Ultra Narrow Band: principles & performances
For our French readers, we recommend this white paper that clearly explains the principles and the performances of the Ultra Narrow Band.
Beyond all marketing aspects, a bit a physic is often meaningful to have a deep understanding of the technology.
For non-french readers, here are some key takeways
Why LPWAN will exist?
The Internet of Things (IoT) is a recent component of communication systems. Its objective is to connect physical devices, vehicles, buildings, and even animals to the Internet without direct human intervention to control them. To achieve this, the existing connectivity in computers and phones needs to be extended to practically any object by equipping them with a chip and a radio interface.
Initially, IoT services were limited to home automation, connecting devices such as thermostats, energy meters, lighting control systems, music distribution and control systems, and remote video streaming devices. Over time, the scope of applications has expanded to include all objects and even living beings, such as health monitoring, smart grids, smart cities, logistics, wildlife monitoring and tracking, home automation, and security. All these applications require dedicated base stations to establish a gateway between the communicating objects within a cell and the Internet bubble that handles the collection and processing of the gathered data. It is important to note that IoT is characterized by the small size of the data to be transmitted (often reduced to a few bytes) compared to “human” communications.
To limit the operational cost for operators (and maintain a reasonable price relative to the amount of data transmitted), the infrastructure needs to be as minimal as possible. This necessitates base stations (BS Base Station) covering the largest possible area. The commonly adopted approach to achieve this requirement is a star topology with long-range coverage, where devices communicate directly with the base station. There are several strategies to extend the range of wireless transmission:
- Increase transmission power: This solution is not easily acceptable due to the associated power consumption. Low power consumption is crucial to preserve the lifespan of battery-powered nodes, and the potential health impact of increased exposure to waves must also be considered.
- Design extremely sensitive receivers: However, the cost of such receivers would be higher. They could be deployed on a few base stations but not extensively on all nodes, which would prevent a downlink communication.
- Develop new transmission technologies: This option has been adopted by operators and the scientific community, leading to the concept of LPWAN (Low Power Wide Area Network). LPWAN networks aim to provide very wide coverage (up to several tens of kilometers in rural areas) while minimizing energy consumption for objects. Furthermore, LPWAN networks should be capable of handling small data packets (as the exchanged data is typically of small size) and sporadic transmission (the targeted applications do not require a continuous data stream but rather relevant data packets when necessary).
LPWAN networks generally utilize the ISM bands for transmissions as they are license-exempt, further reducing network costs and enabling the implementation of these new applications.
LPWAN networks have paved the way for new markets and new commercial operators. Thanks to their low infrastructure cost, new entrants have been able to launch their own transmission technologies and enter the market alongside traditional telecom operators. Initially, these operators focused on technologies such as NB-IoT (Narrow Band IoT). Among the new operators, we can mention SemTech and LoRa Alliance with LoRa (Long Range) technology, SigFox with UNB (Ultra Narrow Band), and Ingenu with RPMA (Random Phase Multiple Access). More recently, the Mioty protocol, based on packet fragmentation, has been proposed.
In practice, two main and diametrically opposed directions have been taken by the new operators to define new transmission technologies:
- Spectrum spreading (e.g., LoRa with CSS – Chirp Spread Spectrum or Ingenu with RPMA): Data is sent over a much wider frequency band than its baseband occupancy. A specific spreading code is used to encode a symbol. The diversity obtained in frequencies allows recovering a signal even if its Power Spectral Density (PSD) is lower than the noise floor. Specific patterns are searchedand enable significant decoding gains. Moreover, different codes with good correlation properties can be assigned to different simultaneously transmitting devices for multiple access to the transmission channel.
- Reduction of signal spectral occupancy (e.g., SigFox with UNB): Data is transmitted at a very low rate using a simple modulation scheme to ensure minimal spectrum occupancy. The advantage of this technique is that the perceived noise (after signal filtering) is reduced since it linearly depends on the signal bandwidth.
For both techniques, the required received power for a given quality of service is considerably reduced. Thus, with the same transmission power as traditional systems, greater range can be achieved without degrading performance.
UNB Principles & performances
In UNB, data is transmitted at a very low bitrate with a simple modulation scheme to ensure minimal spectrum occupancy. The advantage of this technique is that the perceived noise (after signal filtering) is reduced as it linearly depends on the signal bandwidth. With this techniques, for the same transmit power as for more conventional systems, the range is considerably increased.
For comparison, if you want to effectively convey a “short” message from one end to the other in a train station hall on the day of departure for vacation, you can try yelling, but the crowd noise of people talking (or trying to do the same thing as you) will prevent your companion from hearing you. So how do you do it? Use a whistle. This simple device will focus the energy of your breath on a single frequency, creating a significant “peak” of energy on that frequency. Even if other people are speaking on that frequency, it will not be significant. Your companion will be able to distinguish the whistle amidst the background noise. Now imagine that the frequency of the whistle is random and that intermittently you vary this signal (yes, yes, we’re talking about modulation): you have just created a complete system of UNB communication.
The main characteristics of these systems are:
- Very simple modulation: with fixed baud rate to minimize spectral occupancy (DSPK, …)
- High resistance to interference: the concentration of energy on one side, and the low perceived noise by the receiver (as it is linear compared to the signal width) provide very good link budgets.
- Long range: for the same reasons, the link budget allows for long range.
- Signal discretion: by nature, a UNB transmission occupies very little spectrum, making the a priori detection of the signal more difficult. It is like searching for a needle in a haystack.
- Very simple transmitter: unlike spread spectrum, a UNB transmitter is very simple and consists simply of a local oscillator (not very accurate) that is modulated.
- Low emission power: since the energy is concentrated in the spectrum, it is possible to transmit with very little energy.
- Predictability of autonomy: a key element for IoT markets, it is possible to have battery-operated devices that not only have a long autonomy but whose lifespan can be calculated. Indeed, the complete lack of synchronisation of the transmitters makes the emissions identical in terms of power consumption, regardless of network conditions.