We present a generic mobility and traffic generation framework that can be incorporated into a tool for modeling and simulating large scale ad~hoc networks. Three components of this framework, namely a mobility data generator (MDG), a graph structure generator (GSG) and an occlusion modification tool (OMT) allow a variety of mobility models to be incorporated into the tool. The MDG module generates positions of transceivers at specified time instants. The GSG module constructs the graph corresponding to the ad hoc network from the mobility data provided by MDG. The OMT module modifies the connectivity of the graph produced by GSG to allow for occlusion effects. With two other modules, namely an activity data generator (ADG) which generates packet transmission activities for transceivers and a packet activity simulator (PAS) which simulates the movement and interaction of packets among the transceivers, the framework allows the modeling and simulation of ad hoc communication networks. The design of the framework allows a user to incorporate various realistic parameters crucial in the simulation. We illustrate the utility of our framework through a comparative study of three mobility models. Two of these are synthetic models (random waypoint and exponentially correlated mobility) proposed in the literature. The third model is based on an urban population mobility modeling tool (TRANSIMS) developed at the Los Alamos National Laboratory. This tool is capable of providing comprehensive information about the demographics, mobility and interactions of members of a large urban population. A comparison of these models is carried out by computing a variety of parameters associated with the graph structures generated by the models. There has recently been interest in the structural properties of graphs that arise in real world systems. We examine two aspects of this for the graphs created by the mobility models: change associated with power control (range of transceivers) and variation in time as transceivers move in space.