Pathvectoralgoritmes

Path-vector algorithms, or path-vector routing, are distributed routing methods that determine routes by carrying information about the path to a destination in the routing updates themselves. For each destination, routers maintain a path vector that records the sequence of nodes (typically autonomous systems) a packet should traverse.

Operation: A router announces a route to its neighbors together with the accompanying path vector. When a router receives such an update, it prepends its own identifier to the path, applies local routing policies, and forwards the update to its peers. A route is considered loop-free if the path does not contain the advertising router; if it does, the route is rejected.

The canonical example is the Border Gateway Protocol (BGP), the de facto inter-domain routing protocol on the Internet. In BGP, the path vector is represented by the AS_PATH attribute, which lists the autonomous systems along the route. Route selection combines path information with policy and other attributes, such as LOCAL_PREF, AS_PATH length, MED, and next-hop reachability. Policy decisions allow operators to express business rules, traffic engineering, and multi-homing requirements.

Advantages and challenges: Path-vector routing supports policy-based routing across administrative boundaries and provides loop avoidance, making it suitable for large-scale inter-domain routing. Its explicit path history helps detect and prevent routing loops, but carrying path information can raise scalability and privacy concerns. Security risks include route hijacking and prefix leaks; mitigations include RPKI, prefix filtering, and path validation mechanisms.

Variants and scalability: Network designs use route reflectors, confederations, and route refresh or graceful restart features to scale a path-vector system and manage policy complexity. Applications are primarily in inter-domain routing, though some private networks adopt path-vector concepts within larger policy frameworks.