Digital Signal Processing Reference
In-Depth Information
Fig. 4.1
Spatial reuse in wireless networks requires high sensitivity receivers and moreover never
achieves optimal adaptation to the real propagation conditions, since safety margins are needed to
avoid interference to the potential receivers with unknown channels. We want to achieve optimal
spatial reuse (i.e., without safety margins that limit the gain), while relaxing the receiver sensitivity
constraints
opportunistic radio. The discussed technique could be used to communicate actual
run time propagation contours of (TV) transmitters and improve on the overly con-
servative power estimates based on design time assumptions.
The challenges of efficient spatial reuse are also illustrated by means of mea-
surements of an outdoor IEEE 802.11 network where one Access Point (AP) is
considered to be a primary AP and a second AP should avoid interfering with it.
The technique relies on the availability of nodal positioning information which
could be obtained through GPS measurements or a localization algorithm. Future
radios are expected to have geolocation capabilities, so this assumption does not put
additional constraints on the cognitive radio. When determining the optimal power,
a limited amount of interference from control messages is tolerated.
4.1.2 Smart Aspect
The cognitive aspect of the proposed technique lies in the fact that the environment
is monitored extensively. The environment in this case is the distance to the primary
user receive contour, or more importantly the actual pathloss that is a function of
distance but also shadowing. Within that contour, no secondary interference can be
tolerated. The secondary network will iteratively adjust it's transmission power until
the coverage of the secondary network is within a certain predefined bound of the
primary contour. If overlap of the contours is determined, the power will of course
be decreased. This means three scenarios are possible: decrease output power, keep
output power or increase output power. For each scenario, the corresponding action
is then taken until the algorithm converges. This is illustrated in Fig.
4.2
.
The approach also relies on a DT model of pathloss and shadowing, that is then
calibrated at run-time to determine the actual shadowing, distance and pathloss ex-
ponent between the cognitive radio and the primary transmitter. Since explicit feed-
back is not possible from the primary transmitter (considered to be legacy tech-
nology), the information for this calibration is determined based on the monitoring
