A polymorphic recogniser for distributed intelligence in wireless sensor networks

We present an innovative scheme for performing pattern recognition in wireless sensor networks. It is based on the concept of a distributed computer that can alter itself to adjust to the user's needs. By emulating the distributed nature of biological computational devices and evenly spreading the computational and communication workload throughout the network, we will reduce energy consumption and enhance the reliability of the recognition process of sensing applications utilising this approach. We introduce the basic concepts of the scheme as a computer that operates via message passing (and not arithmetic operations), and show a theoretical validation of its operation.