Abstract-in SGPMPLS VPNs, Traffic predominantly limited to intra-domain and
Engineering (TE) achieved by using single administrative domain. Nevertheless
Multiprotocd Label Switching (MPIS) was with the rapid expansion of enterprise predominantly limited to intra-domain and scale. a VPN spans a large number single administrative domain. However of administrative domains, on the other with the rapid expansion of enterprise hand, today's operators acknowledge that scale. a Virtual Private Network (VPN) has the edges of their networks spanned a large number of administrative where they forward traffic to other domain domains, so effective management of inkr- is the source of their greatest costs, so domain traffic is urgently demanded. This effective management of their paper presents the simplest step called as inter-domain traffic is demanded. BGP-based inter-domain TE. The method is A route in the Inremet may take a longer on the basis of various Border Gateway AS path than the shortest AS path due to Protocol (BCP) attributes. BCP policy and routing policies. The extent of Autonomous System (AS) relationship. To AS path inflation observed from the Route implement TE, the method applies diverse View routing tables. From an ISP, at least routing policies by configured BGP 55% of As paths are inflated by attributes to control the outgoing and at least one AS hap and AS paths can be incoming trafic in an AS. We ensure that inflated by as long as 6 AS hops [2]. the best route inan AS depends on the In this paper, the authors present the routes coming from its neighboring ASes simplest step called as BGP-based inter- that apply export policies, as well as the domain TE. The method is on the basis import policies of the AS. Simulation shows of various BGP attributes, inbound and that our approach can effectively outbound filter policy and AS relationship. implement inter-domain "E in BGPMPLS We apply diverse routing policies for VPNs. selecting routes from and for propagating Index Terms-AS, BGP, MPLS, VPN reachability information to other corresponding ASes. This policies can be I. INTRODUCTION implemented by configured BGP attributes VPN is a network where customers have to control the outgoing and incoming connectivity A across a shared traffic in an AS in order to implementing infrastructure using the same access TE. Simulation shows that our approach and security policies as a private network. can effectively implement inter-domain TE One of methods to build VI" has been in BGPlMPLS VPNs. We organize the defined in RFC2547[1] (i.e. BGPlMPLS context of this paper as folIowing. VPNs), In BGPlMPLS VPNs, BGP is used to Section I1 overviews BGPNPLS VPNs. distribute VPN routing information and Section 111 presentsBGP attributes and MPLS is used to forward VPN data traffic policies. Secdon IV describes in detail across Service Provider (SP) backbone our approach to implement BGP-based network (it. Internet). As we know, the inter-domain TE in BGP/MPL,S VPNs and Internet consists of a number of ASes section V presents simulation results, (also called as domains). Each AS can use a followed by a conclusion in Section VI. multiple of intra-domain routing protocols, such as OSPF, IS-IS and MPLS, but they 11. OVERVIEW OF BGPMLS VPNS must present themselves as a single The infrastructure of BGP/MPLS VPNs technique administration with a coherent shown in Fig.1 routing plan to the other AS. In BGPlMPLS VPNs, TE achieved by MPLS was consists of three components, Customer Edge (CE) devices, Provider Edge (PE) routers and Provider (P) routers. A CE router is a customer border device that exchanges routing information for network reachability of a customer VPN with its directly connected PE routers using static routing, Interior Gateway Protocol (IGP) or BGP. PE routers are border routers in the SP backbone network that attach directly to CE devices. The PE router first learns local VPN routes from CE devices, and then exchanges VPN routing information with other PE routers or Autonomous System Border Routers (ASBRs) using BGP. P routers are core routers in the SP backbone network that does not attach to CE devices. Two fundamental flows occur in BGPlMPLS VPNs [3]. A control flow is used for VPN route distribution and Label
Switched Path (LSP) establishment, and a
data flow is used to forward customer data traffic. The control flow consists of two sub-flows. The first one is responsible for the exchange of routing information between CE and PE routers ar the edges of the SP backbone, as we11 as the information distribution berween PE routers or PE router and ASBR across the SP backbone. If two sites of a VPN artach io PE routers that are in the same AS, PE routers can exchange these routes with other PE routers by BGP. If two sites of a VPN are connected to different ASes (as shown in Fig.11, the ingress PE routers use BGP to redistribute rouies to an ASBR in source AS. The ASBR then uses BGP to redistribute those routes to ASBRs in other ASes. By this means, finally rhose routes are distributed to the egress PE routers in destination AS. The second control sub- ffow is responsible for establishment of LSPs in an AS by LDP(Label Distribution Protocol), CR-LDP(Constraint-basecl Roucing LDP) or RSVP(Resource Reservation Protocol).