In　this　paper,　we　investigate　a　noise　reduction　permutation　index　DCSK　(NR-PI-DCSK)　to　ease　the　side　effect　of　the　additive　white　Gaussian　noise　(AWGN)　and　improve　the　bit　error　rate　(BER)　performance.　In　the　first　time　slot,　β/R　chaotic　samples　are　generated　and　repeated　R　times　to　form　the　β-length　reference　sequence.　Permuted　replica　of　the　reference　sequence　is　sent　in　the　second　time　slot.　At　the　receiver,　received　chaotic　samples　are　averaged　and　correlated　to　recover　the　bits　transmitted　by　the　predefined　permutations　and　DCSK.　The　average　operation　reduce　the　noise　level　and　thus　improve　the　BER　performance　compared　to　conventional　PI-DCSK.　Simulation　results　show　that　the　proposed　method　can　achieve　favorable　performance　gain　and　can　be　regarded　as　a　candidate　for　low-power　and　low-complexity　communications,　e.g.,　wireless　sensor　network　(WSN)　for　underground　power　pipe　gallery.　　©　2023　IEEE.