Michele PoleseDepartment of Information Engineering

University of Padova, Italypolesemi@dei.unipd.it

Supervisor: Prof. Michele Zorzi

End-to-End Design and Evaluation of mmWave Cellular Networks

End-

to-e

nd m

mW

aves

Outline

• Introduction

• A case for end-to-end, full-stack evaluations

•Architectures for 5G mmWaves

•End-to-end protocols for mmWaves

•Data-driven 5G network optimization

•Conclusions and research directions

End-

to-e

nd m

mW

aves

2

3GPP NR: novelties B

and

wid

th 𝐵−

max

40

0 M

Hz

per

car

rie

r

Frame – 10 ms Ph

ysical Reso

urce B

lock

N su

bcarrie

rs

Subframe – 1 ms

Examples of slot numbers with different subcarrier spacing

Subcarrier spacing 60 kHz

Slot – 0.25 ms

Subcarrier spacing 120 kHz

Slot – 0.125 ms

Symbol – 8.9 μs

Flexible frame structure

LTE

SA deployment5G Core

NSA deployment4G EPC

PGW/SGW

MME

HSS

Network slicing and NFV

AMF SMF

UPF

PCFAMF SMF

UPF

PCFAMF SMF

UPF

PCF

5G Core Network options

mmWave directional communications

Multi RAT access

NR

NR

End-

to-e

nd m

mW

aves

3

3GPP NR: timeline

Goal: deployment by 2020

Non Stand-alonespecifications

Dec. 2017 June 2018

Stand-alonespecifications

5G phase 1 5G phase 2

March 2020

Release 15 Release 16

End-

to-e

nd m

mW

aves

4

3GPP NR: mmWaves in cellular networksEn

d-to

-end

mm

Wav

es

3GPP NR Release 16 will support frequencies up to 52.6 GHz

Z. Pi and F. Khan, "An introductionto millimeter-wave mobile

broadband systems," in IEEE Communications Magazine, vol.

49, no. 6, pp. 101-107, June 2011.

5G increases the

datarate [Gbps] latency [ms]

5G decreases the

5

3GPP NR: challenges for mmWavesEn

d-to

-end

mm

Wav

es

UAV with mmWave radio

Ground station

blockage

6

End-to-end design and performance: why?

• Sometimes, link-level is enough• Real networks, however, have several

components in-between the user, the link and the content he/she needs

PHYMACRLC

TCP/IPAPP

PDCPRRC

Core network

InternetTCP/IP

APP

End-

to-e

nd m

mW

aves

Focus of my researchend-to-end, system-level design &

evaluation of 5G mmWave networks7

End-

to-e

nd m

mW

aves

The ArchitectureSystem Level Design of5G mmWave Networks

The ProtocolsEnd-to-End and Cross-Layer

Analysis of 5G mmWave Networks

The IntelligenceData-Driven 5G Networks

Optimization

8

The tool: ns-3 mmWave module

Channel model

Application and network stack

3GPP cellular stack 3GPP cellular stack

Application and network stack

PacketPropagation

Fading Beamforming

InterferenceSINR

Error model

End-

to-e

nd m

mW

aves

https://github.com/nyuwireless-unipd/ns3-mmwavehttps://github.com/signetlabdei/ns3-mmwave-iab

https://github.com/signetlabdei/quichttps://github.com/signetlabdei/mmwave-psc-scenarios

9

System Level Design of 5G mmWave NetworksMulti connectivity, beam management and Integrated Access and Backhaul

The

Arch

itec

ture

The ArchitectureSystem Level Design of5G mmWave Networks

The ProtocolsEnd-to-End and Cross-

Layer Analysis of 5G mmWave Networks

The IntelligenceData-Driven 5G Networks

Optimization

System-level challenges at mmWavesTh

e Ar

chit

ectu

re

Issues: high propagation loss and blockage

Large antenna arrays increase the link budget, but the power

is focused on narrow beams

Ultra-dense deployments

Provide backhaul to all the base stations

2

1High number of handovers

3

Need to track the narrow

beams when moving

11

System level solutions at mmWavesTh

e Ar

chit

ectu

re

Integrated Access and Backhaul

Low cost, high density mmWave

deployments

1Multi-connectivity

Low-latency, highly reliable handovers

2

3

Beam managementSeamless tracking

12

• Goal: design a system resilient to fluctuations and outages

• Contribution:

Multi-connectivity architectureto combine sub-6 GHz and mmWave benefits

Mobility managementTh

e Ar

chit

ectu

re

13

1

Results: latency with TCP traffic

• No handover (always keep the same BS)

• Single connectivity (traditional HO architecture)

• Multi connectivity (fast handovers – no service interruption)The

Arch

itec

ture

Proposed solution

14

Integrated Access and Backhaul

• Goal: provide wireless backhaul to ultra-dense mmWavenetworks

• Contributions:

IAB module for ns-3 mmWave

Analysis of IAB end-to-end performance

Distributed path selection policies

The

Arch

itec

ture

3GPP Work Item for Release 16

SDAPPDCPRLCMAC

RLCMAC

Adaptation

RLCMAC

Adaptation

RLCMAC

GTP-UUDP

IP

GTP-UUDP

IP

SDAPPDCP

PHYPHY

Adaptation

RLCMAC

Adaptation

RLCMAC

PHY PHY

F1*Uu

IAB-donor

IAB-nodeIAB-node

UE

DU CU-UP

F1*

MT

Core NetworkInternet

2

End-to-end Performance for IABTh

e Ar

chit

ectu

re

Impact of synchronous vs. bursty traffic

Webpage loading time with browsing model

Throughput with full buffer source

20

• Goal: perform directional initial access and tracking

• Contributions:

Study of 3GPP NR beam management schemes

Analysis of their performance with design insights

Beam management in 3GPP NR

28 GHzomnidirectional range

28 GHzdirectional range

The

Arch

itec

ture

15

3

1. Beam sweeping

2. Beam measurement

3. Beam determination

4. Beam reporting

Beam Management in NRThe 3GPP has specified a set of procedures for the control of

multiple beams at mmWave frequencies which are categorized under the term BEAM MANAGEMENT

The

Arch

itec

ture

Initial Access in a standalone deployment

RACH preamble

gNB UE

SS Burst

UE decides which is the best beam

SS Blocks to get RACH resources UE receives RACH

resource allocation

Beam sweep and measurement

Beam determination

Beam reporting

16

Reactiveness: how much time does it take to perform IA?

Accuracy: what is the probability of receiving an SS block?

Accuracy-reactiveness tradeoff in NRTh

e Ar

chit

ectu

re

Number of antennas at gNB and UE

gNBdensity

Number of SS blocks per burst

17

Beam management for UAVsTh

e Ar

chit

ectu

re

Proposed location-based beam management for UAVsExperimental evaluation

End-to-End and Cross-Layer Analysis of 5G mmWave NetworksTCP issues in mmWave networks

The

Prot

ocol

sThe Architecture

System Level Design of5G mmWave Networks

The ProtocolsEnd-to-End and Cross-

Layer Analysis of 5G mmWave Networks

The IntelligenceData-Driven 5G Networks

Optimization

TCP issues in mmWave networksTh

e Pr

otoc

ols PHY

MAC

RLC

TCP

APP

PDCP

SDAP

End-to-end data plane

IP

mmWave channel: volatile, highly variable capacity, large bandwidth

Link view• Retransmissions• Reordering• Buffering

How do these components interact?

Host-to-host “abstract” view• Congestion and flow control• Retransmissions

22

The

Pro

toco

ls

LOSNLOS

After transition from LOS

time

LOS

time

LOS

NLOS

RLC buffer occupancy

congestion window

Large buffer BufferbloatHigh latency

Small bufferBuffer overflowLow throughput

a) DUPACK retx (CW/2)b) RTO retx (CW=1)

time

a)

b)LOS

NLOS

RLC buffer occupancy

congestion window

Slow ramp-up when back in LOS

1

2

3

23

Possible solutions

• Goal: improve TCP end-to-end performance on mmWaves

• Contributions:Edge deployments: a shorter control loop, to react faster

CC algorithms: faster window ramp-up mechanisms

Exploit multiple paths: mobility management or MP-TCP

milliProxy: cross-layer approach to better control the TCP sending rate

The

Pro

toco

ls

Flow Buffer

Flow window management module

ACK management module

milliProxy instance

serverUE

end-to-end flows

server

UE

flow Buffer

flow window management module

ACK management module

milliProxy instance

24

milliProxy – a TCP proxy for mmWaves

Reduce buffering latency + increase goodput▪ Transparent to the end-to-end flow

▪ Installed in the gNB – or at the edge

▪ Cross-layer approach▪ Per-UE data rate▪ RLC buffer occupancy▪ RTT estimation

▪ Modular▪ Plug-in different

flow controlalgorithms(inspired to [1])

Flow Buffer

Flow window management module

ACK management module

milliProxy instance

serverUE

end-to-end flows

server

UE

flow Buffer

flow window management module

ACK management module

milliProxy instance

The

Prot

ocol

s

[1] M. Casoni et al., “Implementation and validation of TCP options and congestion control algorithms for ns- 3,” in Proc. WNS3, 2015

25

milliProxy – flow control

▪ Interaction with the TCP sender

▪ TCP sending rate is min(CW,ARW)

▪ milliProxy modifies the ARW in the ACKs, according to the flow controlpolicy used

▪ Bandwidth-Delay Product (BDP)based ARW = BW*RTT

▪ More conservative ARW = min([RTT*PHYrate]-B, 0)

Congestion window (computed by the sender)

Advertised window (receiver’s feedback sent on ACK packets)

The

Prot

ocol

s

26

Results: scenario with many LOS/NLOS transitions

Throughput Latency

Latency reduction w milliProxyThroughput gain w milliProxy

The

Prot

ocol

s

27

Data-Driven 5G Networks OptimizationMachine Learning at the Edge

The

Inte

llige

nce

The ArchitectureSystem Level Design of5G mmWave Networks

The ProtocolsEnd-to-End and Cross-

Layer Analysis of 5G mmWave Networks

The IntelligenceData-Driven 5G Networks

Optimization

• Goal: deploy intelligent and data-driven techniques in 5G networks

• Contributions: Mobile-edge controller-based architecture

Data-driven dynamic clustering of base stations

Prediction accuracy of the number of UEs per base station

Machine learning at the edgeTh

e In

telli

genc

e

Data collectionPolicy enforcement

PHY-highMACRLC

PDCPSDAPRRC

RU

CU CU CU

RU RU

RAN Controller

CU CU CU

RAN Controller

RU RU RU

RURU

CU

RU

CU CU CU

RURU

RAN Controller

Cloud Network Controller

3GPP RANlayers

Fast control loop

ClusterRAN

Edge

Cloud

PHY-lowRF

DU DU DU DU DU DU DU DU DU DU

29

Data-driven clustering exampleTh

e In

telli

genc

e

Base station locations

Each color is a cluster

30

Prediction of the number of UEs

• Spatial correlation (cluster- vs local-based) is more impactful than temporal correlation

• Exploit geographic constraints on mobility flows

• When considering all the 472 eNBs (in 22 clusters):

The

Inte

llige

nce

53% RMSE reduction 5% RMSE reduction when increasing 𝑊

0 50 100 150 200 250

80

100

120

140

160

180True

Predicted

(a) High number of users

0 50 100 150 200 250

6

8

10

12

14

16 True

Predicted

(b) Low number of users

Fig. 4: Example of predicted vs true time series, for L = 3 (i.e., 15 minutesahead), W = 3 and the cluster-based GPR on two base stations for cluster 0.

W for the two most accurate methods, the local-based BRR

and the cluster-based GPR. It can be seen that, with respect

to Fig. 3a, where W is fixed, the difference is limited for the

GPR and BRR (i.e., below 5%), while it is more considerable

for the local-based RFR.

Furthermore, the spatial dimension (i.e., the usage of a

joint prediction based on the cluster) is more impactful on

the RMSE than the temporal one (i.e., the past history used

as a feature). Indeed, while the RMSE for the GPR and

BRR improves by up to 5% by varying W , it decreases by

up to 50% when comparing the local- and the cluster-based

methods. The target of the prediction is indeed the number of

users at a cell level (contrary to prior work which focused on

single-user mobility prediction [16]), thus the geography of

the scenario in which the base stations are deployed actually

limits the possible movements of users across neighboring

cells. These constraints on the mobility flow translate into a

spatial correlation among the number of users in neighboring

base stations at time t and at time t + L .

However, there are still some limitations to the accuracy of

the prediction. Fig. 4 reports an example of the true and the

predicted (for L = 3, i.e., 15 minutes) time series for two

1 2 3 4 5 6 7 8 96

8

10

12

14

16

Lag L [5 minutes]

RM

SEσ̂

Cluster-based GPR

Local-based BRR

Fig. 5: Average cluster-based GPR vs local-based BRR for all the SanFrancisco base stations.

Fig. 6: Map of the routes. The dots represent the visited base stations. Noticethat, for route 2 (the red one), several base stations are shared with either theblue or the green routes.

different base stations, with a low and high number of users.

It can be seen that the true time series exhibit daily patterns,

but also a high level of noise. As a consequence, the predicted

values manage to track the main trend of the true time series,

but do not represent the exact value of the number of users in

all cases. This is more noticeable with a low number of UEs,

as in Fig. 4b, which also shows more limited daily variations.

Given the promising results of the cluster-based approach

on the sample cluster, we selected the best performing local-

and cluster-based methods, i.e., respectively, the BRR and the

GPR, and performed the prediction on all the base stations of

the San Francisco area, once again clustered according to the

approach in [1]. The average RMSE over all the base stations

is reported in Fig. 5. The cluster-based method consistently

outperforms the local-based one. The reduction in the average

RMSE over all the clusters Ecl ust er s[σ̂] is 18.3% for L =

Approach based on the proposed architecture

31

Conclusions

Conc

lusi

ons

Conclusions

• System–level, end-to-end approach throughout all the topics

• Considered different components of a complex system and introduce novel contributions for• Architectures• Protocols• Intelligence

• Thorough and realistic performance evaluation

• End-to-end, full-stack analysis may uncover unexpected behaviors

Conc

lusi

ons

33

Future work

• Future research will still be focused on a system-level approach, combined with testbeds and experimental results

• What happens when you consider increasingly complex systems?

Conc

lusi

ons

Intel NUC

GNSS, Compass,

Gyro,MAG,

Acceler.

Facebook Terragraph mmWave Radio

Motor 1Motor 2

Motor 3 Motor 5

6 battery slots

Motor 4 Motor 6

mm

BA

C

Python 3.7 to

DJI Onboard

SDK

Python 3.7

L1: PHY

L0: Motion

L2: MAC

Flight mgmtcommands

mmWave Radio

60 GHz RF

Intel NUC

Beamselection

USB

3

.0Et

h0

+_

+_

+_

+_

+_

+_

Power Module

MOTOR 1MOTOR 2MOTOR 3MOTOR 4 ES

C

MOTOR 6MOTOR 5

Eth0

Beam mgmtcommands

DJI

Flig

ht

Co

ntr

olle

r U

nit

(FC

U)

DC power input

DC-DC step up

Locationreadings

Wi-Fi

GN

SSC

om

p.

Gyro

MA

GA

ccel.I/O

Board

DJI M600 Pro

SNR readings

A3 ProDJI FCJT

AG DJI

API

34

Journals• [1] M. Polese, M. Giordani, M. Mezzavilla, S. Rangan, and M. Zorzi, “Improved Handover Through Dual Connectivity in 5G mmWave Mobile

Networks,” IEEE Journal on Selected Areas in Communications, vol. 35, no. 9, pp. 2069–2084, September 2017.

• [2] M. Polese, R. Jana, and M. Zorzi, “TCP and MP-TCP in 5G mmWave Networks,” IEEE Internet Computing, vol. 21, no. 5, pp. 12–19, September 2017.

• [3] M. Mezzavilla, M. Zhang, M. Polese, R. Ford, S. Dutta, S. Rangan, and M. Zorzi, “End-to-end simulation of 5G mmwave networks,” IEEE Communications Surveys and Tutorials, vol. 20, no. 3, pp. 2237–2263, Third quarter 2018.

• [4] M. Mezzavilla, M. Polese, A. Zanella, A. Dhananjay, S. Rangan, C. Kessler, T. S. Rappaport, and M. Zorzi, “Public Safety Communications above 6 GHz: Challenges and Opportunities,” IEEE Access, vol. 6, pp. 316–329, 2018.

• [5] M. Dalla Cia, F. Mason, D. Peron, F. Chiariotti, M. Polese, T. Mahmoodi, M. Zorzi, and A. Zanella, “Using Smart City Data in 5G Self-Organizing Networks,” IEEE Internet of Things Journal, vol. 5, no. 2, pp. 645–654, April 2018.

• [6] M. Zhang, M. Polese, M. Mezzavilla, J. Zhu, S. Rangan, S. Panwar, and a. M. Zorzi, “Will TCP Work in mmWave 5G Cellular Networks?” IEEE Communications Magazine, vol. 57, no. 1, pp. 65–71, January 2019.

• [7] M. Giordani, M. Polese, A. Roy, D. Castor, and M. Zorzi, “Standalone and Non-Standalone Beam Management for 3GPP NR at mmWaves,” IEEE Communications Magazine, vol. 57, no. 4, pp. 123–129, April 2019.

• [8] M. Giordani, M. Polese, A. Roy, D. Castor, and M. Zorzi, “A Tutorial on Beam Management for 3GPP NR at mmWave Frequencies,” IEEE Communications Surveys and Tutorials, vol. 21, no. 1, pp. 173–196, First quarter 2019.

• [9] M. Polese, F. Chiariotti, E. Bonetto, F. Rigotto, A. Zanella, and M. Zorzi, “A Survey on Recent Advances in Transport Layer Protocols,” IEEE Communications Surveys and Tutorials, pp. 1–1, 2019.

• [10] F. Meneghello, M. Calore, D. Zucchetto, M. Polese, and A. Zanella, “IoT: Internet of Threats? A survey of practical security vulnerabilities in real IoT devices,” IEEE Internet of Things Journal, pp. 1–1, 2019.

• [11] M. Polese, R. Jana, V. Kounev, K. Zhang, S. Deb, and M. Zorzi, “Machine Learning at the Edge: A Data-Driven Architecture with Applications to 5G Cellular Networks,” submitted to IEEE Transactions on Mobile Computing, 2019. [Online]. Available: https://arxiv.org/pdf/1808.07647.pdf

• [12] M. Polese, M. Giordani, T. Zugno, A. Roy, S. Goyal, D. Castor, and M. Zorzi, “Integrated Access and Backhaul in 5G mmWave Networks: Potentials and Challenges,” IEEE Communications Magazine (to appear), 2019. [Online]. Available: https://arxiv.org/pdf/1906.01099.pdf

• [13] M. Giordani, M. Polese, M. Mezzavilla, S. Rangan, and M. Zorzi, “Towards 6G Networks: Use Cases and Technologies,” IEEE Communications Magazine (to appear), March 2019. [Online]. Available: https://arxiv.org/pdf/1903.12216.pdfCo

nclu

sion

s

35

Conferences - 1

• [14] M. Polese, M. Centenaro, A. Zanella, and M. Zorzi, “M2M massive access in LTE: RACH performance evaluation in a Smart City scenario,” in 2016 IEEE International Conference on Communications (ICC), May 2016, pp. 1–6.

• [15] M. Polese, M. Mezzavilla, and M. Zorzi, “Performance Comparison of Dual Connectivity and Hard Handover for LTE-5G Tight Integration,” in Proceedings of the 9th EAI International Conference on Simulation Tools and Techniques, ser. SIMUTOOLS’16, Prague, Czech Republic, 2016, pp. 118–123.

• [16] F. Chiariotti, D. Del Testa, M. Polese, A. Zanella, G. M. Di Nunzio, and M. Zorzi, “Learning methods for long-term channel gain prediction in wireless networks,” in International Conference on Computing, Networking and Communications (ICNC2017), January 2017.

• [17] M. Polese, R. Jana, and M. Zorzi, “TCP in 5G mmWave Networks: Link Level Re- transmissions and MP-TCP,” in 2017 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), May 2017.

• [18] E. Lovisotto, E. Vianello, D. Cazzaro, M. Polese, F. Chiariotti, D. Zucchetto, A. Zanella, and M. Zorzi, “Cell Traffic Prediction Using Joint Spatio-Temporal Information,” in 6th International Conference on Circuits and Systems Technologies (MOCAST), May 2017.

• [19] M. Zhang, M. Polese, M. Mezzavilla, S. Rangan, and M. Zorzi, “ns-3 Implementation of the 3GPP MIMO Channel Model for Frequency Spectrum above 6 GHz,” in Proceedings of the 9th Workshop on ns-3, Porto, Portugal, 2017, pp. 71–78.

• [20] T. Azzino, M. Drago, M. Polese, A. Zanella, and M. Zorzi, “X-TCP: A Cross Layer Approach for TCP Uplink Flows in mmWaveNetworks,” in 16th Annual Mediterranean Ad Hoc Networking Workshop (Med-Hoc-Net’17), June 2017.

• [21] M. Dalla Cia, F. Mason, D. Peron, F. Chiariotti, M. Polese, T. Mahmoodi, M. Zorzi, and A. Zanella, “Mobility-aware Handover Strategies in Smart Cities,” in International Symposium on Wireless Communication Systems (ISWCS), August 2017.

• [22] M. Polese, M. Mezzavilla, S. Rangan, and M. Zorzi, “Mobility Management for TCP in mmWave Networks,” in Proceedings of the 1st ACM Workshop on Millimeter-Wave Networks and Sensing Systems 2017, ser. mmNets ’17. Snowbird, Utah, USA: ACM, 2017, pp. 11–16.

• [23] M. Gentil, A. Galeazzi, F. Chiariotti, M. Polese, A. Zanella, and M. Zorzi, “A deep neural network approach for customized prediction of mobile devices discharging time,” in 2017 IEEE Global Communications Conference (GLOBECOM), Dec 2017, pp. 1–6.

Conc

lusi

ons

36

Conferences - 2

• [24] M. Polese, M. Mezzavilla, M. Zhang, J. Zhu, S. Rangan, S. Panwar, and M. Zorzi, “milliProxy: A TCP proxy architecture for 5G mmWave cellular systems,” in 2017 51st Asilomar Conference on Signals, Systems, and Computers, Oct 2017, pp. 951–957.

• [25] M. Polese, M. Mezzavilla, S. Rangan, C. Kessler, and M. Zorzi, “mmwave for future public safety communications,” in Proceedings of the First CoNEXT Workshop on ICT Tools for Emergency Networks and DisastEr Relief, ser. I-TENDER ’17. Incheon, Republic of Korea: ACM, 2017, pp. 44–49. [Online]. Available: http://doi.acm.org/10.1145/3152896.3152905

• [26] M. Drago, T. Azzino, M. Polese, C. Stefanovic, and M. Zorzi, “Reliable Video Streaming over mmWave with Multi Connectivity and Network Coding,” in International Conference on Computing, Networking and Communications (ICNC), March 2018, pp. 508– 512.

• [27] T. Zugno, M. Polese, and M. Zorzi, “Integration of Carrier Aggregation and Dual Connectivity for the ns-3 mmWave Module,” in Proceedings of the 10th Workshop on ns-3, ser. WNS3 ’18, Surathkal, India, 2018, pp. 45–52.

• [28] M. Polese and M. Zorzi, “Impact of Channel Models on the End-to-End Performance of mmWave Cellular Networks,” in Proceedings of the 19th International Workshop on Signal Processing Advances in Wireless Communications (SPAWC), June 2018.

• [29] M. Giordani, M. Polese, A. Roy, D. Castor, and M. Zorzi, “Initial access frameworks for 3GPP NR at mmWave frequencies,” in 2018 17th Annual Mediterranean Ad Hoc Networking Workshop (Med-Hoc-Net), June 2018, pp. 1–8.

• [30] M. Polese, M. Giordani, A. Roy, S. Goyal, D. Castor, and M. Zorzi, “End-to-End Simulation of Integrated Access and Backhaul at mmWaves,” in IEEE 23rd International Workshop on Computer Aided Modeling and Design of Communication Links and Net-works (CAMAD), September 2018.

• [31] M. Polese, M. Giordani, A. Roy, D. Castor, and M. Zorzi, “Distributed Path Selection Strategies for Integrated Access and Backhaul at mmWaves,” in IEEE Global Communications Conference (GLOBECOM), Dec 2018.

• [32] M. Rebato, M. Polese, and M. Zorzi, “Multi-Sector and Multi-Panel Performance in 5G mmWave Cellular Networks,” in IEEE Global Communications Conference (GLOBECOM), Dec 2018.

• [33] M. Polese, R. Jana, V. Kounev, K. Zhang, S. Deb, and M. Zorzi, “Exploiting spatial correlation for improved user prediction in 5G cellular networks,” in Proceedings of the Information Theory and Applications Workshop, ser. ITA ’19, San Diego, 2019.

Conc

lusi

ons

37

Conferences - 3

• [34] W. Xia, M. Polese, M. Mezzavilla, G. Loianno, S. Rangan, and M. Zorzi, “Millimeter Wave Remote UAV Control and Communications for Public Safety Scenarios,” in Proceedings of the 1st International Workshop on Internet of Autonomous Unmanned Vehicles, ser. IAUV ’19, Boston, MA, 2019.

• [35] M. Polese, T. Zugno, and M. Zorzi, “Implementation of Reference Public Safety Scenarios in ns-3,” in Proceedings of the 11th Workshop on ns-3, ser. WNS3 ’19, Florence, Italy, 2019.

• [36] A. De Biasio, F. Chiariotti, M. Polese, A. Zanella, and M. Zorzi, “A QUIC Implementation for ns-3,” in Proceedings of the 11th Workshop on ns-3, ser. WNS3 ’19, Florence, Italy, 2019.

• [37] T. Zugno, M. Polese, M. Lecci, and M. Zorzi, “Simulation of Next-Generation Cellular Networks with ns-3: Open Challenges and New Directions,” in Proceedings of the Work- shop on Next-Generation Wireless with ns-3, ser. WNGW ’19, Florence, Italy, 2019.

• [38] M. Polese, F. Restuccia, A. Gosain, J. Jornet, S. Bhardwaj, V. Ariyarathna, S. Man- dal, K. Zheng, A. Dhananjay, M. Mezzavilla, J. Buckwalter, M. Rodwell, X. Wang, M. Zorzi, A. Madanayake, and T. Melodia, “MillimeTera: Toward A Large-Scale Open- Source mmWave and Terahertz Experimental Testbed,” in Proceedings of the 3rd ACM Workshop on Millimeter-Wave Networks and Sensing Systems, ser. mmNets ’19. Los Cabos, Mexico: ACM, 2019.

• [39] L. Bertizzolo, M. Polese, L. Bonati, A. Gosain, M. Zorzi, and T. Melodia, “mmBAC: Location-aided mmWaveBackhaul Management for UAV-based Aerial Cells,” in Pro- ceedings of the 3rd ACM Workshop on Millimeter-Wave Networks and Sensing Systems, ser. mmNets ’19. Los Cabos, Mexico: ACM, 2019.

• [40] M. Drago, M. Polese, S. Kucera, D. Kozlov, V. Kirillov, and M. Zorzi, “QoS Provisioning in 60 GHz Communications by Physical and Transport Layer Coordination,” IEEE 16th International Conference on Mobile Ad Hoc and Sensor Systems (MASS), Nov 2019.

Conc

lusi

ons

38

Book Chapter

• [41] M. Polese, M. Giordani, and M. Zorzi, “3GPP NR: the standard for 5G cellular networks,” in 5G Italy White eBook: from Research to Market, 2018.

Conc

lusi

ons

39

Michele Polese

Department of Information Engineering

University of Padova, Italy

polesemi@dei.unipd.it

End-to-End Design and Evaluation of mmWave Cellular Networks

polese.iommwave.dei.unipd.it

End-

to-e

nd m

mW

aves