Recently I was doing some tests in a small wireless ad-hoc network and one interesting feature I've found was:
- Suppose you have a hidden node problem among computers A, B and C (A and C are out of range from each other, but B can communicate with both). If A initialize the ad-hoc network and then B joined it, after that C would also be able to join it through B.
At first glance you might think that, as C could join the network created by A, probably C could also communicate with A (of course in a true ad-hoc network it doesn't matter who initiated it, this is just a fail-safe test). In that case, it means that this network has a multi-hop feature. However, this is not the case. Nodes can only communicate with other nodes within WiFi range, therefore, they don't participate in routing or forwarding data.
1st Issue
According to Wikipedia, in a wireless ad-hoc network "each node participates in routing by forwarding data for other nodes". How do we reconcile this definition with the current ad-hoc network of Windows? Technically, is it correct to call a single-hop network by ad-hoc? What features can we expect from the Windows Ad-hoc Network?
2nd Issue
What would be the technical challenges to develop a protocol for multi-hop ad-hoc network (if there is any)? Current computers would be too slow to forward data based on routing tables???
3rd Issue
If your computer has an Ethernet connection with the Internet and you setup a wireless ad-hoc network, you may share your internet connection within nodes in the ad-hoc network. However, the other nodes only in the ad-hoc network are not allowed to share such internet connectivity (a popup window appears complaining about a problem with automatic IP address).
What are the technical challenges to allow all nodes in an ad-hoc network to share internet connectivity (besides the multi-hop issue aforementioned)?
