Skip to main content
eScholarship
Open Access Publications from the University of California

Transmission Power Management for Wireless Health Applications

  • Author(s): Amini, Navid
  • Advisor(s): SARRAFZADEH, MAJID
  • et al.
Abstract

The proliferation of ubiquitous sensing devices along with advances in low power wireless communication technology have resulted in the extensive use of wireless body area networks (WBANs) as the building blocks of the emerging field of wireless health. In these battery-operated WBANs, the sensor devices are strategically placed in/on the human body and the short/mid/long wireless communications are conducted on/off the surface of the body. As the battery energy does not follow Moore's law, energy-efficiency is always one of the design challenges of wireless health-monitoring systems, impacting usability, security, and cost. The idea of transmission power control (TPC) is to automatically reduce the radio amplifier's output power when the transmission power is more than required. Reduced transmission power translates into more energy savings and reduced interference problems. TPC techniques have been used in abundance in cellular networks and wireless LANs. TPC schemes for WBANs, however, are still in their infancy. For example, current IEEE 802.15.4 specifications do not differentiate between mobile and static settings, thus leaving WBAN transmitters in the dark as to what transmission power level they should utilize.

In this dissertation, we have investigated the potential benefits and limitations of TPC as a means to extend the battery lifetime in WBANs at the first three abstraction levels. Physical and MAC layers' approach to TPC perform a local optimization, whereas network layer TPC is capable of a global optimization. At the network layer, we analytically solve an optimization problem whose solution determines an important parameter, i.e., energy-efficient cluster size, for a class of routing/MAC protocols in WBANs. Assuming that the routes are established in an energy-efficient manner, we then experimentally profile the 2.4 GHz on/off-body radio channel under several scenarios regarding mobility states and environments, and we showed that fixed transmission power either wastes energy or hinders reliability. Finally, we devote our attention to an ambulatory medical monitoring WBAN system, which is tied up with different characteristics in terms of mobility, periodicity, and `unforgivingness' of the wireless channel as a result of proximity to the ground as well as to human's body. The target ambulatory WBAN system encompasses a pair of wireless instrumented insoles (known as smart insoles) for gait data collection, plantar pressure monitoring, and gait analysis. We design a sensor-assisted TPC scheme that augments in-network information with information from built-in sensors. To this end, multiple mobility states are defined for the smart insoles and the mobility states are incorporated into transmission power control policies. Available sensor information is leveraged to detect the mobility states, based on which the TPC scheme switches strategies.

We validate this new idea of switching transmission power control strategies by implementing and evaluating the sensor-assisted scheme and comparing it against a frame-based TPC scheme, which adjusts the transmit power solely based on recent information about packet transmission successes and failures. Our testbed experiments involving mixed mobility scenarios show that our TPC scheme obtains up to 50% increase in the battery lifetime, enabling the smart insoles to be used in uncontrolled environments. Such an improvement in battery longevity (from 4.0 hours to 7.8 hours) is made by reducing the average energy consumed for communication of a single packet from 4.51 mJ/pkt to 2.27 mJ/pkt.

Although designed for the smart insoles as a severely energy-constrained device, the sensor-assisted TPC technique is readily deployable on a variety of today's commodity devices to make a connection between the sensing subsystem and the communication subsystem of such devices. In addition, as the underlying mobility state detection methods place relaxed requirements on how the device should be worn in terms of orientation and position, they can be used for a variety of purposes, such as improving the patient's compliance with medical treatments and therapies.

Main Content
Current View