Distributed, not centrally assigned
Each node learns from nearby messages and local link feedback. A gateway may exist, but it does not have to assign every route.
Ad-Hoc Network Self-Organization Workbench
Step through neighbor discovery, link scoring, route selection, packet forwarding, and local repair in an infrastructure-less network.
animation
adhoc
manet
architecture
routing
Interactive ad-hoc network self-organization animation for learning MANET neighbor discovery, distributed route selection, multi-hop forwarding, route repair, and protocol trade-offs.
Ad-hoc networks
Self-organization
MANET routing
Ad-Hoc Network Self-Organization Workbench
Watch a wireless network form without a fixed access point: nodes announce themselves, score usable links, choose a multi-hop route, forward data, then repair the path when a mobility event breaks a link.
Power on
current self-organization phase
-
selected route
-
estimated delivery success
Link quality
route decision policy
Power-on state
Nodes are awake, but each node only knows itself.
discovery
selected route
broken link
Step through the phases to see the network build knowledge locally before it sends data.
0
visible neighbor links
-
route hops
-
route ETX estimate
0 frames/min
control overhead lens
What the learner should notice
Ad-hoc networking starts from local observations. There is no access point handing every node a complete map.
Selected route and local table
No route is selected yet. Step to route selection to compare the policies.
| Neighbor | Link quality | Energy | Local decision |
|---|
Event log
What this animation abstracts
Link quality can beat shortest path
A two-hop path with weak radio links can be worse than a three-hop path with stronger, lower-loss links.
Repair is part of normal operation
Mobility, fading, and battery depletion change the graph. The network detects the change and searches for a usable alternate path.
Protocol families in context
| Family | Routing behavior | Best fit | Trade-off |
|---|---|---|---|
| Proactive | Maintains routes before traffic needs them. | Dense or steady traffic where first-packet delay matters. | Higher periodic control overhead. |
| Reactive | Discovers routes on demand, then maintains active routes. | Sparse, bursty, or highly mobile traffic. | First packet waits for discovery. |
| Hybrid | Keeps nearby routes ready and discovers distant routes only when needed. | Larger networks with local clusters and occasional distant traffic. | Adds zone-management complexity. |
Quick reference and sources
- Neighbor discovery: a node learns which peers are directly reachable and records link evidence such as recent reception, RSSI, loss, or timing.
- Route selection: the best next hop depends on the routing objective. Hop count is simple; link-quality and energy-aware choices often fit IoT better.
- Route repair: when a link breaks, affected routes are invalidated or rediscovered. The packet path can change while the application goal stays the same.
- Scope note: this is a teaching abstraction across MANET-style protocols, not a complete implementation of one RFC.
Practice prompts
- Change the route policy from link quality to hop count. Which scenario changes path, and why?
- Turn off the mobility event, then step to repair. What maintenance work remains even when no link breaks?
- Select a node on the map. Which neighbor would you avoid if battery life matters most?
- For a fixed sensor field with one report every minute, would you prefer proactive, reactive, or hybrid routing?