WSN Routing Tables and best path determinations

[Garrett Cook]

Hello,

My team has just finished placing a total of 10 nodes in an industrial factory setting.  Nine are set as routers and there is one end node.  The gateway is somewhat in the center of this area and there are no stranded nodes.  The area of coverage is four floors with approximately 63000 sqft of area per level.  I am having an issue with one section of the path used to reach the end node: a high-quality signal between between the 3rd and 4th routers from the gateway is being dropped in favor of a much lesser quality signal.  If I take that 4th router and move it very close to the 3rd router and do a reset it will pick up the 3rd router's signal, and remain at about 60% signal strength for several minutes.  However, this 4th router will soon suddenly drop to about 20% signal strength (much less than NI's recommended 30%).  Is the network deciding that 3 hops of relatively good signal is worse than one hop involving a poor signal?  Also, is there a way to read what the routing tables are to determine if this is indeed the case?

Thanks!

Garrett

[Brian_A]

Hi Garret,

The WSN devices auto configure the network mesh they create and we recommend that you use no more than 3 hops to get to a gateway but there is no way to really ensure a certain mesh unless it is set up so it can only take one configuration. The nodes dynamically asign the mesh based on the current conditions so it's hard to say exatly why it is choosing to skip over some gateways. At this time there is no way of accessing the routing tables for these devices. 

Are you using the 9 routers sequentialy or have you simply not set up all the end node? 

You may have read this already but here is an article about Wireless Sensor Network Topologies

<Brian A | Applications Engineering | National Instruments> 

[Garrett Cook]

We were able to use AC power for all but the end node, so we set them as routers to provide redundant paths.  The physical placement of our nodes looks comparable to the star topology diagram in the article link you posted.  However, few of our nodes are placed close together and have significant obstacles in the way (most of our signal strengths are between 40 and 60%).  Even if no nodes can be "stranded", is it more reliable to have fewer possible paths than to have an "inefficient" network?

Garrett

[Brian_A · Accepted solution]

We suggest that you use the least amount of Routers needed for your applicatioin. Routers add the chance for decreased network performance due to uninteded hopping. So even when AC power is available, the default choice for the nodes should be End nodes unless Routers are needed.

<Brian A | Applications Engineering | National Instruments>