RFID Precision/Antennae

[ptrack12]

I am working on a project with the following requirements:

ability to track up to 100 RFID tags carried by humans
ability to determine relative locations of RFID tags
precision of approximately 1 foot, at a distance of up to 100 meters
real-time exporting of reader data to a database located on a local server
My understanding is that I will require at least two antennae. Are there advantages to purchasing a reader with 4 antennae, such as accuracy or ambiguity? Can this be done with one antenna?

Does anyone have any recommendations of brands of tags or readers which may fulfill these requirements at a low cost?

There are discussions about UWB (ultra wide band) RFID technologies. Are these necessary to achieve the precision I require, or will simple UHF technologies be sufficient?

Any information in regards to best practices and/or limitations would be extremely helpful.

[mparadis]

I don't know a lot about UHF or Wide band RFID since we only sell 125 kHz tags and readers, but how do you plan on detecting the approximate location of the tag? It is my understanding that a tag is either in reading range, or outside of reading range, and this is the only information you can get in regards to its position. Even with two antennas, the only information you get is whether the tag is within range of one reader, the other reader, both readers, or neither. You may also have problems with the readers interfering with one another unless you poll them so that only one is active at a time.

Are you sure RFID is a better solution than GPS in this application?

[ptrack12]

I have read that using RSSI with two or more antennae can triangulate a position, and that readers have this built-in at times. Does this sound correct?

[mparadis]

From what I've read, using RSSI with RFID is a relatively new thing and the accuracy doesn't seem to be good enough for your 1-foot precision requirement. This is because there are a lot of factors that affect signal strength. For example, our RFID reader can read a typical card tag at a distance of 10cm if the tag's antenna is parallel to the reader. But if it's turned 90 degrees, the reader can hardly read it when the tag is touching. I don't think active tags have this problem to the same extent since they don't need coupling to power the tag, but I've read that active tags do not transmit evenly in all directions, so the orientation of the tag will still impact the RSSI.

Again, we don't deal with a lot of the active/UHF side of RFID so I don't know if someone has figured out how to get accurate position information for multiple tags in a field.