A low complexity multi-threshold centralized detection strategy for cooperative spectrum sensing

In this paper, we propose a low complexity multi-bit multi-user combining strategy for cooperative spectrum sensing in cognitive radio networks. We consider the optimal detector where the fusion center receives multi-bit sensing information from several nodes and combine them to make a final decision on the status of the channel. Each node employs a linear quantization technique to generate its multi-bit decision before sending it to the fusion center. Our goal is to optimize this quantization to have the minimum error in sensing. To do so, we use the central limit theorem to get an analytical expression to approximate the performance of the optimal detector at the fusion center. Then, we use this expression to find optimal parameter which is used for quantizing the information in each node. Simulation results are provided to show that the performance evaluation expression is very close to the actual values. Furthermore, we show that using the optimized quantization in each node, the sensing accuracy improves significantly by using 2 or 3 quantization bits compared to the single bit hard decision.