Hi, I think there is a way to increase the packet length (search the mailing list, it's there somewhere). Even so you'll run into a limit at some place.
Once the first message is full, I would send a second message with the additional path information (and once this one is full, another one, etc.). You might have to implement some sort of a send queue as you can't send two messages at once. Saving route information is also a possibility. Each node could periodically send its local routing information to the base station. If you want to query nodes, you'd best use another message type which you send from the base station to the desired node (assuming that the routing protocol allows this). Sending messages from a Java program is straight forward. To initially debug a program in my experience it is best to run it on a simulator. You'll even get debugging messages which you wouldn't have otherwise. Cheers, Urs DAE HEE KIM schrieb: > Thanks. > > As you mentioned, I am using array rather than list. It works. > > However, I have one concern for my code. > I am implementing multihop protocol, and I would like to see which > intermediate nodes forwarded packet data to a destination node. > For this reason, I wanted to add forwarding node ids into a packet. Then, I > can see paths by using LISTEN or MsgReader because the destination node is > baseStation. > > I know the maximum packet size is 29bytes. So, when testbed size become > larger (for example 100 x 100 grid nods), I might not be able to check the > path that packet was delivered to destination. > > In short, in this case, what is the best debugging or checking way ? > > PS) I came up with one idea although I don't know whether it is possible or > not. > 1) When a forwarding node forwards a packet, it saves the history into its > RAM. > 2) Afterwards, by using baseStation, I can send request for every node to > send saved history to baseStation by using ??? ( java program ?? or > otherthings, I am not sure) > 3) Every forwarding node should split save history information, and then > send to destination. > 4) When baseStation get the information, process the information, and then > analyze them. > > It was long comment. Could you comment about my idea? If possible, how > can I > do that? Thanks. _______________________________________________ Tinyos-help mailing list [email protected] https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
