With　the　advancement　of　digital　photography,　high-definition　video　has　become　crucial　in　daily　life,　professional　production,　and　security　monitoring.　However,　low　light　or　small　sensors　often　lead　to　noise　in　captured　videos,　impacting　visual　quality　and　subsequent　processing.　To　address　these　issues,　we　propose　a　lightweight　video　denoising　algorithm　(LVDA).　First,　our　method　introduces　SimpleGate,　a　variant　of　the　gated　linear　cell　that　incorporates　nonlinearity　independently　of　sigma,　enabling　direct　division　and　multiplication　of　feature　maps　in　the　channel　dimension　to　reduce　computational　load　while　maintaining　performance.　Second,　we　present　a　simplified　channel　attention　mechanism　as　an　alternative　to　traditional　complex　channel　attention,　further　enhancing　network　efficiency.　Based　on　the　SimpleGate,　we　propose　a　channel　gating　block　to　replace　the　residual　dense　block.　Third,　we　adopt　depthwise　overparameterized　convolution　to　replace　traditional　convolution,　reducing　computation　and　model　parameters　while　maintaining　network　structure　and　performance.　Comprehensive　quantitative　and　qualitative　experiments　demonstrate　the　effectiveness　of　our　LVDA.　(c)　2025　SPIE　and　IS&T