SOURCE:

Faculty: Engineering
Department: Electronic And Computer Engineering

CONTRIBUTORS:

Nosiri, O.C;
Onoh, G.N;
Idigo,V.E;

ABSTRACT:

This thesis identified the major mechanism of sideband noise in a co-located setting involving CDMA 2000 and WCDMA networks at 1.9GHz. A proposed mitigation technique comprised of two different design methods was developed viz: the applications of Butterworth Band pass Filtering Technique (BBFT) and Adaptive Noise Cancellation Technique (ANCT). The Butterworth Band pass Filter (BBPF) design was achieved through the empirical analysis conducted in two different scenarios to obtain the following: (1) the received signal strength degradation, (2) the interference power, (3) the percentage carrier-to-interference ratio degradation and (4) the increase in total noise floor level on the WCDMA receiver from the CDMA 2000 transmitting power. The first scenario involved a standalone unco-located WCDMA network while the second scenario involved a co-located network. A magnitude specification at 52dB attenuation was actualized using the Infinite Impulse Response BBPF. The filter technique was considered more suitable when compared to the counterpart Finite Impulse Response filters due to its attributes of lower filter order specifications and better frequency response. The simulated results obtained using Noise Cancellation Performance (NCP) criteria at 52dB, gave rise to amplitude imbalance of 0.03dB, phase error of 0.140 and time delay mismatch of 0.07ns. The ANCT technique on the other hand operates on the principle of destructive interference between a primary path and a reference path correlated noise signals. The technique was developed as an improvement to the limitations of the BBFT. The shortcomings of the fixed filter include: (a) the inability to adaptively track signals under non-stationary conditions, (b) inefficient power handling capability and (c) Inadequate perfect isolation. The experimental set-up of the ANCT was realized using a developed algorithm implemented in a MATLAB-Simulink Test-Bed. The result obtained demonstrated a theoretical perfect error cancellation at 67dB with amplitude imbalance of 0dB, phase error of 00 and time delay mismatch of 0ns. The net effect is about 28% improvement when compared to the result obtained with the BBFT. The ANCT showed significant improvement compared with the reviewed works in the literature due to the system characteristic potentials to drive the weight variables to perfect cancellation.