Digital Signal Processing Reference
In-Depth Information
Fig. 5.2
The considered scenario is a string topology of IEEE 802.15.4 terminals, where terminals
report to a sink that is placed at one side of the string. Interference is generated by WLAN devices
and is dynamic in both time and space
R
2. Each of the sensor terminals is currently operating in one of the
F
possible
channels. As explained above,
F
=
16 for IEEE 802.15.4 networks operating in the
2.4 GHz ISM band. An
n
-dimensional vector
t
n
keeps track of the current frequency
f
i
(
i
∈[
=
), which the different terminals are using to transmit. Every termi-
nal transmits its beacon at this frequency, according to the beacon-enabled mode.
We assume that the sensors potentially swap frequency every inter-beacon period,
which ranges from 15 ms to above 4 min according to the standard. When and how
to swap will be determined by the distributed adaptation algorithms.
1
,...,F
]
5.2.2 IEEE 802.11 Interference Model
We consider a large IEEE 802.15.4 network, that is affected by IEEE 802.11 in-
terference. This IEEE 802.11 interference is assumed to vary over time, space and
frequency.
The variations in frequency result from the fact that different IEEE 802.11 net-
works can operate simultaneously using a different channel. Dynamic Frequency
Selection (DFS) is a new functionality currently added to most of those IEEE 802.11
networks and Access Points. It is developed to optimize the frequency alloca-
tion of IEEE 802.11 networks that are subject to interference, which can be self-
interference or interference resulting from other networks or devices. As a result of
this adaptive and hence dynamic frequency selection, IEEE 802.11 networks are not
operating at fixed channels over time.
It can be seen in Fig.
5.1
, that each IEEE 802.11 network always covers four
consecutive channels of IEEE 802.15.4 networks. The power distribution over the
channel is more or less flat, especially for the OFDM-based IEEE 802.11g net-
works [59].
