Engineers Turn Human Body Into Bioacoustic Communications Channel

You meet Bob at a party—a friend of a friend. He's way drunk but let's slip that he's a procurement manager at a very large, very flush government contractor. Dude sucks, but all it takes is a wordless handshake to get his name, contact info, schedule, and LinkedIn profile downloaded into the iPhone stored in your front pocket. No wireless technology was involved; all it took to beam Bob's data into your phone was regular old human flesh.

Using the human body as a medium for connecting electronic devices may turn out to be practical, and, eventually, pretty normal. In the current issue of IEEE Computer, researchers at Georgia Institute of Technology describe just such a flesh-based system. It involves the transmission of low-frequency acoustic signals within or among human bodies and even between those bodies and their environments.

"Given the anticipated mass adoption of wearables, natural and convenient communication between devices has never been more important," the GIT group writes. "Each device requires frequent information exchange with other devices and the service provider. Most mobile and wearable devices currently use wireless communication technologies such as Bluetooth and Wi-Fi to transmit and receive data."

There are a couple of problems with wearables and wireless technology, according to the GIT engineers. First, the process of connecting devices across wireless devices is generally pretty ugly and slow. Often it requires human input to work at all. The second problem is just the wasted opportunity of having wearable devices directly in contact with human skin, and, thus, ideally suited to receiving signals from within the human body, but not really doing anything with that fact. It's as if there's this huge swath of bandwidth out there collecting dust.

Of course, the body is really composed of all sorts of different stuff with all sorts of different acoustic properties. Compared to the squishier parts of the body, bone is a natural conductor of sound waves. So, the GIT researchers used a bone transducer as the sender in their experiments and an ultra-low noise accelerometer as the receiver. When the transducer is positioned over an actual bone—such as in the wrist—we get the desired amount of coupling between the sender device and the body. Compared to more typical microelectromechanical (MEMs) microphones, the accelerometer acts as a contact mic, offering fine-tunability and necessary shielding from background noise.

The GIT group tested their system on a group of eight subjects. The experimental setup is as below, where the red dots indicate sensors and the blue dot indicates the position of the bone transducer.

There were several different sub-experiments involving signal transmission within a single human body, between two human bodies (via handshake), and between a human body and a wooden table. Generally, signal transmission was clear in all cases, but there was a lot of variation across transmission frequencies. Lower frequencies usually had better responses.

Muddying things a bit is that signal transmission also occurred when when subjects weren't actually touching. If two subjects held their hands close together without skin-to-skin contact, there was still a response because the bone transducer coupled to the air around it in addition to the actual bone below. The sensor on the other subject's body would then pick this up. This was even more extreme when the transducer was hooked up to the wooden table.

"We believe that the bone transducer is so powerful that it essentially turned the whole table into a huge speaker, which amplified the energy coupled to the air," the GIT group writes. "Therefore, to apply this technology in the future, it will be necessary to eliminate air coupling."

"One potential advantage of a body area network compared with wireless networks is that it is more secure because information is transmitted only when a device is attached to the user's body," the report continues. "If there is coupling between the transducer and air, however, it is still possible to eavesdrop on the information transmission without touching the user's body."

So, ensuring security will likely mean verifying that physical touch has indeed occurred between parties. This is reasonable enough given that the frequency response curves for touch-based transmission and otherwise are significantly different. In an additional experiment, the group demonstrated that this is theoretically possible, but that "more work is required to apply it in real-world scenarios."

In any case, this is an idea still in its infancy and we should probably treat the GIT experiments as proof-of-concept. In the meantime, our wearables will have to fight through wireless rush hour along with everything else.