Text

theses

ETD graduate

Future cellular networks need to reduce their carbon footprint in view of rapidly increasing data traffic from emerging applications including Augmented Reality (AR)/ Virtual Reality (VR), autonomous vehicles and Internet of Things (IoT). A potential solution is the addition of small base stations, overlaid upon the macro base station, installed at locations encountering traffic hotspot or a coverage gap. However, the addition of base stations leads to an increase in the overall static network power consumption. Algorithms for energy efficiency aim to strike a balance between network throughput and power consumption. In this work, we evaluate several existing optimization based user association algorithms focusing on load balancing and power minimization metrics. Verification of the performance for all user association algorithms is carried out using MATLAB based system level simulator for a network topology comprising of one macro and four pico base stations considering static users. Motivated by the gains demonstrated using optimization based schemes, we consider the design of a lower complexity class of heuristic user association algorithms with comparable performance. The proposed E-ASSOC algorithm utilizes load at each base station and the difference amongst the two best Signal to Noise Ratios (SNRs) for each user as its threshold metrics. Simulation results demonstrate a gain of 41.26% in the total system throughput along with reduction of 32.96% in the system power consumption. Eventually indicating an increase of 87.83% in energy efficiency over the baseline max-SNR based association algorithm along with decrease of 7.2% for load balancing algorithm and 10.13% for power minimization algorithm. Fairness of the algorithms examined using Jain's Fairness Index (JFI) also shows marked improvement over the baseline scheme.

ETD_11152

M.S.

Includes bibliographical references

doi:10.7282/t3-fkes-zk18

The author owns the copyright to this work.