0

I have researched about routing metrics for Wireless Networks. However, I have not come across "Path loss" as a routing metric. Since, Path loss is a very important performance factor for networks like Wireless Body Area Networks, is it feasible to use it as a routing metric?

AbbsLord
  • 1

1 Answers1

0

The main problem I see when using path-loss is that it might vary with the traffic load. You could see some path loss at a lower traffic and some other (usually higher) with higher traffic. So, unlike other baisc path characteristics like bandwidth and delay, path-loss might vary with the operating conditions.

However, there would be cases when the path would have some associated loss as compared to others. For example, let us say we have a micro-wave oven that can interfere with 802.11 traffic (those that live on 2.4GHz spectrum) and for such cases, it might make sense to use an average path-loss.

Manoj Pandey
  • 4,528
  • 1
  • 17
  • 18
  • In case Wireless Body Area Networks, where several nodes reside on/in body, path loss varies due to the human body shadowing & mobility of body parts. Does it make sense to use path loss as a routing metric? Because it can be an indication of how far a node is from the sink/source & how good is the channel quality. – AbbsLord Aug 26 '13 at 18:22
  • Your scenario seems to be the case, where loss might be incurred due to environment. For such cases, if it leads to a better route-selection, then using a path-loss metric would make sense. But you would still need to address the concern of path-loss variation. One way to do this would to correlate the loss to the traffic. For example with 802.11, we can see how much retransmission (if using a reliable transport like TCP) we had to do a for a given peer and at what traffic (loosely, inter-packet-gap). If such mechanisms are not possible, then one send occasionally a train of probe packets. – Manoj Pandey Aug 26 '13 at 18:50