Wireless Sensor Networks is a research area which has gained a lot of interest today in academia as well as Industry. One of the topic that has recently caught my interest is: Neural Wireless Sensor Networks.
Neural Networks consist of multiple layers, each layer separated by a non-linear activation function. Each layer consists of multiple ‘neurons’. A visual representation of a Neural Network would be as follows:
These networks are initialized arbitrarily, and trained on a data-set using algorithms like back propagation. Once a neural network is trained on a data-set it can be deployed on real time inputs and the outcome can be evaluated.
Now imagine each neuron as a node in a sensor network. The data collected in real time, can also be processed in-network and the WSN can be used to make smarter decisions.
You can refer to some of these papers for more details:
- Oldewurtel, Frank, and Petri Mahonen. “Neural wireless sensor networks.”Systems and Networks Communications, 2006. ICSNC’06. International Conference on. IEEE, 2006.
- Nehra, Neeraj Kumar, Manoj Kumar, and R. B. Patel. “Neural Network based energy efficient clustering and routing in wireless sensor networks.” Networks and Communications, 2009. NETCOM’09. First International Conference on. IEEE, 2009.