Dynamic Spectrum Access
We are working on cognitive radio techniques for energy
efficient adaptive physical layer and enhanced MAC layer for low power ad-hoc
wireless networks to support QoS for range of applications such as health
monitoring, security, and multimedia.
A key component of cognitive radios is the ability to sense
and be aware of its channel condition and energy consumption. Information such as received signal strength,
packet error rate, energy consumption, and battery life must be available to
PHY and MAC layers. The wireless sensor networks in this research are equipped
with a low power signal processor to carry out localized processing for
distributed computing.
We address the challenge of increasing the throughput of a
wireless ad hoc network by introducing a multi-channel Medium Access Control
protocol to supplement 802.11. Traditionally, in an ad hoc wireless network,
every device operates on the same channel.
This guarantees that devices within the range of each other can always communicate. However, as the number of devices in an area
increases, the throughput share of each device decreases due to
interference. Often, however, several
channels are available. Mathematical
analysis and simulation have shown that our multi-channel MAC protocol can
significantly increase the throughput of an ad hoc network and improve energy
efficiency of the network.
The outcome of this research program will be the
architecture and algorithms for MAC and PHY layer of a cognitive wireless
system with low energy budget constraints. The architecture will be validated
through extensive theoretical analysis and simulation. A major part of this
research will be the implementation of a cognitive ad-hoc wireless network that
facilitates characterization and profiling of energy consumption and QoS
measures under real world channel and interferes in shared spectrum band.
We have developed a sensor node platform to test and develop
these techniques in a real physical environment. Using off the shelf components and on board
processing provided by DSP and microprocessors we intent to demonstrate several
applications and bridge the gap between theory and reality.
Most of the existing and future applications for ad-hoc
wireless networks operating in unlicensed bands demand long battery life and
reliability. We have identified a few applications to highlight energy
efficiency of our design. Location
estimation is one of they key attributes of a wireless sensor network with many
applications such as health monitoring and security. Redundancy of sensor nodes
can provide addition correlated information to estimate the location of another
node or a target. We also intend to
demonstrate the key characteristic of the test-bed with a deployment of a
distributed acoustic beamforming system using asynchronous wireless
microphones.
The research focuses on five tasks: 1) Design of channel
selection algorithm; 2) Design of interference estimation algorithm; 3)
Adaptive optimization of modem parameters 4) Synchronization and coordination;
5) Experimental characterization of channel and QoS.