoRa mesh networks needs careful assessment as it can reduce the available payload size.