For one of my major projects I’m tasked with figuring out how we can locate mobile wireless devices within a limited location, and it seems a lot of other people I’ve talked to are in the process of figuring out the same thing, so here’s an overview of what I’ve learned so far.

Wireless Technologies / Protocols

• WiFi (802.11)
• XBee/ZigBee (802.15.4)
• BlueTooth (802.15.1)
• RFID
• Near Field Communication (Two-Way RFID)

These are four of the principal wireless technologies in mainstream use today.  The goal is to determine within some degree of accuracy the position in 2 or 3 dimensions of a wireless node.  Most, if not all, of the available approaches utilize the RSSI (Received Signal Strength Indicator) of the target node’s connections to nearby nodes.

If you have fixed routers with known locations, you can use the RSSI values of their connections to the target node to calculate the distances from the fixed nodes to the target node.  However, any RSSI value will be heavily influenced by a number of factors that may be out of your control: interfering structures, interfering people (we are walking bags of water, which readily absorbs 2.4ghz radiation) and other interfering radio signals or noise.

Due to these issues, some smart folks developed a system called RADAR that uses ‘location fingerprinting’ to take a bunch of readings of signal strengths under varying conditions and associate them with known locations.  There are a variety of algorithms to tabulate this information, one of which uses neural networks.

Here is a collection of extremely useful papers and links that describe these various approaches, techniques and technologies:

Resources

• Survey of Wireless Indoor Positioning Systems [pdf]
• Indoor Location using 802.15.4
• Bluetooth Triangulation [pdf]
• CS/RADAR: Indoor Location Discovery and Tracking [pdf]
• Indoor location tracking using RSSI readings from a single Wi-Fi access point [pdf]
• Location Estimation in ZigBee Network Based on Fingerprinting [pdf]
• Experimental Analysis of RSSI-based Location Estimation in Wireless Sensor Networks [pdf]
• Position Location Monitoring Using 802.15.4/ZigBee technology [pdf]

Adam and I got our Bluetooth link in the Espionage chain working.

What you see here are two arduinos each with a SparkFun BlueSMiRF RP-SMA module. The transmitter (my end) is configured using Autoconnect Master mode, and the receiver (Adam) is in Slave mode (the default). With this configuration they pair automatically when within range of each other.

Once paired, I trigger the transmission manually and begin sending my message. His arduino receives the message and initiates its own sending procedure for the next link in the chain.

For our NetObjx Espionage class project, I am initiating a chain of data transfer with a collaborative “blow” interface that transmits over Bluetooth.So far my Bluetooth partner and I have managed to get our BlueSMiRF modules to pair (albeit over a short distance since we’re not using external antennae) and we even managed to get terminal-to-terminal bidirectional communication working!  Video here:

We’ve got our BlueSMiRFs connected to the TX and RX pins of some simple USB-to-Serial FTDI chips.  In order to get the modules to pair, I had to configure my BlueSMiRF module to use “Autoconnect Master” mode and set the destination address, which in this case was Adam’s module, which stays in its default mode of “slave.”  To configure a module, you can connect directly to it over Bluetooth with your computer.  Your OS should have some kind of Bluetooth utility that will detect nearby devices and allow you to connect to them.  Once you find yours (I’m not sure what the default names of these BlueSMiRFs are, mine was BlueRadios), try to pair with it.  It will ask for a password, which by default is, amazingly, “default.”  If it works, the utility will give you a serial port that allows you to talk directly to the Bluetooth module.

Now you need a terminal with which to talk and listen to the serial port (this is both similar and different to the “Terminal” program you have in OS X). I used ZTerm for this; with ZTerm you have to hold <shift> as it starts to select which serial port to talk to.  You can also use the “screen” shell program on *nix systems.  First list your serial connections and pick out the one that sounds right, i.e.:

ls /dev/tty.*
screen /dev/tty.BlueRadios-COM0-1

Once you’re talking to your module, you can try typing “AT<cr>” to test your connection, if it responds with “OK” you’re in command mode.  In my experience all commands are followed by a carriage return, so from now on I’ll leave that out.  You may have to type “+++” to get into command mode first.  The command references I looked at claimed the command mode string is “$$$” but none of the modules we used responded to this, weird.  Anyway, to get your module’s address, enter:

ATSI,1

It should hit you back with an address like this: 00A0961B284D.  If you want to set your module to Autoconnect Master mode, first you set the startup address and then set your startup mode:

ATSMA,<destination address>,1101
ATSW25,1,0,0,0

The “1101″ specifies serial communication.  You can change your startup mode back to the default with:

ATSW25,0,1,0,0

Once you’ve entered these commands, restart your circuit, and the module will begin trying to connect to the destination address you specified.  During this time your BlueSMiRF should be blinking green.  If you are monitoring it over serial you will see “CONNECT,<addr>”.  Once both modules pair, they will turn to solid red.  Without antennae, the RP-SMA modules will need to be within 2 feet or so of each other.  The silver and gold modules use better on-chip antennae so they’ll have much better range.

Now if both modules are being monitored with a terminal, you should be able to talk to each other!

I’ll follow up with video of my blow interface and hopefully soon we’ll get serial bluetooth data transfer working!