Flexible and realistic mathematical models of the Earth's magnetospheric field, combined with fast fitting algorithms, can serve as a tool for converting the flow of simultaneous data from a "swarm" of flying space magnetometers into a dynamically evolving three-dimensional picture of the geomagnetic field lines. Here several important aspects of the project are discussed, such as the reliability of the reconstruction of the magnetospheric configurations, the role of the irregular fluctuations in the data, spatial coverage of the magnetosphere by the multi-probes, and an estimate of the number of the spacecraft needed to achieve a reasonably accurate monitoring of the global magnetosphere in real time. The reconstruction of global magnetospheric configurations was simulated using artificial "data" generated by combining a realistic model of the
magnetosphere with added random noise. Even with a high noise level, comparable in magnitude to the external regular field, the reconstructed field line configurations were found quite close to the original ones, provided the multi-probe swarm had a sufficient coverage of the tail lobes. In another set of runs, reconstruction of disturbed field configurations was made, simulating conditions during the substorm growth and expansion phases. Overall, the real-time monitoring of the global magnetosphere based on multi-probe data is quite feasible task, provided spacecraft orbits are thoughtfully planned.