Asynchronous and high-accuracy digital modulated signal detection by sensor networks

Adaptive modulation can be used to significantly enhance the spectrum utilization for both civilian applications, e.g., cognitive radio, and military ones. It requires a receiver to follow the modulation variation of a transmitter dynamically and automatically. Binary automatic modulation detection technology has been applied to detect the modulation scheme for adaptive modulation. Modulation classification was proposed for a multi-sensor scenario in our prior work using a distributed algorithm to enhance the detection accuracy. In this paper, the function of blind detection by the wireless sensor network (WSN) is further extended to support multiple modulation hypotheses. Besides providing spatial diversities, distributed sensors perform complicated calculations of likelihood functions and use the master node (radio) for data fusion and maximum likelihood testing. The proposed method requires no synchronization across the network and works well with low transmission bandwidth. Both analytical and numerical results are presented to validate its effectiveness.