NULL DEPTH TRADE OFF FOR OUTPUT POWER REDUCTION IN A DOWNLINK ADAPTIVE ANTENNA ARRAY By TUAN NGUYEN SUBMITTED IN FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF: MASTER OF ENGINEERING

The zero-forcing algorithm (min-norm algorithm) is one of the proposed methods for downlink beamforming. This algorithm steers the nulls towards interferers and the main beam towards the desired user provided the angles of arrival of these signals are known to the basestation. However, downlink adaptive arrays use additional power when they are required to include null steering in their beam patterns. This excessive transmitted power reduces the effective antenna gain, increases interference in other directions and has implications on the dimensioning of the power amplifiers feeding the antenna elements. In addition, the power distribution among the antenna array elements is no longer equal. The design ratings for the power amplifiers (PAs) on each element can differ by up to 2.9 dB. A distributed amplifier design could solve this problem. This thesis investigates the trade-off of null depth with transmitted power and utility by modifying the zero-forcing algorithm. The performance or utility of the antenna is defined here as the probability that it can accommodate a given angular scenario between the desired user and interfering sources without transmitting excessive power. A –10 dB null increases antenna utility by 6% when steering a single null and 17% when steering 2 nulls if the excessive transmit power is held below 3 dB for a 4-element antenna array. In this work, a modified version of the min-norm algorithm was used to design the antenna weights from angle of arrival information.