University of Oulu

An initial access optimization algorithm for millimetre wave 5G NR networks

Saved in:
Author: Perera, Arakahagodage1
Organizations: 1University of Oulu, Faculty of Information Technology and Electrical Engineering, Communications Engineering
Format: ebook
Version: published version
Access: open
Online Access: PDF Full Text (PDF, 1.3 MB)
Pages: 44
Persistent link:
Language: English
Published: Oulu : A. Perera, 2019
Publish Date: 2019-10-28
Thesis type: Master's thesis (tech)
Tutor: Rajatheva, Rajatheva
Reviewer: Rajatheva, Rajatheva
Kapuruhamy Badalge, Shashika


The fifth generation (5G) of cellular technology is expected to address the ever-increasing traffic requirements of the digital society. Delivering these higher data rates, higher bandwidth is required, thus, moving to the higher frequency millimetre wave (mmWave) spectrum is needed. However, to overcome the high isotropic propagation loss experienced at these frequencies, base station (BS) and the user equipment (UE) need to have highly directional antennas. Therefore, BS and UE are required to find the correct transmission (Tx) and reception (Rx) beam pair that align with each other. Achieving these fine alignment of beams at the initial access phase is quite challenging due to the unavailability of location information about BS and UE.

In mmWave small cells, signals are blocked by obstacles. Hence, signal transmissions may not reach users. Also, some directions may have higher user density while some directions have lower or no user density. Therefore, an intelligent cell search is needed for initial access, which can steer its beams to a known populated area for UEs instead of wasting time and resources emitting towards an obstacle or unpopulated directions.

In this thesis, we provide a dynamic weight-based beam sweeping direction and synchronization signal block (SSB) allocation algorithm to optimize the cell search in the mmWave initial access. The order of beam sweeping directions and the number of SSBs transmitted in each beam sweeping direction depend on previously learned experience. Previous learning is based on the number of detected UEs per SSB for each sweeping direction.

Based on numerical simulations, the proposed algorithm is shown to be capable of detecting more users with a lower misdetection probability. Furthermore, it is possible to achieve the same performance with a smaller number of dynamic resource (i.e., SSB) allocation, compared to constant resource allocation. Therefore, this algorithm has better performance and optimum resource usage.

see all

Copyright information: © Arakahagodage Perera, 2019. This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.