The great developments in applied mathematics and computational capabilities facilitate the design and implementation of robust control. In addition, the huge developments in nanotechnology and its availability in civilian level with less cost, size and weight attract many of the researchers allover the world towards embedded systems especially the embedded flight control. Among the real applications are the guided missiles especially the antitank guided missile systems which are commanded to the line of sight (CLOS) against ground and short range targets. The present work is concerned with improving the performance of an antitank guided missile system belonging to the first generation via robust synthesis of autopilot and guidance systems. The design and analysis necessitates somehow accurate model with different uncertainties (objective of Part-1 of the paper) for the system, a robust autopilot design (objective of Part-2 of the paper) and implementation via hardware in the loop (HIL) simulation (objective of Part-3 of the paper). This part of the paper is devoted to the derivation of the missile-control system transfer functions representing the system dynamics in pitch plane based on the designed 6DOF simulation model in Part-1 of the paper. These transfer functions are augmented with
mathematical formulation for the system uncertainty to be considered during the robust (H) design. Then, it presents the results of 6DOF simulation with justification and validation against previous work. The next objective for this part is the autopilot design using  H technique with justification against previous work and reference flight data concerning the
performance requirements of time responses and flight path characteristics. The new design is implemented within the 6DOF simulation from which the obtained results clarify its capability to stabilize the system in presence of un-modeled dynamics and satisfy the performance requirements with disturbance rejection and measurement noise attenuation.
However, the selection of the weighting functions necessary for the  H design is cumbersome and necessitates more investigation to stay on some rules of thumb as guidelines to subsequent research. Towards this objective the next part of the paper is a first trial via conducting the simulation with some hardware (pitch and yaw) in the loop.