User's Manual

User Manual
3000 SERIES 3
rd
GENERATION ACTIVE RFID
The aforementioned bottleneck is a typical Infrastructure bottleneck and can easily be avoided by
placing additional MRs around the problem area.
Avoid Infrastructure bottlenecks by all means.
7.2.2 SP Load Bottleneck
Another type of bottleneck can be introduced by SPs, but these are a bit harder to find and depend
heavily on the configuration of an SP. By default, a SP3000 will send a message every 5 minutes. By
default, a GW and MR can handle 2 SP interactions per second
1
. As a result you should then be able
to place 60*5*2 = 600 SPs around a single Infrastructure node. However, packet loss can occur, SPs
can wake up at the same time, and so forth. As a rule-of-thumb, it is advisable not to exceed the load
of a GW or MR with more than 50 SPs in its range. Your Infrastructure deployment needs to
incorporate this SP load and provide enough bandwidth to handle the SP payload. A way of
determining the network load is described below.
We determine the load of a GW or MR, indicated as the 'SP Load', by determining how many SPs are
in range of an Infrastructure node. With this approach we assume that every SP is configured
similarly, i.e. all SPs transmit the same number of packets within the same time period. A single SP
provides an SP Load of 1. If an SP is in range of more Infrastructure nodes, we calculate the load by
dividing 1 by the number of Infrastructure nodes in its range. We can now determine the SP load of
an infrastructure node by summing the SP loads of the SPs in its range. For the Infrastructure nodes
we can now determine an accumulative load by summing all the SP loads of Infrastructure nodes that
are one hop further away. As with the SPs, depending on the number of Infrastructure nodes in
range the SP load is divided by the number of Infrastructure nodes that are one hop closer to the GW
(see Illustration 28).
1 This can be change to 4 by configuring the appropriate driver, driver network:DNCP
48/73 Ambient Systems B.V.