Inspiring wireless sensor network from brain connectome

Abstract

Wireless Sensor Networks (WSN) are usually large - scale self - organized networks that can dynamically change with no pre - established infrastructure or a topology. In order to inference information from it, data collected by different sensors should be aggreg ated, known as Data Fusion (FC). This can happen in a centralized mode by broadcasting all data to a FC or in a distributed way. The centralized approach needs high communication bandwidth and transmission power, which is usually lacking due to limited cap abilities of sensor nodes. In distributed processing, instead of transmitting all the data to a FC in order to accomplish the final goal of the network, each sensor should rely only on local information received by itself and the sen sors in its vicinity. O n the other hand, relying only on the information received by a single sensor (or a small group of them) might not necessarily lead to the overall precision required by the network. Thus, appropriate information sharing and collaborative processing algorit hms should also be put in place to make sure of reliable inferencing. Distributed processing makes large - scale sensor networking possible by striking a proper trade - off between performance and resource utilization. The proposed methodology in this research is to use the idea of sparse structures which the best example of it, is human brain network of neurons known as connectome. Many studies demonstrate inferencing reliability (performance) and energy efficiency (resource utilization) of connectome. In this research a review of the possibility of using brain connectome in wireless sensor network design has been presented .