Wireless sensor networks (WSNs) are being widely used in sensing, collecting, and disseminating data for various applications in environmental, industrial, and military applications. In this paper we present MAConf, a layer-2 detection mechanism for mobile WSNs. We propose the use of radio channel analysis to predict the underlying MAC protocol and addressing scheme, facilitating interconnections among WSN while ensuring an energy-efficient detection strategy. By implementing the intelligence using a neural network, we show that a 99% hit rate can be achieved in a variety of scenarios.