This blog article was created to give the reader insight in the practicability of LoRaWAN in a tracking use case in an urban area setting through the setup, observations, and conclusions of each test case, ending with a wrap-up that summarizes the significant findings of the experiments and a final conclusion that gives a broader perspective on the key influences on transmission in LoRaWAN in different scenarios.
What is LoRa and LoRaWAN
LoRa, short for “Long Range“, is a wireless data communication technology designed to enable low-power, long-range communication between devices in an IoT network. With its chirp spread spectrum modulation scheme, LoRa operates between 400 MHz and 950 MHz, depending on the region. In Europe, it is the 868 MHz band that is commonly used.
LoRaWAN stands for Long Range Wide Area Network. It is a protocol that organizes LoRa-enabled devices into a network. Messages from devices are sent to gateways, which then relay the messages to a central network server, mostly organized in a star-of-stars topology.
LoRaWAN finds use in applications like smart farming, environmental monitoring, and smart city projects, where the devices are typically stationary, battery-operated, and send small packets of data infrequently.
Key specifications and testing setup for LoRaWAN
In this section we go through some of the important specifications to later on understand the results of the field tests and learn fundamentals on what is important when usingto deploy a LoRaWAN.
RSSI: The signal strength indicator
RSSI, or Received Signal Strength Indicator, is a measure of the power level received by the antenna. It is a key metric for understanding the quality of a connection in a wireless network. Stronger signals (closer to 0 dBm) typically imply clearer data reception and lower chances of packet loss. While values closer to -125 dBm imply a worse signal strength where packet loss might occur.
Spreading factor: Balancing range and data rate
The Spreading Factor (SF) in LoRa modulation plays an important role, serving as a dial that network designers can turn to regulate the balance among communication range, data rate, and energy consumption. A higher SF value can boost the communication range, making signals more robust and capable of being detected even at very low signal-to-noise ratios. However, this comes with a trade-off: as SF increases, the data rate decreases, which means that it takes longer to transmit a given amount of data.
In our tests, we opted for an SF of 10. This choice was made to give a harmonious balance between achieving a respectable communication range and maintaining a practical data rate. With an SF of 10, we observed that data packets were reliably transmitted every 30 seconds – a frequency that suited our field tests and overall tracking use case quite well.
In contrast, when we experimented with an SF of 12, the data rate slowed significantly, with packets being transmitted at intervals ranging from 3 to 5 minutes with no significant gain in range. For our purposes, this lower and less predictable frequency was not practical.
Antenna specifications
The antenna’s gain, measured in dBi, plays a vital role in the range and reliability of LoRa transmissions. A higher gain antenna can focus radio waves more effectively, extending the transmission range.
Setting the stage for testing
To evaluate the tracking capabilities of LoRaWAN, we crafted a tailored and portable testing setup. Central to our equipment list were a “Wisgate Edge Pro“ Gateway and a “ELV-LW-GPS1“ GPS Tracker. Our Gateway, with specifications including a maximum transmit power of 27 dBm and a minimum receiver sensitivity of -139 dBm, came supplied with two 5dBi gain fiberglass antennas. The GPS Tracker, initially equipped with its standard antenna (Range > 500m according to its data sheet), was later adapted with an N-Type connection, allowing us to experiment with different antennas – specifically, a small LTE antenna with a 2dBi gain, and a high-performance fiberglass antenna with a 5dBi gain.
Powering this setup to be portable presented its own challenge, which we met by employing drill batteries. A 20V to 12V DC/DC converter made the Gateway compatible with these batteries, while the GPS Tracker was powered using another drill battery. To enable real-time data communication and visualization, we connected the Gateway to a computer via LAN in a point-to-point configuration. On the software side, ChirpStack served as our network server, with InfluxDB handling data storage duties. For visually mapping our data, we turned to Grafana and its GeoMapPlugin as well a simple react-leaflet app.
Our comprehensive setup was designed to not only test the basic functionality of LoRaWAN but also to explore how different variables – such as the desired range or the choice of antenna – could impact performance.
Urban vs rural environments: The effect on LoRaWAN range
To elucidate how urban structures and line-of-sight obstructions between the gateway and GPS Tracker affect transmission quality, we conducted a comparative analysis through a series of field tests.
Test case 1: Urban environment
The initial test unfolded in the urban locale of Pasing,Munich, with the gateway positioned 15 meters above the ground. Under these conditions, the maximum observed range was a mere 0.91 km.
Observations
In this densely populated urban environment, the signal had to navigate through a complex maze of buildings to reach the gateway. Despite employing a 5 dBi fiberglass antenna, signals were weakend drastically beyond the railway, indicating significant obstruction-related challenges.
Test case 2: Rural environment
Contrastingly, the second test was orchestrated near Langwied, a rural setting near Munich. Here, the gateway was positioned just 2 meters above ground, yet the transmission range impressively extended to 4.49 km.
Observations:
Positioned to an S-Bahn station and above a high voltage line, the initial RSSI values in this test were recorded between -44 to -47 dBm. While these are not poor signal strengths per se, they were noticeably weaker compared to the -33 to -37 dBm range observed in Test Case 1, suggesting possible interference from electromagnetic sources.
Comparative evaluation of both test cases
These experiments illustrate the impact that obstructing the line of sight has on transmission quality and reliability. In rural environments, where obstructions are minimal, the range was expansive at 4.49 km. In stark contrast, urban settings, rife with obstructions, saw this range plummet to less than 1 km.
Further Test Case 1 reveals a sharp deterioration in RSSI values after the signal passed through an underpass beyond the railways, plummeting into the -80 dBm range and continuing to degrade. This pattern led to an hypothesis: Could the magnetic interference from railways and high-voltage conductors further negatively impact the signal degradation?
This hypothesis found further support in Test Case 2. Despite its rural setting, the proximity of the gateway to an S-Bahn station and high-voltage lines appeared to influence the initial RSSI values negatively.
A consistent observation across both tests was the correlation between RSSI values and packet loss. Specifically, when the RSSI value dipped below -100 dBm, packet loss escalated dramatically, often resulting in 7 out of 8 packets being lost.
Key takeaway
LoRaWAN’s range and reliability are heavily influenced by both physical obstructions and potential electromagnetic interference, highlighting the importance of strategic gateway placement to optimize performance.
Importance of LoRa-Gateway height
In the next series of tests, we turned our focus to the height of the gateway. In the LoRa community, it is a widely held belief, backed by many papers and manuals, that elevating the gateway improves both the transmission range and reliability. To put this theory to the test, we contrasted two different setups: in Test Case 3, the gateway was placed 60 meters above ground on the Olympia Mountain in Munich, while in Test Case 4, it was positioned just 1 meter above ground, at the base of the mountain.
Test case 3: Gateway positioned 60 meters above ground
In Test Case 3, the setup yielded an impressive range of almost 5 km, aligning with the maximum range suggested in the Wisgate datasheet for urban areas. Though there were some packet losses, the overall transmission was notably more reliable compared to Test Cases 1. At the farthest measured point, the packet loss surged to 84% though, resulting in 5 out of 6 packets being lost, rendering further transmissions not possible.
Test case 4: Gateway positioned 1 meter above ground
For this test, the gateway was nonchalantly placed on a bench right next to the Olympia Mountain, and the same route was traversed again. This configuration yielded a much more modest range of 1.75 km. While this outperformed Test Case 1, it was significantly less than half the range achieved in Test Case 3.
Comparison of test cases 3 and 4
The stark contrast between these two test cases underscores the profound impact of gateway height on LoRaWAN’s performance. The elevated position in Test Case 3 – 60 meters above ground – demonstrated a boost in range. Remarkably, this elevated setup achieved a range comparable to that of rural areas, where a clear line of sight between the gateway and GPS Tracker were more easily maintained, even though the gateway was placed much lower in the rural setting.
On the other hand, Test Case 4 highlighted the limitations imposed by a low gateway placement. The potential range was more than halved compared to when the gateway was positioned at a higher elevation in Test Case 3.
Key insights:
With the findings from Test Cases 1, 3, and 4, we can confidently affirm that both obstructions in the line of sight and the height of the gateway are important factors for the transmission range and reliability of LoRaWAN systems. Elevated gateway placement can dramatically extend the viable range, while obstructions and lower placement can significantly impair performance.
The difference in antennas and their effect on range
With the understanding that height and line of sight significantly influence the communication range between IoT devices and the LoRa gateway, we shifted our focus to another factor: The influence of antenna quality on a device’s range and reliability within the LoRa network. To assess this, we conducted three tests in a similar location, measuring the maximum range with different antennas. The results of all three tests are summarized in Test Case 5.
Test case 5: Different antennas, same location
Test Case 5 offers insights into how a high dBi gain and quality antenna can amplify the signal range between the gateway and IoT devices. Although the additional distance achieved with the 5 dBi fiberglass antenna may seem marginal at first glance, it represents a whopping 75% improvement in range compared to the default antenna. This improvement is noticeable considering that the high dBi test had to contend with building obstructions, while the default test enjoyed a largely unobstructed line of sight.
While a superior antenna can significantly enhance range and reliability (mainly through reduced packet loss up to 1500 meters), it is not as impactful as maintaining line of sight or elevating the gateway. It’s also worth noting that the fiberglass antenna, measuring 2.5 cm in diameter and 60 cm in height, may be inconveniently large for some IoT applications.
Conclusion and Wrap Up
Overall, LoRaWAN emerges as an excellent choice for long-range transmission in rural and open areas – especially where the gateway can be positioned at heights exceeding 50 meters and a clear line of sight can be consistently maintained.
However, LoRaWAN faces challenges in urban environments. Even with gateway placement at heights over 50 meters, the typical range, as backed by multiple tests and reports, hovers between 2 km and 5 km. The most critical issue undermining LoRaWAN’s reliability in urban settings is the unpredictable packet loss and transmission inconsistency when line of sight is compromised or when the network faces interference from high-voltage lines or noise from machinery, especially in industrial settings.
That said, LoRaWAN can still serve as a reliable solution in urban contexts where IoT devices are stationary and the overall RSSI values remain higher than -80 dBm (the threshold at which heavy packet losses were first observed in our tests).
Upgrading to a better antenna can further extend range, but the gains here are relatively modest compared to the effects of line of sight and gateway elevation.
For tracking use cases in urban environments, achieving comprehensive coverage with LoRaWAN using a single gateway is challenging. Assuming such a feat is possible, the costs would likely be enormous. For instance, to ensure consistent signal coverage across Munich (which has an area of 310.71 km²), a Wisgate gateway would need to be installed approximately every 800 m². This would mean deploying around 311 gateways, translating to enormous costs.
Moreover, even if all these gateways could be positioned 60 meters above ground, the total number of required gateways would remain substantial, exceeding 70 units. This level of infrastructure is likely to be impractical for most deployments.
