A form of least squares solution in which the observed baseline vectors are treated as "observations" in a secondary network adjustment (see Network Adjustment), and only one coordinate must be held fixed to its known value while all others are allowed to adjust. Typically GPS surveys measure more baselines than the minimum needed to coordinate all the points in the network. These extra "observations" are redundant information that a minimally constrained network adjustment uses to derive optimum estimates of the coordinate parameters, as well as valuable quality information in the form of parameter standard deviations and error ellipses (or ellipsoids).

This definition is in context to GPS. See more contextual defintions for Minimally Constrained.