This project is a simulator of the communication between a base station and a fully distributed wireless sensor network.
The wireless sensor network comprises 20 nodes and a base station. These nodes are arranged in a 4 x 5 grid.
In this WSN Project, assuming that a set of 21 MPI processes represents the whole WSN and each MPI process maps a certain node in WSN. There’re totally 21 processes required to run the application. The ‘rank 0’ MPI process represents the base station while ‘rank 1’ – ‘rank 20’ MPI process represents all network nodes.
There're typically two communication models in the mesh data structure.
communication between Adjacent node →Adjacent node. The program applies blocking(synchronous) communication between all adjacent neighbor Nodes. The following figure shows the simulation:
From this communication model, each node can only communicate with the adjacent nodes directly (i.e. top node, right node, bottom node, left node), but cannot chat with non-adjacent nodes (i.e. diagonal nodes).
The message passed between each adjacent node are random generated number in the range of 10, 11 or 12. There’re 3 categories of node which has been clarified in the program structure part.
For a specific node, for example, Node 1 generates an random number and sends the message(random number) to it right neighbor – Node 2 and its bottom neighbor – Node 6. Also, at the same time, Node 1 receives random number from Node 2 and Node 6 as well. For another instance, Node 7 generates an random number and sends the number to its neighbor – Node 2, Node 8, Node 12 and Node 6. Simultaneously, it receives random number from these four nodes.
The whole procedure would take 10 iterations (rounds). Between each round, there’s 1 second’s interval to make sure every processes can finish the task.
On Model 1, communication uses MPI_Sendrecv() which send and receive messages from adjacent neighbor.
communication between Base station →Node.
The program applies non-blocking (asynchronous) communication between base station and each grid Node. The following figure demonstrated the communication between base station and grid nodes when event triggered.
In terms of Model 2, whenever there is an event firing in a node in the network, it reports a group of 6 messages, which are the iteration, the random number, top node random number, right node random number and left node random number, to the base station(Rank 0).
In figure 2, as an example, in a specific time interval round, Node 7(Rank 7) and Node 17(Rank 17) received at least 3 repeated random number from the neighbor nodes. When the event triggered, the two nodes non- blocking(asynchronous) send message packets to base station (Rank 0). In comparison, all other nodes that don’t receive at least 3 repeated random number won’t send messages to base station, which increases the efficiency of the wireless sensor network message passing system.
On Model 2, the communication is taken play by the following MPI operations:
MPI_Isend(): non-blocking send message from nodes to base stationMPI_Irecv(): non-blocking receive message by base stationMPI_Wait(): wait for MPI request to completeMPI_Test(): test whether a request has been receivedMPI_Cancel(): cancels a communication request from node to base stationMPI_Request_free():frees a communication request object
-
open terminal
-
input commands
make clean
make
make run
- see results
cat base_station.log
cat message_stat.log