SNR estimation and decision making using hypothesis testing in energy-efficient adaptive modulation

To improve energy efficiency in low power short-range communication systems such as Wireless Sensor Networks (WSNs) and Wireless Body Area Networks (WBANs), adaptive techniques have been proposed to adjust the transmit waveforms according to the channel conditions. The estimation of channel quality, such as Signal-to-Noise Ratio (SNR), plays an important role in the waveform switching. Errors introduced by the estimation result in wrong decisions on transmit mode, which would further affect the overall energy efficiency. In this paper, we consider a low power system that adopts Link Adaptation (LA) switching between multiple modulation schemes. We point out that the decision of switch between different modulations only requires the knowledge of whether the SNR falls into a certain range, rather than that of an exact value. Therefore we propose a novel SNR estimation and decision making framework based on Hypothesis Testing (HT) theory. This framework provides a generic way to both evaluate and control the probabilities of different types of estimation errors. Hence approach can be adjusted for different applications to improve energy efficiency by reducing the probability of the errors associated with significant energy cost. We also highlight that the conventional estimation method is a special case of the proposed framework.