Toward Energy-efficient Communication in Wireless Sensor Networks

Participating Faculty:

Faramarz Fekri (ECE), Umakishore Ramachandran (CoC), Raghupathy Sivakumar (ECE)

Optimizing the energy consumption in wireless sensor networks is the joint focus of the wireless sensor network (WSN) umbrella project. The project involves three complimentary aspects of energy consumption in WSNs and deals with redefining the functionalities embodied in the protocol stack of each sensor network node. In wireless sensor networks, two of the fundamental tasks are broadcasting of commands and queries to sensors, and gathering of sensed data in the field followed by the delivery of the data to the sink– a central coordinating entity. This data communication is typically characterized by a high degree of temporal and spatial correlation. Under such conditions, it can be highly beneficial to eliminate any redundancy much before the data reaches the final destination at the sink (data gathering) or the sensor nodes (broadcasting). Essentially, eliminating the redundancy near the original sources will save considerable amounts of scarce bandwidth and energy resources in the wireless sensor network. The first two aspects explored in the project deal with evolving correlation aware communication strategies that span the coding and routing layers of the protocol stack. The proposed strategies are complementary to each other, and can result in several folds of energy savings to the nodes of the sensor network. The third aspect investigated adds a new dimension to wireless sensor networks, namely, mobility. With advances in micro air vehicles (MAVs), mobility offers a thus-far unmined opportunity to optimize the energy consumption in wireless sensor networks. The transport and network layers of the protocol stack can use the inherent mobility attribute of the sensor network nodes in route selection and hence energy optimization.

Distributed Source-Channel Coding: In this work, Dr. Fekri will investigate distributed source-channel coding techniques that will provide effective data compression and channel coding for wireless sensor networks where sensor devices send correlated information back to the sink and the sink broadcasts commands (through multiple paths) to the sensor nodes.

Scalable Correlation Aware Aggregation Tree: While the above work focuses on the data coding techniques, Dr. Sivakumar’s work will focus on constructing appropriate routing structures back to the sink that will enable to the fullest extent, and only to the extent possible, any fusion of information possible due to correlation in wireless sensor networks.

Mobility Modeling, Mobility Virtualization, and Route Selection: In this work, Dr. Ramachandran will explore optimizing computation, communication, and control (i.e. mobility) from the point of view of conserving energy in wireless sensor networks. Key issues to be investigated include adaptive role assignment, distributed load balancing, mobility modeling, and mobility virtualization. These issues are complementary to the correlation aware coding and routing issues identified by the previous two sub-problems.