NB-IoT is considered a licensed low power wide area networks (LPWANs) technology supported by your local telecom operator. That means each device requires a SIM card and monthly or annual payments to your operator just like your cell phone. The benefit is that you don’t need to manage the infrastructure; you do not need to install your own base stations.
The key advantage of NB-IoT is the protocol is synchronous and designed to optimize the spectral usage and throughput of the network. This optimization for spectral utilization comes at the cost of compromised battery life and recurring costs (monthly or yearly).
Unlicensed LPWANs, such as LoRaWAN® and Weightless™ are optimal for longer battery lifetime. If sensor data is small and infrequent (one or twice a day), LoRaWAN could be the optimal choice being an asynchronous protocol. If data is larger and data transmissions must be acknowledged then Weightless stands alone as the option for a synchronous protocol for private networks.
The effect of asynchronous versus synchronous protocols has significant impact on the battery lifetime of sensors.
Semtech conducted a comparison using the T-Mobile NB-IoT network available in the U.S. The NB-IoT sensor consistently took more than 20 seconds of active time to negotiate a slot to communicate an 11-byte packet. The average current consumption over this 20 second period was 40mA. In comparison, sending the same 11-byte packet over LoRaWAN required an active time of only 1.6 seconds, with an average active current consumption of 6.4mA. This translates into greater than 50 times advantage in battery lifetime for LoRaWAN.
Take an example application: wireless, battery powered, pushbutton. A LoRaWAN-enabled pushbutton and an NB-IoT pushbutton each were equipped with a 600 maH battery. The LoRaWAN device could support roughly 70,000 button presses on a single battery, while the NB-IoT button could handle only about 2,000 button presses on a single battery. The difference is quite drastic.
When choosing a LPWAN technology, be sure to thoroughly review the application requirements. One size does not fit all.
Nearly all uplink messages are unacknowledged (you won’t know if the message was ever successfully delivered)
All gateways in range see all uplink traffic (not safe)
LoRaWAN requires an enormous amount of bandwidth (tough to scale, culprit for ISM band traffic, subject to interference from other LoRaWAN gateways)
1. Nearly all uplink messages are unacknowledged
LoRaWAN has 1% Duty Cycle Limit for both end devices and the gateway (A duty cycle is the fraction of one period in which a signal or system is active. A period is the time it takes for a signal to complete an on-and-off cycle) In order to support the 1% duty cycle limitation for the gateway, all uplink messages are unacknowledged and uncoordinated, LoRaWAN is considered a “pure-aloha” scheme.
What is pure-aloha: “The idea is applicable to systems in which uncoordinated users are competing for a single channel (shared resource). ALOHA permits users to transmit any time they feel like. Collisions will occur and therefore colliding frames will be destroyed. However, if feedback is available on the destruction, then users will be made aware of their frames have not been transmitted (received) successfully (1)”
“Such a network has about 18% efficiency. This means that 82% of packets are lost when a LoRaWAN network is fully utilized. Since most messages are unacknowledged, the end node does not know its message was missed”(2). The Things Network also confirms this statement by stating “the capacity for downlink messages is even lower than for uplink messages, so don’t waste it.”
2. All gateways in range see all uplink traffic (not safe)
No explanation required.
3. LoRaWAN requires an enormous amount of bandwidth
In wireless communication, the Hata Model for urban areas, also known as the Okumura–Hata model for being a developed version of the Okumura model, is the most widely used radio frequency propagation model for predicting the behaviour of cellular transmissions in built up areas. This model incorporates the graphical information from Okumura model and develops it further to realize the effects of diffraction, reflection and scattering caused by city structures. This model also has two more varieties for transmission in suburban areas and open areas. (source: Wikipedia)
The Hata Model simulation was conducted for Sigfox, LORA, and Weightless-P with the base station height set at 30m and the end devices heights set at 0.5m. The following simulation was conducted at Ubiik (hardware developers for Weightless-P) but we have checked their math and our team has confirmed the numbers are accurate and unbiased.
Let’s first take a look at the U.S Results (902-928MHz)
Now let’s take a look at the results in Europe (863-870MHz). The only difference is LORA is only able to use a smaller bandwidth.
One of 3GPP’s chief low-power, wide-area (LPWA) technologies under development is NB-IOT (narrowband IOT) . Many have been speculating over the differences between NB-IOT and the current LPWAN technologies in the unlicensed frequencies such as LORA, Weightless-P, Sigfox, RPMA. Some individuals have even gone as far as saying NB-IOT will be the death of LPWAN technologies. But that is likely not going to be the case as there will always be a huge difference in use-cases of licensed and unlicensed technologies. The best analogy is WiFi (unlicensed) vs 4G (licensed). The business models and use-cases built around WiFI and 4G are “night and day” .
NB-IOT may not be as robust as we are expecting it be. Check out the following features that are likely to be a slight let-down to NB-IoT enthusiasts
1.No full acknowledgement: By design (found in 3GPP Specification TR45.820) NB-IOT is planned to only acknowledge 50% of messages serviced by the wireless technology. This is due to limited downlink capacity. Unlicensed technologies like Weightless-P allows 100% full acknowledgement of every message. If every message is of high value, you will need to know if your messages are successfully sent/received via an acknowledgement.
2. Long Latency: Transmit packet aggregation from buffering of messages and data. NB IOT will not be able to support “real time” responses therefore not suitable for time sensitive applications.
3. IoT devices in the network will not be the priority. The licensed spectrum is EXPENSIVE. Ingenu mentioned “$4.6 billion in a recent auction for only 20 MHz of spectrum!” IoT traffic will always come second to high profit margin, cellular traffic.
4. Long battery Life? The actual battery life will remain unknown until the Cellular LPWA networks are commercially available.
5. Availability: NB IOT is a technology that will be ready a few years down the line.
6. Compatibility: NB-IOT will differ across regions and carriers. Huawei initially pushed for a clean slate NB-IOT technology that would not be backwards compatible with 4G etc. This actually makes a lot of sense as it would be eliminating a lot of the unnecessary overhead. But just as Huawei began making progress, Nokia and Ericsson began insisting on building upon the frameworks of LTE which means significantly more complexity and unnecessary overhead. Not a very nice foundation for such a huge project.