Fusing Dependent Decisions for Hypothesis Testing With Heterogeneous Sensors

In this paper, we consider a binary decentralized detection problem where the local sensor observations are quantized before their transmission to the fusion center. Sensor observations, and hence their quantized versions, may be heterogeneous as well as statistically dependent. A composite binary hypothesis testing problem is formulated, and a copula-based generalized likelihood ratio test (GLRT) based fusion rule is derived given that the local sensors are uniform multilevel quantizers. An alternative computationally efficient fusion rule is also designed which involves injecting a deliberate random disturbance to the local sensor decisions before fusion. Although the introduction of external noise causes a reduction in the received signal-to-noise ratio (SNR), it is shown that the proposed approach can result in a detection performance comparable to the GLRT detector without external noise, especially when the number of quantization levels is large.