The concept of body-area networks (BANs) - in which sensors that collect and wirelessly transmit data from human bodies for apps like health monitoring ?has typically been hampered by power consumption issues. Researchers at Qualcomm claim they've come up with an algorithm to fix that.
According to EETimes, Qualcomm's healthcare R&D team has developed an app-layer technique called a "Compressed Sensing" algorithm that reduces the amount of data acquired and sent via a BAN.
The idea, said Qualcomm engineer Harinath Garudadri, is to selectively sample a pulse oximeter, which results in gathering less data and, in turn, lowering power for both the sensor and the wireless network "by as much as a factor of 30". The technique is suited for health metrics that don't change very rapidly, like blood pressure and heart rate.
For things like electroencephalogram (ECG) sensors, which require richer data, Qualcomm used pre-coded signals, resulting in 99% accuracy despite a packet loss of 20%. Garudadri told the EE Times that Qualcomm's techniques could lower sensor power consumption down to "tens of micro-Amps".