This master thesis investigates synchronization in coupled semiconductor lasers using theoretical modeling and numerical simulations. Starting from a single laser with feedback and extending to two mutually coupled lasers and a four-laser all-to-all network, the study uses Heun integration to incorporate multiplicative noise. For the two-laser system, the effects of frequency detuning, delay mismatches, and laser parameter heterogeneities are examined, with synchronization assessed through cross-correlation and statistical measures. Temporal and spectral analyses reveal quasi-periodic, chaotic, and interference-driven behaviors, yet strong synchronization often persists despite noise. In the four-laser network, simulations show clustered synchronization, fewer chaotic regimes, and sensitivity of synchronization quality to small detuning changes. The results clarify how heterogeneities influence networked laser dynamics and support the design of more robust photonic reservoir computing systems.