中文摘要：

在TBF的基础上提出了用于多属性元数据快速查询的双层体系结构（DBF）和基于分组的延迟更新策略（GDA）,GDA配合DBF中的确认验证机制,在不会附加错误查询的前提下大幅度的减少了属性更新所需要的操作,从而能够有效地减小系统更新开销.边数量更新策略解决了DBF中消减操作的更新冲突问题,同时,最大限度的减少消减操作次数,提高了更新效率.

英文摘要：

A multi-attribute double quick search metadata architecture DBF（Double-level Bloom Filters） and the packet-based update delay strategy GDA（Group-based Delayed-update Approach） is presented on the basis of TBF.GDA combining with the authentication mechanism of DBF can reduce updating operation of object attribute under the premise that the additional error check will not be significantly increased,which can effectively reduce the cost of system updating operation.Updating strategy of edges number proposed reduces operations to minimize the number reduction and improve the updating efficiency while addressing the update operation of the conflict,since the shared edge of DBF ring or chain can not exist simultaneously.