Abstract
In this paper, we study the problem of cooperative abnormality detection using mobile sensors in a fluidic medium, based on a molecular communication setup. The sensors are injected into the medium to search the environment for the abnormality. To reduce the effects of sensor imperfection, we propose a cooperative scheme where the sensors activate each other by releasing some molecules (i.e., markers), into the medium after they sense an abnormality. A number of fusion centers (FC) are placed at specific locations in the medium, which absorb all sensors arrived at their locations. By observing the states of the received sensors, each FC decides whether an abnormality exists in its corresponding region or not. Then, it resets the sensors' activation flags and releases them again into the medium to proceed with the next regions. In our model, both sensors' imperfection and markers' background noise are taken into account. We consider two sensor types, the memoryless sensors that get active based on the number of received markers in each sampling time and the aggregate sensors that get active based on the summation of received markers in all sampling times, before they reach the FC. We analyze the related binary hypothesis testing problem and obtain the probabilities of false alarm and misdetection for the memoryless and aggregate sensors. Then, we obtain the probability of error. It is shown that using sensors with the ability of activating each other can significantly improves the performance in terms of probability of error.