Resource　allocation　aims　at　using　the　storage,　computing　and　bandwidth　between　edge　servers　and　end　users　while　satisfying　the　requirement　of　different　application　scenarios.　However,　there　are　some　security　threats　in　the　allocation　process,　such　as　unreal　resources　and　denial　of　service,　in　which　the　trusted　resource　allocation　becoming　a　critical　research　content.　Blockchain　technology　is　one　important　solution　for　this　problem,　but　the　high-delay　and　low-throughput　shortcomings　in　blockchain　cannot　fulfill　the　requirements　of　time-sensitive　applications.　In　this　paper,　we　firstly　utilize　a　reputation　evaluation　mechanism　to　calculate　and　rate　the　servers＇　reputation　values　in　the　resource　allocation　process.　The　high-reputation　servers　are　added　as　candidate　leaders　into　a　block　generation　group,　while　the　remaining　low-reputation　edge　servers　are　added　as　verifiers　into　a　block　verification　group.　Secondly,　we　utilize　a　round-robin　mechanism　to　choose　a　leader　from　the　block　generation　group.　The　leader　is　responsible　for　generating　a　new　block　in　each　timeslot　and　is　updated　in　different　timeslots.　Thirdly,　we　utilize　a　reward　and　punishment　mechanism　to　encourage　low-reputation　servers　to　participate　in　consensus.　Finally,　we　propose　a　proof-of-reputation　consensus　mechanism　to　realize　the　trusted　resource　allocation　based　on　smart　contracts,　which　can　match　the　end　users　with　high-reputation　edge　servers　automatically.　Simulation　results　show　the　consensus　delay　and　throughput　of　resource　allocation　achieve　significant　performance　improvements　in　comparison　with　current　schemes,　and　the　system　is　more　sensitive　to　dishonest　edge　servers　and　quickly　reduces　their　reputation　values.