The two-plasmon-decay (TPD) instability is a significant concern in direct-drive inertial confinement fusion (ICF) experiments for its low threshold and highenergy electron generation. We study the TPD instability for parameters relevant to ICF using an existing particle-in-cell (PIC) code OSIRIS and a fluid code LTS developed in this thesis work. Both the linear and nonlinear regimes of TPD are explored. In the linear regime, a convective gain formula retaining the dependence on electron temperature and perpendicular mode number is derived and is shown in good agreement with the fluid simulation results. The growth rates and thresholds predicted by the linear theories are verified in both PIC and fluid simulations. Pump depletion due to convective modes' domination are observed in the well-above-threshold PIC simulations. The PIC simulations show that both the absolute and convective modes saturate due to ion density fluctuations, which can suppress TPD by raising the instability threshold through mode coupling. A series of PIC simulations have been done to study the long-term (∼ 10ps) nonlinear behaviors of the TPD instability. The simulations have reached a quasi-steady state where the laser flux absorbed in the simulation box is balanced by the particle energy flux leaving the box. When the TPD threshold is exceeded, the simulation results show that significant laser absorption and energetic electron (> 50keV) generation occur in the nonlinear stage. The hot electrons are stage-accelerated from the low-density region to the high-density region. New modes with small phase velocities develop in the low-density region after saturation. These modes can more effectively couple to background thermal electrons and form the first stage for electron acceleration. The fluid simulations show that similar new TPD modes can develop under static ion-density fluctuations. The laser absorption and hot-electron production from these 2-D plane-wave-driven PIC simulations are higher than experimental observations, which could indicate the importance of non-ideal factors such as speckle structure in the actual laser profile and collisions, which are not included in the simulations. This work gives us a deeper understanding in the hot-electron generation in TPD, which is an important part of direct-drive ICF research.