With　the　proliferation　of　Things　connected　to　the　Internet　(IoT),　network　vulnerabilities　to　Attacks　known　as　distributed　denial　of　service　(DDoS)　have　escalated.　Conventional　DDoS　detection　methods　often　falter　in　the　multifaceted　IoT　landscape.　Addressing　this,　our　research　introduces　a　novel　hybrid　deep　learning　model,　termed　CNN-LSTM-GRU,　which　synergistically　integrates　(CNN)　Convolutional　Neural　Networks,　(LSTM)　Long　Short-Term　Memory,　and　(GRU)　Gated　Recurrent　Units.　Early　findings　indicate　a　marked　enhancement　in　detection　precision　and　a　reduction　in　false　alarms　when　juxtaposed　with　existing　methodologies.　This　paper　champions　a　cutting-edge,　versatile　deep　learning　strategy　utilizing　the　CNN-LSTM-GRU　fusion　to　adeptly　discern　varied　network　threats.　Our　methodology　harnesses　feature　clusters　from　UNSW-NB15　and　BOT-IoT　Flow　datasets,　encompassing　protocols　like　DNS,　FTP,　HTTP,　MQTT,　and　TCP.　Based　on　metrics　like　accuracy,　recall,　precision,　and　F1-score,　performance　evaluation　reveals　that　our　hybrid　deep　learning　model　boasts　a　98.45%　detection　rate　against　IoT-centric　threats.　Additionally,　a　comparative　analysis　underscores　the　superiority　of　our　model　against　other　leading　detection　frameworks.　©　(2024),　(International　Journal　of　Network　Security).　All　rights　reserved.