Tech Briefs: What drew you to this project?
Professor Patrick Mercier: I've been doing a lot of research in enabling new ways to make low-power communication between devices. This is useful for wearable applications, Internet of Things, applications, and so on. A lot of the best academic work on low-power wireless communication leverages custom modulation schemes, which means that you can't communicate with existing devices — you need specific radios to communicate with each other, which increases deployment costs.
What we were trying to do in this work is enable low-power communication, but using standards that are already popular, like Bluetooth and Wi-Fi. But unfortunately, if you build a Bluetooth or a Wi-Fi radio in the normal way, it's going to consume quite a lot of power. Even Bluetooth Low Energy (BLE), which has the words low energy in the title of the standard, is not actually very low power.
My colleague Dinesh Bharadia and I have worked on what is called backscatter communication, where we leverage existing transmissions that our phones, our access points, our routers, and so on are already making. And we're just piggybacking on top of those existing transmissions. What this allows us to do is communicate directly with these commercial devices at 1,000 times lower power than you could otherwise do — that can enable some really interesting new applications.
Tech Briefs: Regarding backscatter, aren’t all these transmissions at different frequencies?
Mercier: That's right. What we do is take an incident Bluetooth signal — it arrives at the antenna on the chip that we've developed. We receive that information and then we modulate the impedance seen at the antenna to re-radiate the signal in a format that's compatible with Wi-Fi. So, we do a format translation and a frequency translation, because Bluetooth and Wi-Fi are similar in frequencies, but not exactly the same. So, we frequency translate the signal to a Wi-Fi channel, and we reformat it into a format that's compatible with Wi-Fi.
Tech Briefs: You said you modulate the impedance. Could you explain that a little?
Mercier: Normally what happens with a radio transmitter is, you take a signal, amplify it to a large voltage, and send it to an antenna, which then radiates it into the electromagnetic space around it. That requires quite a bit of power, especially if you want to achieve a large communication distance. What we do is take the incident signal coming in from the existing Bluetooth transmission and if we don't terminate the antenna with the appropriate impedance, some of that incident signal is going to reflect back out into the environment. What we're doing is modulating the amount of reflection that's happening in a very specific manner and sequence — to make it reflect as if it were translated in frequency and now looks like a Wi-Fi signal.
Tech Briefs: So, is all of that translation and modulation powered by the incident signal?
Mercier: In this particular work, we're harvesting energy from incident mobile phone LTE signals at a slightly different frequency. So, the phone transmits LTE signals. We receive those, that powers up the tag, and then when the phone sends a Bluetooth packet, we frequency translate and reformat that to send back a Wi-Fi packet for reception by the Wi-Fi receiver in the phone.
In many ways this is very similar to standard RFID technology. It sends out a signal that powers the tag and then the tag does this backscatter modulation — this reflection modulation, if you will — to send the signal back. But in this case, we don't need a dedicated RFID system, we can just use our phone.
Tech Briefs: So, then do you have to install software on the phone?
Mercier: That's right; the phone does need some firmware modifications to enable this to work.
Tech Briefs: Why do you use both Bluetooth and Wi-Fi?
Mercier: It's a good question. The reason we have to do that is: Let's imagine we wanted to not use Bluetooth at all — we wanted to send Wi-Fi and have Wi-Fi come back. Unfortunately, most phones only have one Wi-Fi radio, and that radio is not capable of both transmitting information and receiving it at the same time. That's called a full duplex system and no commercial phones have that capability for Wi-Fi.
But they do have a separate Bluetooth radio in most cases. So, the Bluetooth radio can send a signal and we can receive a Wi-Fi signal on a different radio, which is also on the phone.
Tech Briefs: Do you have any thoughts about possible applications?
Mercier: I think any application that currently uses RFID could potentially use this type of technology. Instead of having to buy a dedicated RFID reader system, which is usually big and bulky and not that convenient, now you can enable that functionality with just your cell phone.
An application where this might be relevant is, for example, a grocery store, where you put your purchases into your cart, and you just walk out of the store. If each food item has one of these tags on it, it can talk to your phone and it will know exactly what items are in your cart, so you don't have to physically go through the checkout process — your phone just knows all of this information and charges you as you leave the store.
We could then keep track of how old those food items are as they're in the fridge. Perhaps we could even put temperature sensors on the food so that we can ensure the temperature is well-regulated and the food hasn't spoiled.
It would also be very useful anywhere that requires very, very low-power wireless communications, like wearable devices.
Tech Briefs: I guess the transmission distances are limited.
Mercier: That’s a good point. If we want to power the tag from the LTE signals, the distance we can get is about one meter or so. If, on the other hand, we can install a very small battery on the system, then that one-meter range goes to 10s of meters, maybe even 100 meters depending on the scenario because we wouldn’t have to rely on the harvested energy.
The key bottleneck in terms of range is the energy harvesting from the LTE signals. If we get rid of that requirement, then the backscatter range can be significantly larger.
Tech Briefs: Would the battery be part of the tag?
Mercier: That's right.
Tech Briefs: Wouldn’t that make it heavier and more expensive?
Mercier: Yes, but the nice thing about RFID systems is they're very cheap. They cost about a penny, and you could put them on every food product at the grocery store. If you then put batteries on all these things, it obviously increases the cost and the environmental impact.
That would be OK for a wearable application — you could afford a rechargeable battery. However, instead of every couple of days, with this system you could probably recharge it once a month because of the power savings. So, it really depends on the application.
Tech Briefs: The very first time I ever heard of RFID is with the E-ZPass tag on my car.
Mercier: In that case, they have large RFID readers on the tops of the toll booths and as you pass through, it reads your tag. The technology works great in that application where you buy a dedicated RFID reader and put it on a toll station. But if you want to enable that functionality in your home, you don't want to have to buy a $1,000 RFID reader.
Tech Briefs: Where are you now in terms of commercializing this?
Mercier: There's a lot of interest around this technology. We have a number of different companies that we're talking to.
We're also looking at perhaps spinning it off into our own startup company. We're looking for the right partners to do that with. So, I think the future is exciting for this technology, and we'll see what happens.
Tech Briefs: Are you thinking about food as a first application?
Mercier: Not necessarily. I think that makes sense for the self-powered applications, where you need the cost to be very low. But for applications like wearables where you have a battery, this still really makes a difference. So, think about going from a couple of days of battery life to a month — that really changes the use case for your wearable, especially for things like sleep tracking and so on.
Tech Briefs: How would your technology extend the battery life?
Mercier: We don't need to harvest energy to enable backscatter modulation. We can completely get rid of the LTE path and the system will still work as long as it has power from some source, whether it's the battery or the energy harvester from the LTE signals — either is OK.
Tech Briefs: So, it would extend the battery life because your device consumes so little power. Is that it?
Mercier: Yes — just to give you some numbers, a Wi-Fi transmitter consumes about 100 milliwatts of power, while our backscatter modulator consumes 10 microwatts. So, it's multiple orders of magnitude lower power, which can really improve battery life. Even BLE is 1 to 10 milliwatts. Again, we're 10 microwatts, so we're still multiple orders of magnitude lower.
Tech Briefs: Is there anything you want to add?
Mercier: Our ability to work within the confines of existing standards and protocols that are already in our phones is really interesting. The ability to use them to communicate at much lower power will enable a lot of new applications. So, I'm excited to keep working on this.