Abstract:
Power Optimization of Bandwidth Efficient WCDMA Transceiver
High data rates combined with high mobility represent a challenge when designing tactical
networks. Advanced algorithms are required, resulting in higher complexity, more chip area, and
increased power consumption. However, this contrasts to the limited power supply of mobile
devices using spread spectrum techniques.
The selected thesis topic is about the application of Point to Point WCDMA transceiver deployed
in the Tactical Networks, optimized regarding its power consumption, with the focus on the
dynamic precision scaling. This device is also made bandwidth efficient, as it uses adaptive
modulation scheme and chooses the appropriate modulation scheme, sending more data using the
same power as per channel conditions. So a tradeoff has been studied between modulation
schemes, word lengths and number of users in a WCDMA system in tactical environment.
This thesis was dealt with in three parts. Firstly, it deals with the simplified physical layer of the
WCDMA. This involves the exclusion of a few components and control section to emphasize more
on the baseband portion, and to make it convenient for MATLAB® implementation. Necessary
parts involving spreading, scrambling, digital modulation, and Gaussian channelization are being
incorporated in this section.
Secondly, the behavior of P2P WCDMA based physical layer is evaluated in the tactical
environment. This is intended to be done by analyzing the underline system in multipath fading
channels of tactical terrains. SUI channels were studied and implemented for the system evaluation
in this phase, considering SUI channels as an approximation of tactical environment. To
accommodate the needs of multipath fading effects, Rake receiver has been studied and added in
the WCDMA transceiver module and effects of higher order modulation and delay spread of a
channel have been studied in it. To improve the performance of classical rake in high delay spread;
some amendments had been studied and incorporated apart from complex multiuser detection
techniques (MUD).
By utilizing the idea that reduced wordlength means reduced area which leads to reduced power
consumption; in the last phase the system has been mapped from floating point to fixed point using
MATLAB® Fixed point Toolbox. Optimum wordlength was found using Gradient Search method
using Fixed point toolbox and Transformation toolbox by Han [1]. Simulation results were
compared for floating and fixed points WCDMA systems.