What is an autonomous BGP system

Border Gateway Protocol

With the development of the Border Gateway Protocol (BGP) an attempt was made to solve the problems of the EGP protocol and to create a routing protocol that is capable of making intelligent routing decisions. For this reason, a transition from EGP to BGP is currently taking place on the Internet.

The main task of the BGP protocol is the exchange of routing information between exterior gateways in the core network. For the transmission of this information, the BGP protocol uses network reachability information, which, in contrast to the EGP protocol, contains additional information in addition to information about the accessibility of a specific target network. This information, known as attributes, includes

Origin: Information about the origin of the routing information (e.g. from an IGP protocol).
AS Path: A list of the autonomous systems that have to be traversed on the way to the destination network.
Next Hop: The nearest router to be contacted on the way to a specific destination.
Unreachable: An attribute that indicates whether a previously specified route is still correct.
Inter-AS Metric: A unit of measure that an EGP router propagates in order to announce costs for routes that lie within its autonomous system. This information may be used by routers from other Autonomous Systems in order to find a better entry point into the Autonomous System from which the information originates. However, forwarding this information is not permitted.

With the help of this information, a router then builds a graph that contains information about the accessibility and costs of a specific autonomous target system. Any existing cycles in the graph are eliminated so that there is only a single path from a router to a target network.

Although BGP creates a very extensive view of the network topology with the help of the AS Path attribute, it cannot be clearly assigned to the link state method class, as it only passes on information to other routers about paths that the BGP protocol itself uses. This is actually a characteristic of the distance vector method. To reduce the bandwidth required for the exchange of routing information, the BGP only uses update information that describes changes in the network after it has initially transmitted its complete routing table. If there are no changes in the network, so-called keep-alive messages are transmitted at fixed time intervals, which ensure that a neighboring router can still be reached.

For the transmission of the routing messages, the BGP uses a TCP connection, which ensures that routing information reaches the neighboring router. In addition to your own security mechanisms, you can use all authentication mechanisms that the protocol makes available to the transport layer. The BGP uses very different mechanisms for authentication internally, which also vary greatly in their complexity. As part of the BGP mechanism for authentication, the protocol allows e.g. the sending of digital signatures that are present in the BGP messages.

BGP allows the manager to set up specific strategies through additional mechanisms. This includes the possibility that outsiders only find out certain routes so that the internal structure of a network is not disclosed, which can prevent certain attacks. In addition, each autonomous system can determine for itself whether it is to be used as a Transit system is used and forwards data traffic that is not intended for its own autonomous system, or whether it is used as a Stub system works, which does not pass any external data traffic. The prerequisite for working as a transit system is of course that an autonomous system has at least two gateways to the core network. Autonomous transit systems are also called Internal service provider (ISP) designated.