Digital Signal Processing Reference
In-Depth Information
Chapter 5
Coexistence: The Whole Is Greater
than the Sum of Its Parts
5.1 Introduction
We have argued in Chap. 3 that RT solutions based purely on DT models are inef-
fective. In a complex environment solutions need to be learned at RT. In this chapter,
we build further on this observation. We focus on the horizontal sharing of two stan-
dards that operate in the 2.4 GHz ISM band: IEEE 802.11g Wireless LAN [59] and
IEEE 802.15.4 Sensor Networks [60], illustrated in Fig.
5.1
. As mentioned in the
previous chapter, in the ISM bands wireless spectrum sharing is the most flexible
and, hence, the most difficult to control. Though the IEEE 802.11 is the most pop-
ular standard and intra-coexistence is important (also see Chap. 7), the ISM bands
are horizontally shared by many other devices like IEEE 802.15.4 sensor terminals.
From Fig.
5.1
we can see that, while these two standards have equal rights to the
spectrum, the characteristics of both standards are very different and the problem
turns out to be asymmetric in nature.
Indeed, the output power of IEEE 802.15.4 devices is usually as low as 0 dBm
[61], while the output power of IEEE 802.11g devices is typically above 15 dBm
[62]. Also, IEEE 802.15.4 sensor networks are designed to monitor the environ-
ment or buildings, and are typically very large. IEEE 802.11 networks are typically
hotspots centered around an Access Point (AP), and hence more local. Finally, sen-
sor network applications are not demanding in terms of throughput, but however
require a high reliability and robustness against attacks or unknown events. They
should also be self-organizing since it is impossible to maintain such large net-
works manually. IEEE 802.11 networks are typically used by a limited number of
throughput-intensive applications.
Although recent studies have shown that sensor networks do have an impact on
IEEE 802.11 networks [63], it is well accepted that coexistence of both networks af-
fects the sensor networks most [64]. In this chapter, we design distributed protocols
for IEEE 802.15.4 networks to optimize their performance under varying interfer-
ence. The proposed algorithms should be distributed in order to improve scalability,
robustness and adaptability. More specifically, we design distributed channel selec-
tion algorithms that allow the sensor terminals to dynamically adapt their channel
