Internet of Things, typically defined as modern solutions to collect real-world data at scale, is still in its infancy. There are a vast number of solutions available, with different use cases in mind and different pros and cons.
To deliver on the promise of a true Internet of Things, the complete solution needs to be built with security and scalability at its core. Scalability in this blog refers to a high density of sensors in a deployment. For example, tens of thousands of sensors in a single building and millions across a campus.
Wiring even a handful of sensors can be costly and limits where sensors can be located. Wireless sensors enable cost-effective and scalable IoT deployments. Before deploying a sensor solution, you need to understand the key trade-offs between sensor range and scalability.
Today there are only three wireless technologies that can be viewed as true standards:
All other wireless technologies are proprietary networks with limitations like a single vendor for key parts of the system, limited regional support, etc.
The benefit of using a proprietary technology is that it can be tailored to e.g. small wireless sensors, instead of being generic, with support for many different types of communication needs. This article focuses on these proprietary technologies as a way to achieve scale.
A single long range message takes 1000x longer to send compared to systems designed for scalability
The lower system can handle 1000 sensors in the same time the long range system can handle one sensor.
Scalable sensor system = many sensors transmit in bursts
Length of long-range messages vs. length of messages for scalable IOT systems
When deciding which wireless technology to use, the wireless range is a key parameter to consider. At first glance, it seems logical that the more range you can get, the better. The choice is, however, not that simple. There are important trade-offs: range and scalability drive in opposite directions.
There are many factors that affect the wireless range. Given the strictly defined rules for RF emission levels, the main practical parameter available to a system designer that affects range is the data rate, i.e. the speed of communication.
The way radio communication reaches further is to “speak slower”. This makes it easier for the receiver to understand the message at longer distance, where naturally the signal is weaker. The downside is that it takes a longer time to send the information.
For example, using a long range wireless system with 1 kilobit per second will use 1000x longer time to get the same message transmitted compared to a Bluetooth device operating at 1 megabit per second. This means that a high-speed receiver can get data from 1000 different sensors in the same time as a long-range receiver can get a single message from one sensor.
From this simple comparison, it is obvious that low data-rates and long-range are very strong limitations on scaling the number of sensors1. If the geographical aspect is also put on top, the scaling challenge gets exponentially worse. A 1000x difference in data-rate translates to
around 32x the range2.
Consider the following simple example:
This is the simple, physical reason why long-range systems cannot scale to the needs of modern IoT solutions.
1Simplified model as high rates use more bandwidth. Generally high rate systems are much more bandwidth efficient compared to long range / low rate systems.
21000x energy needed, i.e. 30 db. 6 db approx. 2x range line of sight gives 2^5=32x range
Consider this result in a practical application in a building:
If you have a long-range system that e.g. can cover a building with a single gateway, it will not be able to scale to more than a few 10s, or maybe up to 100+ sensors in the whole building.
A sensor solution built for scale will logically require more gateways, but can then easily be deployed to cover tens of thousands of sensors. This capability is needed to support use cases like temperature monitoring, door/window monitoring, feedback buttons, etc. Once the infrastructure is in place, it is vital that new applications can be added without the need to change or replace the infrastructure (gateways).
Long-range wireless systems are built for applications with very few sensors in a very large area. For use cases in e.g. building automation, where you easily find hundreds and even thousands of sensors in a single building, the higher data-rate on sensors is needed to be able to support the number of sensors. The downside is the shorter range, which means more receivers/gateways.
This is a known trade-off. It explains why cellular networks have very dense base station deployments in urban areas, with support for many devices and high data-rates and have low-density base station deployments in rural areas to support only low data rates.
There is a common misconception that long-range systems, like LoRa, use spread spectrum techniques to achieve high data rates and long range.
The simple fact is that spread spectrum does not itself give longer range. Long range is achieved by using low data rates. It is physically not possible to have high data rate and long range, no matter what technology is used. The relation between data rate and range is defined in signal processing theory and different techniques are used to come as close to the theoretical limit as possible.
Spread spectrum is a well known technique to get closer to the theoretical limit for very low data rates. For higher data rate systems, however, spread spectrum has typically limited value and drawbacks such as excessive bandwidth usage often make it counterproductive.
In long-range systems like Lora, the technology supports extremely low data-rates, down to as low as 18 bits per second. This naturally gives very long range performance, but also places very strong limitations on the number of sensors in a wide area.
In practical deployments, LoRa cannot support more than approximately 200 sensors in a wide area. This limitation will typically not be seen during test and piloting, but will become a critical issue in real-world deployments. We recommend you investigate this issue before starting a LoRa implementation.
A communication system can in many ways be compared to a transportation system:
Consider a highway with two lanes in each direction. It can handle a lot of cars, up to a certain point. When this point is reached, you get congestion and the traffic gets stuck and very few cars come through.
The same is the case for a wireless system. If your road, i.e. your radio channel, gets congested, very little data will get through. You can always build a new road in parallel. Or in our analogy, use another RF channel, but as long as you do not use the roads efficiently, you will very quickly run into congestion issues again.
As radio spectrum is in its nature shared by everyone, using a wireless sensor technology that does not make efficient use of the spectrum will not only be a problem for the inefficient technology, but for everyone else using the same spectrum!
If traffic is stuck, no one gets through, no matter what type of car they use.
This is the simple reason why cellular operators invest heavily in radio technologies to improve spectrum efficiency - to allow more wireless devices to share the limited radio spectrum available.
Disruptive Technologies have designed extremely small and energy efficient sensors, built to
scale to billions of sensors.
Due to the very limited energy available in the miniature sensors, the available standard radio technologies (cellular, wifi, Bluetooth) were not a direct fit.
Inspired by cellular network technology, Disruptive Technologies have designed a tailored protocol for miniature sensors called Secure Data ShotTM (SDS). SDS provides end-to-end encryption and seamless roaming across base stations, in SDS called Cloud Connectors.
The data rate chosen has a range similar to wifi, as the deployment of wifi is well known and
proven highly scalable.
The SDS protocol is designed to allow up to one million sensors to operate in a small, geographical area. The SDS protocol is designed for highly efficient radio spectrum usage to enable it to share spectrum with other sensor technologies that also operate with fair spectrum usage.
This article is also available as a white paper in PDF.
There are key trade-off between wireless range and scalability.
Radio spectrum is a limited, shared resource and efficient use has always been the key to
successful wireless systems.
|Secure Data ShotTM (SDS)||Long range solutions (e.g. LoRa)|
|Wireless range||up to 25 m indoor and up to 1 km outdoor||Kilometers (for few sensors)|
|Density, number of sensors in a small area||10s of thousands per cloud connector||Up to 200, issues reported with 100 sensors per building|
|Number of gateways||Higher||Lower|
|Bandwidth efficiency||High||Low to very low|
|Applications||Best for high data collection, high
density applications, such as:
|Best for low data collection, long
range applications, such as:
Follow this blog for a peek of our future!
* By subscribing to our newsletter, you agree to receive digital communications. You may withdraw this consent at any time.
33-40% of the world’s food is lost or wasted annually at some point along the production spectrum: from farm..
Whether you’re equipping an office building with sensors to understand usage or implementing remote..
Technology is making food spoilage a thing of the past, protecting retailers from profit loss, ensuring food..