With　the　rapid　development　of　network　technology,　network　security　is　facing　serious　problems.　Distributed　Denial　of　Service　(DDoS)　attack　is　one　of　the　most　difficult　security　threats　to　guard　against.　In　this　paper,　we　propose　a　DDoS　detection　method　based　on　improved　generalized　entropy.　The　model　includes　a　preliminary　detection　module　based　on　improved　generalized　entropy　and　a　DDoS　detector　based　on　deep　neural　networks　(DNN).　The　preliminary　detection　module　filters　as　much　normal　traffic　as　possible　while　ensuring　the　accuracy　of　the　model　by　calculating　the　generalized　entropy　threshold　of　the　traffic.　The　DNN-based　DDoS　detector　takes　the　filtered　data　as　input　and　detects　DDoS　attacks　more　accurately.　The　experimental　results　show　that　the　method　achieves　more　than　99%　accuracy,　precision,　and　recall　on　the　dataset　of　this　paper.　©　2023,　The　Author(s),　under　exclusive　license　to　Springer　Nature　Switzerland　AG.