Digital Signal Processing Reference
In-Depth Information
Fig. 7.11
In a legacy
network, an SL terminal will
perform significantly above
average. With the
introduction of more SL
terminals, the average
throughput of SL terminals
begins to decrease as less
terminals can now be
exploited. However, a full SL
network still outperforms a
full legacy network
7.5 Conclusions
In this chapter we have presented a throughput model for large 802.11 networks.
This illustrates that the DT/RT method can handle very complex environments for
CRs. The proposed model is shown to match simulation results very accurately, and
hence gives the correct insight in the relative importance of the different starvation
sources on the expected throughput of large networks.
This model is however too rigid for RT control of IEEE 802.11 terminals. Hence,
we have presented a hybrid DT/RT control algorithm for tuning power, rate and car-
rier sense threshold. It is shown to outperform the reference algorithm Spatial Back-
off in network-wide throughput and power consumption. We believe distributed
optimization is a key topic in wireless communications as it is able to cut down
infrastructure and management cost. This is key for CR.
IEEE 802.11 has been selected as the supporting standard, since it is the most
prominent wireless standard that enables ad-hoc communication. In the future, we
expect a synergy between purely distributed optimization based on local information
and distributed optimization based on information exchange between the terminals.
