摘要： 从空间数据之间的拓扑关系出发，提出了一种基于角色和权限的GIS 并发控制体系，利用用户的权限解决空间数据的拓扑关系问题，将拓扑关系问题演变为不同类型的空间数据集合问题。针对不同类型的空间数据集合，采用了基于角色与权限的封锁粒度，来实现对空间数据及其拓扑关系的统一封锁。由于采用了基于角色与权限的封锁粒度，大大增加了并发控制的复杂性。该文应用多维粒度树，有效地解决了封锁粒度的复杂性问题。

关键词: 并发控制；空间数据；角色和权限；多维粒度树

Abstract: Logical relativities between spatial data are analyzed and then a concurrency control system based on role and privilege(R&P) is presented. It can resolve the problem of topology relativities between spatial data with R&P by converting topology to spatial data sets of variant types. Corresponding these types, lock granularity based on R&P is used to realize uniform lock on spatial data and topology relativities between them. As using of lock granularity based on role and privilege, it is much more complex than common currency control system. So the conception of multi-dimension granularity tree (MDG-Tree) is used to resolve the problem

Key words: Concurrency control; Spatial data; Role and privilege; Multi-dimension granularity tree