In this paper we consider distributed power control in a multicell wireless
data system and study the effect of pricing transmit power. Drawing on our
earlier work in [1], we formulate the QoS of a data user via a utility fun
ction measured in bits per Joule. We consider distributed power control, mo
deled as a noncooperative game, where users maximize their utilities in a m
ulticell system. Base station assignment based on received signal strength
as well as received signal-to-interference ratio (SIR) are considered joint
ly with power control. Our results indicate that for both assignment scheme
s, such a procedure results in an inefficient operating point (Nash equilib
rium) for the entire system. We introduce pricing of transmit power as a me
chanism for influencing data user behavior and our results show that the di
stributed power control based on maximizing the net utility (utility minus
the price) results in improving the Pareto efficiency of the resulting oper
ating point. Variations of pricing based on global and local loading in cel
ls are considered as a means of improving the efficiency of wireless data n
etworks. Finally, we discuss the improvement in utilities through a central
ized scheme where each base station (BS) calculates the best SIR to be targ
eted by the terminals it is assigned.