The 802.11 Hidden Node Problem

Apparently, the CSMA/CA method used by 802.11 networks has a problem when there are multiple nodes accessing a base station that can’t see or hear each other. When they transmit at the same time. The “popular” solution appears to be RTS/CTS, but apparently this doesn’t work all that well in crowded networks. Joy.

Russian explanation and analysis of the problem

http://www.securitymagazine.com/CDA/ArticleInformation/features/BNPFeaturesItem/0,5411,77206,00.html – Better english explanation of the phenomenon, as well as Karlnet’s Turbocell. Wonder if Turbocell is licensed for use in Linksys or Netgear products… Hm….

http://www.pbs.org/cringely/pulpit/pulpit20010628.html – Funny but relevant story about a guy and his mission for high speed internet. Also mentions hidden node problem and Doug Karl.

http://www.karlnet.com/products/software/wlTurboCellHiddenNode.html – From the horse’s mouth, Karlnet’s Turbocell solution.

http://www.airmagnet.com – Site survey tools and training. Nice.

2 thoughts on “The 802.11 Hidden Node Problem”

I want to ask a question ? the question is.. What is ” Hidden node problem “? please define clearly.

Admittedly, I had a hard time determining this as well. From the research I did, here’s a practical example that I’ll try to explain with. Imagine a simple 802.11g network with a single access point at point A. There are two nodes, N and P, which both are connected to the access point at A. However, N is 150 ft. west of point A, and P is 150 ft. east of A.

N –(150 ft)– A –(150 ft)– P

In this network, N can communicate with A, and P can communicate with P, but N and P are out of range of communications with each other, or in other words, the P node is “hidden” from N, and vice versa.

The problem arises when, in a CSMA/CA controlled network, N first checks the “line” (checks on the 2.4Ghz spectrum) to see if there is communication currently going on. This is the “Carrier Sense” part going on. It doesn’t see anyone else communicating on the “line”, so it goes ahead and begins communications with the access point at A. This is a problem if the P node is communicating with the access point at the same time.

That’s my understanding of the hidden node problem. Solutions and additional discussion are linked above.

M.S.I. Polash on February 18, 2004 at 9:59 pm said:

I want to ask a question ? the question is.. What is ” Hidden node problem “? please define clearly.
Gordon on February 22, 2004 at 8:45 pm said:

Admittedly, I had a hard time determining this as well. From the research I did, here’s a practical example that I’ll try to explain with. Imagine a simple 802.11g network with a single access point at point A. There are two nodes, N and P, which both are connected to the access point at A. However, N is 150 ft. west of point A, and P is 150 ft. east of A.

N –(150 ft)– A –(150 ft)– P

In this network, N can communicate with A, and P can communicate with P, but N and P are out of range of communications with each other, or in other words, the P node is “hidden” from N, and vice versa.

The problem arises when, in a CSMA/CA controlled network, N first checks the “line” (checks on the 2.4Ghz spectrum) to see if there is communication currently going on. This is the “Carrier Sense” part going on. It doesn’t see anyone else communicating on the “line”, so it goes ahead and begins communications with the access point at A. This is a problem if the P node is communicating with the access point at the same time.

That’s my understanding of the hidden node problem. Solutions and additional discussion are linked above.