Distributed Signal Processing Algorithms for the Physical Layer of Large‐Scale Sensor Networks
An-swol Hu, S.D. Servetto · 2005
The ability to move the sensed data out of the network (the reachback communication) is one of the basic communication primitives that must be supported by every sensor network. This is most commonly achieved by routing of information through the network to some central data collection point that will act on the sensed information. Authors propose an approach to this problem by developing a method to allow all nodes in the network to cooperatively generate a strong information bearing signal to communicate with a distant data collection point. Such a solution would be extremely robust to the failure of nodes in the network and would allow for the deployment of a homogeneous network of extremely small, low-power nodes for a variety of applications. This approach to the sensor reachback problem is called cooperative reachback. In this chapter two aspects of the cooperative reachback problem are considered: time synchronization and reachback modulation schemes. Time synchronization facilitates the design of cooperative reachback modulation schemes. A system model that will apply to both the time synchronization problem and reachback communication is described. Authors then study the properties of waveforms generated by asymptotically dense networks and then move into the development of a time synchronization mechanism for dense networks using these waveform properties. In the asymptotic regime, this time synchronization framework can keep the synchronization mean squared error from increasing with distance from the ideal time source. Next, authors study the performance of this asymptotically optimal scheme for networks of finite size. They then develop a modulation scheme for reachback communication again using these waveform properties and study its performance using simulations.