In this paper, a novel communication constraint is introduced for the synchronization problem of coupled agents over finite field networks. Given a network of interconnected agents, the state information of agents may contain multiple levels, therefore multiple communication channels between two connected agents are needed to transmit corresponding level information. Compared with existing results over finite fields, the model considered in this paper is assumed that there are inactive communication channels in all communication channels between two agents, that is, only part of state information can be transmitted. Moreover, due to the disturbance or irregular environmental variation, the switching dynamics of active channels is also considered in the model. The proposed communication constraints bring some difficulties to the analysis of network synchronization since less information transmission between agents. To overcome these difficulties, we introduce a group of channel matrices to describe the activation of communication channels, and a state layered method is utilized to derive the main results. Then, two criteria are obtained to judge the synchronization of agents over finite field networks. Finally, two numerical examples are given to illustrate the effectiveness of our theoretical results.