Energy Efficient Sequential Sensing for Wideband Multi-Channel Cognitive Network

Miao Xu,He Li,Xiaoying Gan

DOI: https://doi.org/10.1109/icc.2011.5962519

2011-01-01

Abstract:Recently, Cognitive network is growing more and more attractive due to its capability of dynamic spectrum usage. When the dynamic spectrum is wideband, a sequential sensing scheme is proposed to sense the spectrum with low system complexity. Conventionally, parameters of sequential sensing are optimized subject to sensing time and detection performance constraints, such as false alarm and miss detection probability. However, when energy consumption is considered, tradeoff involved in sensing parameters design have to be studied. Apart from sensing time, we find that the energy consumption of a cognitive handset is greatly affected by the sampling rate of A/D&D/A convertor, which will affect detection performance as well. For the purpose of figuring out the optimal sequential sensing parameters, we establish Energy Optimization and Detection Optimization problems based on the relationship among energy consumption, sampling rate and detection performance. By exploiting the convexity property of energy consumption function and monotonic decreasing property of detection performance function with respect to sampling rate, the optimal solution of a practical cognitive handset utilizing AD9051(ADC) is obtained. Numerical results are provided to investigate the impact of sampling rate on energy consumption and detection performance. It is shown that the optimal sampling rate will bring up to 50% energy saving and 3 times detection performance improvement.