Forums

Hi,

cc2530 has no location engine.

Hi Igor ,

Thanks for reply.

Yes you are right but, Is there any way to run z-Location engine using CC2530 ?

Thanks

sam

Hi Igor,

Back to you again.

For Location using CC2530, consider the following simple scenario:

1- A blind node periodically broadcasts a request to reference nodes ( distributed in known places)

2- Reference nodes respond with their RSSI vaues

3- Blind node calculates its distance from all reference nodes and then sends this info to the coordinator

4- Coordinator (connected to a PC) sends this info to a visualized application to indicate the blind node location on a simple map.

Note: Accuracy required here is just to say which is the nearest reference node to the blind node.

Questions:

 Is this scenario makes sense ?

I am thinking of modifying sensor Demo for that, am I right ?

Thanks for your time

Hi,

Samih Eisa

2- Reference nodes respond with their RSSI vaues

A blind node is probably a ZigBee end device (ZED). The parent of this ZED will handle the 

routing on its behalf, which means that you won't be able to get true RSSI values from reference

nodes since the first node to get these response messages will be the parent router, not a blind node.

Here are several suggestions to improve your scenario:

1. Blind node is considered as a battery powered node, therefore you probably should broadcast a dummy
   message to nearest reference nodes (which are probably a main powered routers). Then each router will
   send RSSI value as a unicast message to coordinator (or any other node which is used as gateway to your
   GUI application).
2. Since cc2530 is a slightly modified (enhanced) 8051 microcontroller, you might consider to do the math on a
   PC, or some other device with a higher "horse power".
    

Hi Igor,

Thanks for your inputs.

I agree there are several pitfalls in the scenario. However, let's consider your suggestions and assume the following:

1- Blind node periodically sends a broadcast message. Then each router sends RSSI value to coordinator

2- Coordinator forwards this info to a PC for location calculation

Pitfalls 

1- potential avalanche effect of too many routers trying to send unicast message to coordinator at once

- Assume the intended location area is a small house. Therefore, a few routers will be deployed to cover the entire house.

- random jitter in each router seams as a good suggestion

2- How much time to wait before deciding that the RSSI values from all the participating routers (reference nodes) have reached the application?

- Real question !!!!- I guess reasonable time should be given proportional to the jitter time in each router. I need to be more specific in this part. any advice ?

3-  Three reference nodes (ZRs) are located within the broadcast area of a blind node (ZED). Where one ZR is relatively far from ZED, the second ZR is in the middle range and the third ZR is very close to ZED, but the first and second ZRs are in an object free line of site to ZED and there is a concrete wall between the third ZR and the ZED? By measuring the RSSI values there is a very high probability that the  second ZR will be identified as the closest reference node instead of the third.

Yes most likely to occur. RSSI is not always the best criteria for location estimation. any  suggestion ? 

Is the sensor Demo a good starting point for me ?

Thank you

Sam

Hi,

Samih Eisa

2- How much time to wait before deciding that the RSSI values from all the participating routers (reference nodes) have reached the application?

- Real question !!!!- I guess reasonable time should be given proportional to the jitter time in each router. I need to be more specific in this part. any advice ?

The jitter time  + polling time of USB (gateway) + computation time on the application side

(is it a synchronous or asynchronous application)

Samih Eisa

3-  Three reference nodes (ZRs) are located within the broadcast area of a blind node (ZED). Where one ZR is relatively far from ZED, the second ZR is in the middle range and the third ZR is very close to ZED, but the first and second ZRs are in an object free line of site to ZED and there is a concrete wall between the third ZR and the ZED? By measuring the RSSI values there is a very high probability that the  second ZR will be identified as the closest reference node instead of the third.

Yes most likely to occur. RSSI is not always the best criteria for location estimation. any  suggestion ? 

One possible solution is a smart application. For example, mapping of the area, in this case your

application can actually learn the topography of a house and be aware of different obstacles that are

near/close to reference nodes. This solution adds a very high level of complexity to your application.

There is no magic solution for this kind of behavior, if you have another idea or a question about this

one, feel free to elaborate/ask.

Samih Eisa

Is the sensor Demo a good starting point for me ?

You can start with it, if it doesn't fit your purpose you can always switch to any other example,

it should be pretty quick and painless :)