Abstract:
With increase in demand for wireless data services, it becomes very essential to
manage scarce radio resources. Especially in CDMA systems, where all the users share
the same radio spectrum, mutual interference plays a crucial role in determining system
capacity. Good codewords with small cross-correlation help in reducing mutual
interference and power control mitigates near-far effect bringing energy efficiency. Joint
codeword adaptation and power control can help users to achieve their desired Quality of
Service (QoS) as well as in increasing the' system capacity. The QoS can be defined in
terms of target Signal to Interference plus Noise Ratio (SINR), acceptable bit error rate.
In this dissertation work non co-operative game theoretic approach has been
used to deal with joint power control and codeword adaptation in wireless synchronous
CDMA system. In CDMA system, every user tries to achieve its desired QoS with least
possible transmitting power and best codeword sequence. Thus, all the users have
conflicting interests and with their actions being interrelated. Non co-operative game
theory can be used to model this scenario. In this dissertation, Non co-operative Joint
power control and codeword adaptation game (NPCG) is considered where each user
adjusts its codeword and power to minimize its cost while achieving its desired QoS. The
NPCG is formulated as separable game with respect to codeword and power.
Simulations have been carried out for continuous and quantized codeword and
power profiles for single cell and multi-cell systems. Incremental update strategy has
been used for power and codeword update so that receiver can follow changes in
transmitter. The performance for dynamic wireless systems under varying QoS and
varying number of users has been evaluated. The algorithm effectively tracks variable
target SINR and variable number of users. In multi-cell systems, the algorithm reaches
equilibrium with increase in signal dimension. For quantized codewords, different signal
space dimensions with different quantization level have been used. Coarser quantization
has been used for larger signal dimension to compensate for the bandwidth increment. Its
effect on various performance parameters at equilibrium has been studied.
