—As it is well known, multimodality is a very common task in human-robot communication. Human conversation is also considered multimodal, and a great amount of research is done worldwide to engineer novel robotic systems, with more and more intelligence for human gestures or speech recognition abilities enhancements embedded within them. This paper presents a Fractional Fourier transform-based strategy for multimodal communication abilities improvement of pervasive mobile robots. Using a special hardware architecture, based on the standard configuration of the NI SbRIO-9631 prototype robot, a novel voice signals recognition algorithm has been tested and implemented. The experiments prove that the pervasive mobile robot endowed with these additional voice signals analyzing abilities displays more intelligence and cooperativeness in its environment significantly improving human-robot multimodal communication.