Joint Compressive Data Gathering and Scheduling in Wireless Sensor Networks under the Physical Interference Model

Compressive data gathering (CDG) has emerged as a useful method for collecting sensory data in large scale sensor networks; this technique is able to reduce global scale communication cost without introducing intensive computation, and is capable of extending the lifetime of the entire sensor network by balancing the aggregation and forwarding load across the network. With CDG, multiple forwarding trees are constructed, each for aggregating a coded measurement, and these measurements are collected at the sink for recovering the uncoded measurements from the sensors. This paper studies the problem of constructing forwarding trees for collecting and aggregating sensed data in the network under the physical interference model. The problem of aggregation tree construction and link scheduling is addressed jointly, through a mathematical formulation, and its complexity is underlined. Our objective is to collect data at the sink with minimal delays and fewer transmissions. Owing to the complexity of the joint problem, we present a decentralized method for solving the tree construction and the link scheduling sub-problems. Our link scheduling sub-problem relies on defining an interference neighbourhood for each link and coordinating transmissions among network links to control the interference. Numerical results are presented to compare the performance of the decentralized solution with the joint model as well as prior work from the literature.