When executing safety-critical missions, tracking algorithms must be dependable, accurate, and have quantifiable tracking performance in non-ideal environments (e.g., actuation noise, delayed communication, and noisy measurement). Trajectory tracking has already been widely explored using sophisticated control tools (e.g., feedback linearization, optimal and adaptive control, and sliding-mode control). However, tracking stability analyses are often centered around Lyapunov designs in an ideal environment with full state observation. The proposed state-feedback trajectory tracking control guarantees high precision trajectory tracking for differential drive robots. Operational bounds on output velocities, heading angle error, and tracking error are formulated and explored to guarantee tolerances in non-ideal conditions. It is common that autonomous systems will operate using communication devices that provide non-instantaneous positional measurements subjected to package loss, delay time, and inaccuracies. The latter part of this thesis will introduce a new control function to the state-feedback controller to be carried out as an intermittent feedback controller. The control function is designed as the optimal control that minimizes the average energy in the system during the time between each triggering event. The event-generator determines each triggering instance that provides sensor information to the plant based on the main objective of guaranteeing tracking stability. The analysis develops the appropriate communication policy (an integral part of the event-generator) for the open-loop critical time instance such that the tracking error asymptotically converges to subsequently remain bounded under a user-defined tolerance. The communication policy provides a direct relationship between the system parameters (e.g., actuation noise, control parameters, and reference velocity) and the sporadic communication frequency as well as the demanded velocities on the robot. The theoretical, simulation, and hardware experimentation results demonstrate operability and efficiency of the proposed algorithms in non-ideal environments with set physical limitations and constraints.