An Energy-Efficient Bit Allocation Scheme in Wireless Sensor Networks

Abstract

In wireless sensor networks (WSN), a large number of sensor nodes which are capable of sensing, data processing and communicating are densely deployed in an area to measure some physical phenomenon. Generally, wireless sensor nodes carry very limited irreplaceable power sources. Thus, two primary concerns in WSN are to save the overall energy consumption and to prolong the network lifetime, namely the time when all the nodes are functional.

Motivated by these two concerns, this thesis mainly focuses on the energy efficient transmission and bit allocation schemes in multi-source single-sink WSN from an information theoretic point of view. Specifically, this thesis investigates the interactions between source coding and channel coding to gain cooperation between them in terms of energy efficiency.

For transmission through additive white Gaussian noise (AWGN) channel with path loss, this work shows that the overall energy consumption can be minimized if each source transmits with minimum power and cooperates with other sensors in TDMA (time-division multiple access) mode. From the source coding perspective, the Slepian-Wolf coding theorem is applied for efficient bit allocation since sources are usually highly correlated in WSN. Combining the transmission with correlated source coding, we derive an optimal closed form bit allocation scheme to minimize the overall energy consumption. The fundamental idea is to allocate more bits to the nodes with better channel conditions and less bits to the nodes with worse channel conditions. Based on this scheme, we further maximize the network lifetime and develop a heuristic algorithm to average the distribution of energy consumption among all sensors. Both analytical and simulation results are presented to show the superiority of our schemes.