Modern satellite Attitude Determination and Control Systems (ADCS) rely heavily on star trackers to solve the "Lost-in-Space" (LIS) problem. However, traditional star-matching algorithms, such as the Triangle or Geometric Voting methods, require extensive star catalogs and significant onboard memory. These requirements pose a major challenge for resource-constrained Nano-satellites and CubeSats. This paper introduces a novel Vedic-Grid Algorithm, a computationally efficient framework that utilizes the ancient Indian 27-Nakshatra celestial partitioning system as a pre-indexing grid for star identification. By dividing the celestial ecliptic into 27 discrete sidereal sectors—each spanning 13 degrees and 20 minutes—the algorithm implements a hierarchical search strategy. The system first identifies a primary "Yogatara" (a junction star or anchor star) to determine the specific Nakshatra sector where the satellite is currently pointed. This initial step effectively reduces the active search space by approximately 96.3 percent. Once the sector is localized, a secondary refinement process uses small, localized sub-catalogs to calculate the precise three-axis orientation of the spacecraft. Preliminary simulations indicate that this Vedic-inspired partitioning significantly decreases the Time-to-First-Fix (TTFF) and minimizes CPU cycles compared to non-indexed global searches. This approach demonstrates that bridging ancient celestial geometry with modern computational astrophysics offers a robust, lightweight, and high-speed solution for autonomous satellite navigation in deep space.