In real world use they give great range at very low data rates. I looked at several youtube presentations and I'm still none the wiser as to whether they are fundamentally different from prior technologies.
I realise there is a trade off between bandwidth and range. Is the lack of similar products using older technology due to lack of demand for low data rate devices, or is there something truly new about these devices?
Well, it's rather different from other technologies certainly.
When designing a wireless system there is a set of variables that impose some tradeoffs to make.
Distance, data throughput, and transmission power.
- At constant data throughput, more range requires more transmission power.
- At constant transmission power, more data throughput implies less range.
If your energy budget is low, which affects not only the physical size of the sensor but also the manteinance costs associated with the frequency of battery replacements, your power must be low.
It works on shared spectrum, which means that it should coexist more or less peacefully with other applications. That's why the designers chose a spread spectrum technique: it's reasonably resistant to interference from other modulation techniques and farily robust even when the S/N ratio is extremely low. Of course, at the price of a very low data throughput.
And it has another rather unique aspect compared to other wireless network technologies. In order to reduce power consumption even more, a LoRaWAN network with class A devices is not the typical star shaped network with a master arbitrating communication with the slave devices or polling them. Instead, Class A devices (the typical sensor gathering data) decide when to communicate with the gateways. Either periodically when reporting regular measurements, for example, or whenever its software detects an unusual event that must be reported.
A class A LoRaWAN device is like a submarine in the middle of the ocean. Maybe it will send a situation report every few days, if any. Most of the time it won't be easy for it to receive unsolicited communications from the submarine base either. However, if the submarine detects something unusual it has means to send an unsolicited message to the submarine base and even wait for an answer.
The way the communication works makes it more secure against intentional than other techniques. Most wireless transmission techniques implement some sort of arbitration/flow control scheme which means that a device sending a message requires a clearance from the master station. That makes them highly vulnerable to jamming. A good example are cell phones. What happens if you jam an end device? You will see a "no network" error and it will be unable to transmit.
Imagine an alarm system that uses UMTS or LTE to send alarm events. A burglar turns on a jamming device and your alarm is suddenly unable to transmit an event. LoRaWAN is different. Most of the messages sent by class A devices will be unacknowledged messages. Maybe you are sending temperature readings every 30 minutes. You can afford to lose a routine message now and then, it's not a big issue.
But in order to send an unusual/important event, say, an alarm, the device will likely use acknowledged messages. It will send the event and wait for a confirmation. What happens if the burglar is using a jamming device at your home? Your alarm won't receive the confirmation message and it will resend the event. But the transmissions have been made without requiring a previous clearance, and the messages have been received by the base station ("gateway" in LoRaWAN parlance) which can certainly process it regardless of the jamming.
In order to prevent those messages to be received a burglar needs to jam the base station, not the end device. And jamming a base station has some issues of its own. First, a well deployed LoRaWAN network will have enough density of gateways so that more than one gateway will pick up the messages of an end device. A backend decides which of the gateways will acknowledge a message requiring confirmation, but you need to jam several gateways in order to make sure that the alarm messages won't be received by the network.
A jammed gateway is likely to attract some attention from the network operator, hence from the authorities. Moreover, at least some gateways will be placed at sensible locations (telecommunications relay stations, for example) and probably you can expect a prompt police response in case a kind of "attack" is detected.
The "lone wolf" device model itself has some interesting security properties. End devices offer a very small attack surface for mischief like the exploitation of security vulnerabilities. It's not only that they don't talk among themselves, but they won't even listen at all most of the time. Compare that to the constant chatter of other wireless technologies.
And, finally, before LoRa only public networks on exclusive spectrum could be considered for wide area deployments (where by "wide area" I mean something larger than a campus or). Technologies using shared spectrum (Bluetooth, Zigbee, WiFi) are meant for short range communications. LoRa, working on shared spectrum, can work over quite large distances (20 Km on open terrain, maybe 1 Km in urban areas). That is pretty unique when compared to the others!
So, well, I guess it's really different from prior open, shared spectrum technologies. And while it won't be able to stream CCTV video there are plenty of applications for which it is very well suited. Don't you think?
Home
Help
Search
Login
Register
