由于现有VLAN(虚拟局域网)自动学习方案VTP(虚拟终端协议)和GVRP(GARP VLAN注册协议)存在各自的不足之处，使VLAN自动学习功能在实际应用中受到限制。文章分析了现有两种VLAN动态学习方案的工作原理及其不足，提出一种改进的VLAN动态学习方案。该方案能解决VTP和GVRP在实际应用中带来的问题，使VLAN自动学习功能能够得到更好的应用。

As the existing Virtual Local Area Network (VLAN)-based automatic learning schemes, i.e. Virtual Terminal Protocol(VTP)and GARP VLAN Registration Protocol (GVRP) have their respective deficiencies, the automatic learning functions of VLAN are restricted in practical applications. This paper analyses the working principles and shortcomings of these two VLAN-based learning schemes and presents an improved one, which can resolve the problems of VTP and GVRP in practical applications so that the VLAN automatic learning functions can be better applied.