<p>Cognitive monitoring systems are now playing a vital role in cyber-physical systems’ security and smart cities’ safety. For envisioning the secure cities and areas through detecting devices at the edge, collaborative networked detectors for cognitive radars have the ability of provisioning multiple-input and multiple-output (MIMO) models’ capability. They can change the shape of their transmitted information; therefore, the design of the transmitted waveform by these devices has recently attracted the attention of researchers. In this study, an optimally designed waveform based on the Kullback-Leibler divergence (KLD) model is proposed which leads to an increase in the probability of detection of the distributed nodes at the edge of the monitoring system with limited device resources. For this purpose, first, the probability density functions (PDFs) of the assumptions in the hypothesis test are extracted and then, at each step, the transmitted waveform is designed in such a way that distance of the probability density functions of the hypotheses is maximized. In fact, the distance of the probability density functions of the hypotheses increases and the probability of separating the true hypothesis from the false hypotheses also increases. The simulation results showed that by using the designed waveforms, the probability of detection increases.</p>