This study describes an Intrusion Detection System (IDS) called MOVICAB-IDS (MObile VIsualization Connectionist Agent-Based IDS). This system is based on a dynamic multiagent architecture combining case-base reasoning and an unsupervised neural projection model to visualize and analyze the flow of network traffic data. The formulation of the underlying Intrusion Detection framework is presented in advance. The described IDS enables the most interesting projections of a massive traffic data set to be extracted and depicted through a functional and mobile visualization interface. By its advanced visualization facilities, MOVICAB-IDS allows providing an overview of the network traffic as well as identifying anomalous situations tackled by computer networks, responding to the challenges presented by traffic volume and diversity. To show the performance of the described IDS, it has been tested in different domains containing several interesting attacks and anomalous situations.