Advice needed selecting propper xbee radio

XBee & XBee Gateway XBee - DigiMesh
1

hi my name is mile,

i’m doing a research for my masters and seeing i have no experience with radios, i need an advice on what kind of radio modules to use for my project.

NOTE:
-i didn’t just opened this thread blindly, as first I took a look around the web manufacturers web site, the forum and plenty other sites as well. however when a know little about a certain topic, i always ask help from people who are more competent than me and whose knowledge in this ‘unknown area’ is greater than mine.

GOAL:
-to find a wireless radio with 400m range for the sensor nodes that i’m going to use for the fire detection system.

USE CASE:
-the sensor nodes will be placed inside the forest and they will have smoke sensors attached to them in order to detect potential fire.
-once smoke is detected, the sensor node will generate a warning and he will pass it to his neighbor, who will pass it on to the next sensor node, and it will go all the way to the base station. the signal from the base station will go throgh wireless backhaul, but that’s another topic that currently doesn’t need to be discussed

REQUIREMENTS:
-the maximum distance between two sensor nodes will be 400m at most
-the sensor nodes will be put in a pinetree forest where the trees grow upto 40m
-the sensor nodes should be able to communicate with all of their neighbor-sensor nodes that are within the 400m radio range
-the sensor node is actually a microcontroler development board with ARM7 LPC2368 processor and the radio should be able to communicate with it. here’s the board’s specifications >> http://www.micro4you.com/files/lpc2368/ARM7_LPC2368.pdf

SOLUTION:
-tell me i’m doing overkill with the 400m reqirement. can i use certain antenna to get to the 400m mark? what if there’s a sensor node that will be even further away, let’s say 1.5km
-i’m inclining towards this choice, but i have no idea how will the forest trees affect the radio transmission >> http://www.digi.com/products/wireless-wired-embedded-solutions/zigbee-rf-modules/zigbee-mesh-module/xbee-digimesh-900#specs

FINAL NOTE:
-if you have better ideas please share.
-if you know other place where i can ask for help, let me know in your replies
-if you have other more appropriate manufactures in mind, please share. (i’m not sure if i’m allowed to ask this, but i hope it won’t get someone mad) what are some other companies that are xbee’s competitors on the market

P.S. i do most of my shopping on ebay, but i think other specialized web sites would have better prices than ebay in this case. Please recommend me some websites that sell radio equipment and have low prices as well.

thanks. mile

2

Also keep in mind that wet pine trees are perhaps the worst vegatation for radio signals! So you will want signal overkill at lower frequencies (900Mhz is better than 2.4GHz) … although forest fires within wet trees isn’t so common :slight_smile:

Timing and battery life will be important to look at. DigiMesh (DM) allows sleeping routers, but of course they don’t talk or route while sleeping. So if you have them wake every 60 seconds for 4 seconds … you need to consider the power requirements. The longer the sleep, the longer the battery life but the higher the lag/latency for a detection.

During a detected event, you could just use a DM broadcast and you’d have latency up to 1 minute. You’d also want the nodes to periodically (but not once a minute) report back that they are healthy via Unicast. I would use a protocol to assign the nodes to a ‘group’, so the group would report health togather and you could prevent them from bunching up and having too much data to fit in your short ‘awake’ time slice.

As for antenna, for the ‘omni-directional’ antenna, the signal is much like a donut skewered by the antenna. You can use ‘higher gain’ antenna, which just mean the donut is flatter, but bigger around. The problem with this is vertical changes between peers would now have a much worse signal. So if your trees are in a flat area, higher gain would give better performance … but if the trees follow hills or even cliffs, you’d have trouble going ‘up/down’ hill.

3

Also keep in mind that wet pine trees are perhaps the worst vegatation for radio signals! So you will want signal overkill at lower frequencies (900Mhz is better than 2.4GHz) … although forest fires within wet trees isn’t so common :slight_smile:

Timing and battery life will be important to look at. DigiMesh (DM) allows sleeping routers, but of course they don’t talk or route while sleeping. So if you have them wake every 60 seconds for 4 seconds … you need to consider the power requirements. The longer the sleep, the longer the battery life but the higher the lag/latency for a detection.

During a detected event, you could just use a DM broadcast and you’d have latency up to 1 minute. You’d also want the nodes to periodically (but not once a minute) report back that they are healthy via Unicast. I would use a protocol to assign the nodes to a ‘group’, so the group would report health togather and you could prevent them from bunching up and having too much data to fit in your short ‘awake’ time slice.

As for antenna, for the ‘omni-directional’ antenna, the signal is much like a donut skewered by the antenna. You can use ‘higher gain’ antenna, which just mean the donut is flatter, but bigger around. The problem with this is vertical changes between peers would now have a much worse signal. So if your trees are in a flat area, higher gain would give better performance … but if the trees follow hills or even cliffs, you’d have trouble going ‘up/down’ hill.

can’t i leave the nodes in sleep mode all the time, and have them wake up only when the sensor detects smoke? but they will wake up periodically to report on their help status like you suggested

i am planing to use this radio module >> Wireless Mesh Networking RF Module | Digi International
XBee-PRO® DigiMesh® 2.4

4

can’t i leave the nodes in sleep mode all the time, and have them wake up only when the sensor detects smoke?

Well, sure. You can use an external PIC or signal to ‘wake’ the module.

But then you do NOT have a mesh. Your lone module is now awake with no peers. It will require line-of-sight back to some fully powered device to receive the message.

In this case, you’ll want some high-power 900Mhz non-mesh product which can punch back a dozen miles - not Digi Mesh.

For a mesh, you need all nodes to wake up at the same time & help ‘mesh’.