When building a VPN based on p-to-p overlays

When building a VPN (Ferguson et al., 1998) based on p-to-p overlays, connection-oriented (like ATM or frame relay, tunneling-on-IP techniques) scalability is a main problem, while VPNs based on MPLS are used to address scalability issues (as they are purely designed on the basis of connection-less, peer-based architecture). Since, a customer-site in a peer-based architecture requires the peer simply within a single provider-edge router in place of the entire customer-edge/provider-edge routers which are associated with the VPN that results in the reduction of large number of VCs. In addition, MPLS-based VPNs naturally use connectionless approach. The Internet is to be obliged its worth to its fundamental approach which is based on connection-less, packetswitching network topology (i.e., TCP/IP). Thus, it does not require any prior act to make association possible in a flexible and useful way among the hosts. In an IP-based connection-less setup the traditional VPNs require initial connection establishment process over p-to-p, connection-oriented overlay networks. When it utilizes under a connection-less environment it still can not get benefit from the connection simplification and service expandability offered by connection-less network. In contrast, if a connectionless VPN is built, to guarantee the network privacy the use of tunnels and encryptions are not required, hence it eliminates the considerable complications.