With　the　rapid　development　of　the　communication,　traditional　LFM　and　phase-coded　signal　are　gradually　falling　short　of　the　demand.　The　conventional　radar　signal　is　easy　to　be　intercepted　and　sorted.　So,　nowadays　we　usually　try　to　increase　the　complexity　of　signal　to　promise　the　efficiency　and　safety.　Meanwhile,　it　is　also　essential　to　optimize　the　signal　to　avoid　the　influence　from　side　lobes.　This　paper　designs　a　mismatched　filter　based　on　convex　optimization　for　frequency-phase　hybrid　coded　waveform.　Frequency-phase　coding　has　significant　advantages　in　improving　data　transmission　efficiency　and　anti-interference　capability,　but　its　complex　coding　structure　poses　challenges　for　filter　design.　To　find　the　optimal　solution,　this　paper　converts　the　problem　into　a　second-order　cone　programming　(SOCP)　problem.　Simulation　results　show　that　compared　with　the　matched　filter,　it　can　significantly　improve　PSLR　and　ISLR　of　the　waveform.　Experimental　results　indicate　that　by　adjusting　the　similarity　between　the　mismatched　filter　(MMF)　and　the　matched　filter　(MF),　the　optimization　effect　can　be　effectively　improved.　©　2024　IEEE.