In the paper, a cooperative differential network game is considered. We suppose that players choose simultaneously and independently the neighbors with whom they intend to interact during the game. Each player can choose neighbors from a fixed subset of players. Such subsets can be different for different players and, for each player, the number of its possible neighbors is limited. The players create the network to maximize the joint payoff. But the network that is optimal at the initial time instant may cease to be such afterwards. To avoid this, we propose an analogue of IDP (see [6])—payoff distribution procedure which changed the instantaneous payments of players, keeping the total payoff of players in the game under cooperation fixed (unchanged).