University of Waterloo
200 University Ave W, Waterloo, ON
N2L 3G1
Phone: (519) 888-4567
Staff and Faculty Directory
Contact the Department of Electrical and Computer Engineering
Thursday, August 12, 2021 — 1:00 PM EDT
Candidate: Andrew Lappalainen
Title: Dimensioning 5G Networks for Fixed Wireless Access Services at Rural Homes
Date: August 12, 2021
Time: 1:00 pm
Place: MS Teams
Supervisor(s): Rosenberg, Catherine
Abstract:
We study the problem of dimensioning a rural 5G multi-user massive MIMO (MU-MIMO) cell using a 3.5 GHz time-division duplex (TDD) band to offer minimum bit rates (MBRs) jointly in the downlink (DL) and the uplink (UL) to multi-antenna fixed wireless access
(FWA) homes. We seek to answer the following research questions: RQ1) Given a set of network resources, what is the user limit, i.e., the maximum number of FWA homes that can simultaneously receive a target DL MBR and target UL MBR (or better)? RQ2: How should a network be configured to obtain such user limits? What bandwidth, transmit power, and number of antennas are necessary?
How should time resources be shared between the DL and UL and how should users and streams be selected?
To answer these questions, we propose a static user grouping strategy that organizes homes into fixed groups and allocates an equal share of resources to each group. Under the static grouping strategy, we develop a simple process to determine the user limit by decoupling the DL and UL problems and determining the minimum number of resources required to give a group of users the MBR. We first study how DL and UL group sizes impact performance and show that smaller group sizes enable more streams to be used and a larger total number of users to be given the MBRs. Using the simplified dimensioning process we developed, we then show how the user limit at different cell radii is impacted by the following: the system bandwidth, the number of antennas at the base station and homes, the base station transmit power, and the DL and UL MBRs. Finally, we show how these results can also provide insight into how the network should be operated when the number of active homes is less than the user limit.