A distributed sensing/fusion network consists of more than one (spatially) separated sensors, each with possibly different characteristics and not all of them sensing the same environment. Due to their vast applicability, there has been a flurry of recent activity in the area of network design with respect to distributed sensing/fusion. Issues involved in the design of efficient networks include sensor mobility, reliability of links and capacity. This work builds on the Dynamic Expected Coverage Model proposed earlier and incorporates the issue of bandwidth capacity in the model. A Mixed Integer Linear Programming (MILP) formulation is proposed that includes first order preferential assignment with coverage and relocation of sensors. A modified column generation (CG) heuristic is developed for this problem. Computational results indicate that CG performs faster than standard commercial solvers and the typical optimality gap for large size problems is less than 10%.