Distributed power control and spreading gain allocation
in CDMA data networks
Abstract
We study the radio resource allocation problem of distributed joint transmission power control
and spreading gain allocation in a DS-CDMA mobile data network. The network consists of
K base stations and M wireless data users. The data streams generated by the users are treated
as best-effort traffic, in the sense that there are no prespecified constraints on the quality of the
radio channels. We are interested in designing a distributed algorithm that achieves maximal
(or near-maximal in some reasonable sense) aggregate throughput, subject to peak power constraints.
We provide an algorithm where neighboring base stations coordinate in a distributed
fashion to control the powers and spreading gains of the users, and show that it converges to
a Nash equilibrium. In general, there may be multiple equilibria; however, certain structural
properties of the throughput expression can be exploited to significantly trim the search space.
The numerical results indicate that with these modifications, the algorithm frequently converges
in just a few iterations to the throughput maximizing (globally optimal) power and spreading
gain allocation.
1 Introduction
The tremendous success of voice-based cellular telephone networks has spawned an increasing interest
in mobile wireless data communication. Indeed, a number of wireless data networks have recently
emerged in the marketplace spanning the domains of cellular networks (e.g., CDPD, GPRS, and
EDGE) to metropolitan, local area, and ad hoc networks operating in the ISM band1 (e.g., ARDIS,
Metricom Richochet, Cisco Aironet, AT&T WaveLAN, Utilicom Longranger, Rooftop, Bluetooth,
and HomeRF). As compared to voice traffic, data traffic is less senstive to delays, but more sensitive
to transmission errors; reliability is assured via retransmissions. From a network layer perspective,
data communication is often treated as a best-effort service, and therefore achieving high aggregate
data throughput is one of the primary goals. Towards this, the delay tolerance of data traffic can be exploited to allow the design of spectrum-efficient radio resource allocation algorithms adapting
the received energy per bit (ratio of the received power per bit to the transmission rate) to the
current interference and channel conditions.
In this paper, we study the radio resource allocation problem of distributed power control and
spreading gain (transmission rate) allocation in a Direct-Sequence Code Division Multiple Access
(DS-CDMA) mobile data network. The network consists of M wireless data users and K stationary
base stations (access points). At any given point in time, each user is connected, that is, assigned,
to at most one base station; however, each base station is capable of simultaneously receiving
transmissions from several different users. The users generate non-real-time streams of data packets
(e.g., paging, electronic mail, facsimile, file transfer, etc.), and transmit to appropriately chosen base
stations using different spreading codes over a common wideband radio channel. We are interested
in designing a joint transmission power control and spreading gain allocation algorithm for the users
that achieves maximal (or near maximal in some reasonable sense) aggregate throughput, subject
to peak power constraints. We focus on the uplink (user-to-base station), but the downlink can be
treated similarly.
In terms of maximizing aggregate throughput, or any other global network-wide objective, it
is best for the base stations fully coordinate in order to determine the optimal allocations; but
this is, of course, not practical. In [8], a radio resource allocation algorithm is provided for a
DS-CDMA network in which the base stations act autonomously, that is, the base stations do not
coordinate (except for soft handoffs, etc.), to control the transmission powers and spreading gains
of the users. Each base station attempts to maximize the aggregate throughput of just the users
in its cell, subject to constraints on peak power and, to prevent excessive intercell interference,
total received power. The simulations of the algorithm presented in [8] show that it outperforms
other autonomous resource allocation schemes, such as equal received power and round robin, with
respect to throughput and average delay, even for users with low quality channels.
With a moderate increase in complexity, even higher performance gains would be expected by
allowing a limited amount of coordination between base stations. The basic idea is that light loading
or low interference in neighboring cells can be exploited via the exchange of information between
neighboring base stations to achieve a higher network spectral efficiency than can be achieved via
less complex greedy localized maximizations by each base station separately. In this paper, we are
interested in developing an algorithm where neighboring base stations coordinate in a distributed
fashion to jointly control the transmission powers and spreading gain allocations of the users.
The problem of distributed power control in wireless communication networks has received
considerable attention in the past. For a network in which the QoS requirement of each user is
a prespecified target Eb/I0 value, Hanly [1] and Yates and Huang [2] have developed distributed
power control and base station assignment algorithms based on interference and channel measurements
local to the base stations, that minimize the total transmitted power subject to target Eb/I0
values, and provided conditions for their convergence. Grandhi et al. [3] have developed a distributed
power control algorithm that maximizes the minimum Eb/I0 value of the users, subject to
constraints on peak transmission power. For a network in which each user generates a bursty traffic
stream, and specifies its QoS requirement in terms of the probability that the Eb/I0 falls below a
target level, Mitra and Morrison [4] have developed a distributed power control algorithm based
on measurements of the mean and variance of the interference at the base stations, and provided
conditions for its convergence.

Download full report
http://citeseerx.ist.psu.edu/viewdoc/dow...1&type=pdf