Distance Vector Routing Protocol (Explain with Diagram)

Distance Vector Routing is a dynamic routing approach where each router learns “best routes” by exchanging distance vectors with its directly connected neighbors, rather than knowing the full network topology. The protocol is mathematically grounded in the Bellman–Ford shortest-path update rule: routers update their distance to each destination using the metric advertised by neighbors. As updates propagate, routing tables eventually stabilize to a state called convergence.

Key idea: each router maintains, for every destination, a pair like (distance, next-hop), derived from neighbors’ advertised distances.

1) Core Mechanics (What a Distance Vector Router Does)

A distance-vector router operates with three essential components:

1. Neighbor exchange: periodically (and/or when changes occur) it sends its distance vector to directly connected neighbors.
2. Metric-based update: for each destination $y$, router $x$ computes a new distance estimate via each neighbor $v$, then takes the minimum. This is the Bellman–Ford logic used in distance-vector routing.
3. Routing table selection: it stores the best next hop based on that computed minimum, and advertises updated distances in future messages.

A common formal form of the update is: $D_x(y) = \min_{v \in neighbors(x)} \{c(x,v) + D_v(y)\}$
where:

• $D_x(y)$ = router $x$’s current estimated distance to destination $y$
• $c(x,v)$ = cost from $x$ to neighbor $v$
• $D_v(y)$ = distance to $y$ as advertised by neighbor $v$

Distance-vector routing’s defining limitation is that it does not require global topology knowledge—routers “route by rumor” (neighbor information only), which makes it simpler but can lead to slower convergence and loop issues.

Keywords:

• distance vector
• next hop
• convergence

2) Why Loops Happen (Count-to-Infinity)

Because distance-vector routing relies on neighbor-reported information, a router can temporarily believe a destination is reachable via a neighbor when that neighbor also has stale or incorrect information. This can create routing loops where metrics keep increasing until they hit a protocol-defined infinity value.

RIP, for example, defines an “infinity” hop-count limit used to bound the count-to-infinity behavior (maximum metric is 15, and 16 is used as unreachable/infinity).

Keywords:

• count-to-infinity
• routing loop
• infinity

3) Loop Prevention Techniques (Split Horizon, Poison Reverse, Infinity, Timers)

Distance-vector protocols typically add mechanisms to reduce or mitigate loops:

Split Horizon

A router does not advertise to a neighbor routes learned from that same neighbor on that same interface, preventing simple “back-and-forth” advertisement loops.

Poison Reverse (Route Poisoning)

A stronger method: if a route learned via neighbor $B$ would otherwise be advertised back to $B$, the router explicitly advertises it as unreachable using an infinity metric. RIP-style behavior includes “poisoning” failed routes with an infinite metric.

Bounded infinity (e.g., RIP uses maximum hops)

By limiting the hop count, distance-vector protocols prevent metrics from increasing without bound. RIP’s metric range and infinity handling are specified in RFC 2453.

Timers (Timeout and Garbage Collection in RIP)

RIP maintains per-route timers:

• A route expires after a timeout period, but is kept temporarily to inform neighbors.
• A garbage-collection timer later removes the route entry.

FRR documentation also summarizes common RIP timers (e.g., update timer 30s, timeout 180s).

Keywords:

• split horizon
• poison reverse
• hold-down timer

4) Neat End-to-End Diagram: Message Exchange & Bellman–Ford Computation

Assume three routers A, B, C where B is between A and C. B receives A’s advertisement and C’s advertisement, then computes its own best distances and next hops.

This diagram corresponds directly to the distributed update described for distance-vector routing: neighbors exchange distance vectors, routers update their own distance vectors, and repeat until convergence.