(eng): In the paper, an algorithm for synchronization of a neural network composed of interconnected Hindmarsh-Rose neurons is proposed. The interconnections of the neurons are subject to delays and an additive noise is present. The convex optimization is the tool for finding the synchronization control, the result is formulated using linear matrix inequalities. The synchronization of the recovery and adaptation variables is also guaranteed thanks to the minimum-phase property of the Hindmarsch-Rose neuron. An example illustrates the results.