This paper presents an optimal communication relay positioning method to improve the communication performance of mobile multi-node networks in complex environments. The communication channel quality prediction between nodes is of primary concern to find the optimal relay node positions while considering the uncertain and dynamic nature of environments. To this end, the learning-based or the distance model-based method is used for the channel prediction depending on the mobility of the communication nodes of interest. The global message connectivity and the worst case connectivity are introduced as the communication performance metric of networked agents. The optimal relay positions are found by maximizing the performance with respect to the relay positions through a heuristic optimization technique. This algorithm outperforms a recently-developed relay positioning algorithm in the simulations. The indoor experiments are conducted to show that the proposed approach using mobile relays improves the communication performance of the complex network significantly with the accurate channel prediction. (C) 2020 Elsevier B.V. All rights reserved.