PETAR POPOVSKI AALBORG UNIVERSITY - CHALLENGES AND POSSIBILITIES FOR WIRELESS IOT - WILAB

challenges and possibilities for wireless IoT
 connectivity on the path beyond 5G

 Petar Popovski
 Aalborg University
 Denmark
 petarp@es.aau.dk

 WiLab - Huawei JIC Workshop, January 11, 2021

outline

connectivity types

 IoT and time

 IoT and space

 transforming the IoT traffic

 WiLab - Huawei JIC Workshop, January 11, 2021 2

outline

connectivity types

 IoT and time

 IoT and space

 transforming the IoT traffic

 WiLab - Huawei JIC Workshop, January 11, 2021 3

the complex connectivity ecosystem

 Video
 server

 internet

 Logistics
 data

 WiLab - Huawei JIC Workshop, January 11, 2021 4

how 5G deals with this complexity

 enhanced Mobile Broadband

 Ultra-Reliable
 Low Latency
Communication

 massive
 Machine Type Communication

perhaps the main innovation in 5G: offer a platform for
flexible support of heterogeneous services
 WiLab - Huawei JIC Workshop, January 11, 2021 5

the space of connectivity services

URLLC mMTC

 IA: Industrial Automation
 V2V-3
 IA-1

 VR: Virtual Reality

 VR-2 VR-1
 V2V: Vehicle-to-Vehicle

 eMBB

 WiLab - Huawei JIC Workshop, January 11, 2021 6

decoupling reliability from low latency

resilient connections with large latency budget

 disaster and rescue
mobile health,
remote monitoring
 smart grid
 this decoupling is visible in the
 new 5G requirements posed by 3GPP (Sep 2020)

 WiLab - Huawei JIC Workshop, January 11, 2021 7

connectivity space revisited

reliability data rate

 tactile robotics

 health monitors
 augmented
 reality

 latency massiveness

 WiLab - Huawei JIC Workshop, January 11, 2021 8

the traffic model challenge

need for improved understanding of the connection ecosystem
 population of
 connections

 generic reliability latency
 connectivity
 data rate massiveness
 types

characterization of the composite connections
§ example: low latency traffic arrivals from the same device
 correlated with broadband requests

 time

 WiLab - Huawei JIC Workshop, January 11, 2021 9

simultaneous massive and ultra-reliable?

 mMTC URLLC

 possible when the information across
 devices is correlated (e.g. alarm)

 WiLab - Huawei JIC Workshop, January 11, 2021 10

illustrative scenario that cobines mMTC and URLLC

 ps
 pd

 pd ps

 Physical pd
 ps
 Phenomenon pd

 ps
 pd

 ps

g. 1. System model with common alarm and standard messages. pd denotes the probability of detecting an alarm, and ps is the probability of send
andard message.

 a device generates two message types
recludes the identification of users and the error measure is done on a per-device basis. This has also been called unsou
 § individual update, independent of others
andom access [11].
 § we
 In this alarm-type message,
 build upon the model correlated
 in [1] with an forweallbring
 important extension: sensors
 in the correlation of activation and mes
ontent across different devices. This is different from the mainstream view on massive random access, where the de
ctivation
 K.and message
 Stern, A. content is independent
 E. Kalør, B. Soret,across
 andtheP.devices. An exemplary
 Popovski, case is
 "Massive as follows:Access
 Random IoT devices can s
andardwith Common
 messages or alarm Alarm Messages",
 messages, the latter with in Proc.
 critical IEEErequirement
 reliability ISIT, Paris, France,byJuly
 and triggered 2019. obse
 a commonly
henomenon. In normal operation, standard uncorrelated messages are sent. Upon the alarm activation, a number of
evices will prioritize it and send WiLab - Huawei
 the same JICThis
 message. Workshop,
 reflectsJanuary 11, 2021
 the extreme all-or-nothing correlation where11devices

heterogeneity brings a new dimension to spectrum usage

the first topic listed in EU Commission call for projects
5G PPP – Smart Connectivity beyond 5G
“Provision of seemingly infinite network capacity
including innovative spectrum use and management …”

 Basil

but how do we measure
the spectrum capacity
when one user wants a high rate
 Zoya Yoshi
and another one wants a fixed rate,
 broadband low-latency
low latency and high reliability? connectivity critical
 connectivity

 WiLab - Huawei JIC Workshop, January 11, 2021 12

heterogeneity brings a new dimension to spectrum usage

 eMBB data rate

 mMTC arrivals

 slicing in spectrum domain
 • orthogonal
 • non-orthogonal

 WiLab - Huawei JIC Workshop, January 11, 2021 13

outline

connectivity types

 IoT and time

 IoT and space

 transforming the IoT traffic

 WiLab - Huawei JIC Workshop, January 11, 2021 14

latency-reliability characterization
 fixed data size

 reliability=Pr( ≤ )

 1
 1-Pe

 latency t
 WiLab - Huawei JIC Workshop, January 11, 2021

design targets

 broadband rate-oriented ultra-reliable low latency
 systems communication URLLC
reliability reliability

1 1

 latency latency
 WiLab - Huawei JIC Workshop, January 11, 2021

timing requirements
 Fig. 4: Relationship beyond
 between application device and latency
 communication device (3GPP TS 22.104 [2]).

 AF = Application function (source and target) HCL = Higher communication layer
 Source: 3GPP TS 22.104 [2] OS = Operating system (source and target) LCL = Lower communication layer
Relationship between application device and communication device [3]
 In control applications, transmissions of data and control commands often have to be com-
[3] 3GPP Technical
 pleted Specification
 within a given 22.104. “Service
 time period. requirements
 This time for cyber-physical
 period determines control
 the maximum applications in
 permissible
vertical domains” https://www.3gpp.org/ftp/Specs/archive/22_series/22.104/
 end-to-end latency.
 WiLab - Huawei JIC Workshop, January 11, 2021
 5.4 Diagnostics

latency vs. age
§ latency performance historically characterized with packet delays
§ other timing requirements
 – tracking applications and sense-compute-actuate cycles
 are not sensitive to packet delay,
 but to the freshness of the information at the receiver

 example:
 satellite-based tracking

 WiLab - Huawei JIC Workshop, January 11, 2021 18

Age of Information
 § Age of Information (AoI) and its byproducts
 are better metrics to capture the freshness of the information
 § exogeneous vs. controlled sampling (generate-at-will)
 AoI

 ...

 t1 t2 t3 t4 time

 tִ 1 tִ 2 tִ 3 tִ 4

 inter-arrival time system t ime
[1] S. Kaul, R. Yates, and M. Gruteser, “Real-time status: How often should one update?” in International Conference on
Computer Communications. IEEE, Mar. 2012, pp. 2731–2735.
[2] A. Kosta, N. Pappas, and V. Angelakis. "Age of information: A new concept, metric, and tool." Foundations and Trends in
Networking 12.3 (2017): 162-259.

 WiLab - Huawei JIC Workshop, January 11, 2021 19

Peak Age of Information (PAoI)
§ Peak Age of Information (AoI) measures how old
 the last packet is when the new one arrives at the destination

 AoI PAoI1 PAoI2

 ...

 t1 t2 t3 t4 time

 tִ 1 tִ 2 tִ 3 tִ 4

 inter-arrival time system t ime

M. Costa, M. Codreanu, and A. Ephremides, “Age of information with packet management,” in
2014 IEEE International Symposium on Information Theory, pp. 1583–1587, June 2014.

 WiLab - Huawei JIC Workshop, January 11, 2021 20

case study: PAoI in edge computing 3

 model with a tandem queue

 µ1 µ2
 4

Fig. 1: An example of the IoT edge computing use case: the sensor transmits data to a computing-enabled edge
node, which needs to maintain information freshness.
 AoI
 PAoI 1 PAoI 3
representing the model in the edge computing scenario: bothPAoI
 the2 transmission and the
 PAoI edge
 4
 PAoI n
computation can be buffered, and the system is a tandem queue.
 If the load on the computing-enabled edge node is time constant, while communication is
 Aoi

less predictable due to, e.g., dynamic channel variations and random access, then the M/M/1 end node
– M/D/1 tandem queue is an appropriate model. Computation time is often modeled as a
linear function of the data size in the literature [7], and updates of the same size would have a intermediate
constant and deterministic service time. Modeling the channel as an M/M/1 queue is correct
if we assume an ALOHA system [8] with perfect Multi-Packet Reception (MPR), in which // gn 1[9]gn
 g1 g2 g3 g4
the packets are not lost due to collisions and can only be lost due to channel errors. This model time
 F. Chiariotti, O. Vikhrova, B. Soret, and P. Popovski. t "Peak Age of Information Distribution in
is suitable for IoT systems based on Ultra-Narrowband (UNB) transmissions, such as SigFox
 Tandem Queue Systems."
 Fig. 2: arXiv
 Change the y-axis frompreprint arXiv:2004.05088
 Aoi to AoIThe time-evolution of the Age (2020).
[10]. Another option is to consider the computing load as also variable over time, leading to
 of Information in a 2-node tandem queue.
 The age at the destination is plotted in black, the age at the intermediate node is plotted in red. Departures from
stochastic computing time, and represent the two systems by M/M/1 queues with different
 each node are marked by a circle.
 WiLab - Huawei JIC Workshop, January 11, 2021 21
service rates [11].

study of the distribution of PAoI
derived the PAoI pdf for
§ M/M/1 and M/D/1 tandem queue (constant computation time)
§ M/M/1 and M/M/1 tandem queue (two independent satellite links) 9

 µ1 µ2
 Transmission buffer Communication link Edge node queue Computing

 gi 1

 ⌦i 1,1
 Yi,1

 gi Si 1,1
 ⌦i,1
 Time ⌦i i
 1,2

 Si,1 Si 1,1
 ri
 ⌦i,2
 Si,1
 ri 1

 Fig. 3: Schematic of the four steps a packet goes through in a tandem queue, highlighting the components of the
F.PAoI.
 Chiariotti, O. Vikhrova, B. Soret, and P. Popovski. "Peak Age of Information Distribution in
Tandem Queue Systems." arXiv preprint arXiv:2004.05088 (2020).

 The reason we named ⌦WiLab
 i,j the- extended waiting
 Huawei JIC time is
 Workshop, that W11,
 January i,j = [⌦i,j ]+ , where [x]+ is equal
 2021 22

Age of Incorrect Information (AAoI)

 92 5.3. System Overview
 Chapter 5. Age of Incorrect Information: Analysis and Optimization

 (a) Age penalty function. (b) Error penalty function. Figure 5.3: Illustration of the proposed penalty function.

 age penalty
 Figure functionof the age and error
 5.2: Illustrations errorpenalty
 penalty function
 functions. combined penalty function
 5.3 System Overview
 5.3.1 System Model
 error penalty function is its failure to capture the following phenomena that arises
 in numerous applications: staying in an erroneous state should have an Inincreasing
 this chapter, we consider a transmitter-receiver pair where the transmitter s
 status updates about the process of interest to the receiver side over an unrel
 penalty effect. In fact, the function in (5.2) treats all instances of error equally, no
 channel. Time is considered to be slotted and normalized to the slot duration
 matter how long the time elapsed since their start is. In other words, the penalty
 the slot duration is taken as 1). The information process of interest is an N s
A.of being in an erroneous state after 1 time slot or 100 time slots is the same
 Maatouk, S. Kriouile, M. Assaad, and A. Ephremides, discrete
 1. Because of this observation, we can see that the long-time average error
 “The value
 age of
 Markov
 penalty
 incorrect
 chain information:
 X(t) t2N depicted A that
 in Fig. 5.4. To new extent, we defin
 probability of remaining at the same state in the next time echelon as Pr(X(t +
performance metric
 due to a burst error is thefor
 samestatus updates,”
 as the one IEEE/ACM
 resulting from Transactions
 several isolated errors
 X(t)) = pRof on Networking,
 . Similarly, the probability ofpp. 1–14,to2020.
 transitioning another state is define
 Pr(X(t
 the same duration. However, this is not always the case. There is a vast amount of + 1) 6
 = X(t)) = p t . Since the process in question can have one of N diffe
 possible
 applications where the penalty grows the longer the monitor has incorrect informa- values, the following always holds:
 tion. For example, let us suppose that X(t) = 1 refers to the case where a machine pR + (N 1)pt = 1. (5
 is at a normal temperature at time
 WiLabt and- X(t) = 2 isJIC
 Huawei the case where the January
 Workshop, machine is11, 2021 23
 overheating. This information has to be transferred to a monitor that can,Asconse- for the unreliable channel model, we suppose that the channel realiza

+
 Department of Computer Science, Aalborg University
 Query Process
 AoI in pull-based communication
 Selma Lagerløfs Vej 300, 9220 Aalborg, Denmark, email: {skj,tbp}@cs.aau.dk

 tti ∗ , Beatriz
 stract—Age Soret∗ , Søren
 of Information (AoI) has Jensenan+,important
 K. become Q1 Q2 Q3
 common
 §freshness
 pt in communications,
 etarthePopovski
 ure
 ∗ ascommunication models are push-based
 it allows system designers
 of the information available to remote
 to

 , Aalborg
 toring
med
 §
 or pull-based
 University
 control communication:
 processes. However, its definition tacitly
 that new information is used at any time, which is not
 1 2 3 4 5 6 7

 il: {jho,aek,fchi,bsa,petarp}@es.aau.dk
 ys the casereading process
 and the instants at whichthat is different
 information is collectedfrom permanent subscription
 (a) Permanent query transmissions
 , Aalborg
 used University
 are dependent on a certain query process. We propose Q1 Q2 Q3
 del that
mark, – satellites, cloud-based
 {skj,tbp}@cs.aau.dk
 accounts
 email: queries
 for the discrete time nature of many to the edge devices, data fetching in a control loop
 toring processes, considering a pull-based communication 1 2 3 4 5 6
 l in which the freshness of information is only important
 the receiver generates a query. We then define the Age (b) Query-aware transmissions
 formation at Q1Query (QAoI), a Q2
 more general metricQ3 that Permanent query
 he pull-based scenario, and show how its optimization can
 Query-aware
 to very different choices from traditional push-based AoI
mization when
 1 using
 2 a Packet
 3 Erasure
 4 Channel
 5 6(PEC). 7
 dex Terms—Age of Information, networked control systems

 AoI
 (a) Permanent query transmissions
 Q1 Q2 Q3
 I. I NTRODUCTION
 ver the past
 1 2 few years, the concept
 3 4 of information5 6fresh-
 has received a significant attention in relation to cyber-
 ical systems that (b) rely
 Query-aware transmissions
 on communication of various up-time Q1 Q2 Q3 time

 in real time. This has led to the introduction
 Permanent query of Age of t
 mation (AoI) [1] as a measure that reflects the freshness
 Query-aware (c) Age for the two systems
 e receiver
 J. Holm,with
 A. E. respect to the sender,
 Kalør, F. Chiariotti, andS.denotes
 B. Soret, theT. B. Pedersen, and P. Popovski. "Freshness on Demand: Optimizing
 K. Jensen,
 renceAge between the current time andProcess."
 the timearXiv
 when Fig. 1: Example of the difference between a system assuming
 the arXiv:2011.00917
 of Information for the Query preprint a permanent query (2020).
 and one that is aware of the query arrival
 AoI

 recently received update was generated at the sender.
 e first works to deal with AoI considered simple queuing process. The same packets are lost (depicted in red) in both
 ms, deriving analytical formulas WiLab - Huawei fresh-
 for information JIC Workshop,
 systems,January
 and the 11, 2021 indicate the age at the query24arrival
 markers
 instants.

outline

connectivity types

 IoT and time

 IoT and space

 transforming the IoT traffic

 WiLab - Huawei JIC Workshop, January 11, 2021 25

systems with massive number of antennas
 1

§ massive MIMO Massive MIMO for Internet of Things (IoT) Connectivity
§ Large Intelligent Surfaces Alexandru-Sabin Bana1 , Elisabeth de Carvalho1 , Beatriz Soret1 ,
 Taufik Abrão2 , José Carlos Marinello2 , Erik G. Larsson3 , and Petar Popovski1
 1
 Department of Electronic Systems, Aalborg University, Denmark
§ Reconfigurable Intelligent Surfaces
 New generation of access points:
 3
 • Low cost, low weight, easy to deploy electromagnetic panels
 2
 Electrical Engineering Department, Londrina State University, Parana, Brazil
 Department of Electrical Engineering (ISY), Linköping University, 581 83 Linköping, Sweden
 • Easily integrable in the surroundings
 • Access points closer to the users
 Abstract—Massive MIMO is considered to be one of the key tech-
 nologies in the emerging 5G systems, but also a concept applicable to
 other wireless systems. Exploiting the large number of degrees of freedom
 905.06205v1 [cs.IT] 15 May 2019

 (DoFs) of massive MIMO essential for achieving high spectral efficiency,
 high data rates and extreme spatial multiplexing of densely distributed
 users. On the one hand, the benefits of applying massive MIMO for short packets
 broadband communication are well known and there has been a large
 body of research on designing communication schemes to support high
 rates. On the other hand, using massive MIMO for Internet-of-Things low latency
 (IoT) is still a developing topic, as IoT connectivity has requirements eMBB
 VR
 URLLC
 and constraints that are significantly different from the broadband high data rate
 connections. In this paper we investigate the applicability of massive high reliability
 MIMO to IoT connectivity. Specifically, we treat the two generic types of
 IoT connections envisioned in 5G: massive machine-type communication
 (mMTC) and ultra-reliable low-latency communication (URLLC). This
 paper fills this important gap by identifying the opportunities and mMTC
 challenges in exploiting massive MIMO for IoT connectivity. We provide
 insights into the trade-offs that emerge when massive MIMO is applied
 to mMTC or URLLC and present a number of suitable communication numerous
 schemes. The discussion continues to the questions of network slicing of devices short packets
 the wireless resources and the use of massive MIMO to simultaneously
 support IoT connections with very heterogeneous requirements. The main sporadic traffic
 conclusion is that massive MIMO can bring benefits to the scenarios with
 IoT connectivity, but it requires tight integration of the physical-layer Figure 1. Massive MIMO and the 5G services.
 techniques with the protocol design.

 Index Terms—mMTC; URLLC; Massive MIMO; 5G; Activity de- The other two services, URLLC and mMTC, are the cornerstones
 tection; Collision resolution; Extended coverage; short packets; NR;
 Random Access (RA); Grant-based RA; Grant-free RA; Unsourced of machine-type traffic and thus enablers of various types of Internet
 RA; Network Slicing; compressing sensing; sparsification; covariance of Things (IoT) connectivity. URLLC, also known as mission-critical
 methods; Cross-layer optimization design IoT, envisions transmission of moderately small data packets (in
 WiLab - Huawei JIC Workshop, the order
 January 11, of2021
 tens of bytes) with extremely high-reliability, ranging
 26
 between 99.999% and 99.9999999%, i.e. down to 10 9 packet
 I. I NTRODUCTION error probability [1]. The user plane latency requirement is most

massive MIMO and IoT

pros
§ very high SNR links
§ quasi-deterministic links, fading immunity
§ extreme spatial multiplexing capability

cons
§ expensive CSI acquisition procedure
§ additional protocol steps

 WiLab - Huawei JIC Workshop, January 11, 2021 27

massive MIMO and IoT 9

 M ULTI -A NTENNA S YSTEMS
VE
 factory scenario with
 predefined
 Multi-Antenna spatial
 Systems for URLLC
he base station orchannels
 terminals of a
 fficient mechanisms at the physical
 low latency communications. They
ent to the higher layer methods de-
section focuses on massive antenna
a very large number of antennas at
 terminals, at high frequency bands,
d as a major enabler towards the
works [37]. They are largely viewed
g the data rates and/or increasing
 users that can be simultaneously
me bandwidth. However, they are
building the two other 5G services,
 Fig. 6. Factory scenario where a massive MIMO access points serves multiple
 communications [38] and URLLC terminals (workstations).
 A.-S. Bana, E. de Carvalho, B. Soret, T. Abrao, J. C. Marinello, E. G. Larsson, P. Popovski,
 “Massive
 antenna MIMO lie
 systems for in
 Internet of Things (IoT) Connectivity”, https://arxiv.org/abs/1905.06205
 their ability
mber of spatial Degrees-of-Freedom B. Channel Structure
 he following remarkable properties To illustrate the main concepts of this section, we assume a
 WiLab - Huawei JIC Workshop, January 11, 2021 28
LLC: factory-type environment as pictured in Fig. 6. An access point

Collection of closely spaced tiny antenna elements over a large surface
simple IoT devices and Reconfigurable Intelligent Surfaces
Provides a high resolution image of the propagation environment

 RIS

Base Station 14

 simple IoT devices can
 benefit significantly from the
 favorable and reconfigurable
 wireless environment
 WiLab - Huawei JIC Workshop, January 11, 2021 29

Sensing
case study: RIS helps satellite IoT
 Collection of closely spaced tiny antenna elements over a large surface
 Provides a high resolution image of the propagation environment
 5

 Isotropic IRS Planar IRS, 45° tilt Diffuse Reflector 10° 25° 50° 90° 50° 25° 10°
 Without IRS Planar IRS, 0° tilt Specular Reflector 8

 IRS Gain [dB]
 10° 25° 50° 90° 50° 25° 10° 6

 Channel Gain [dB]
 155 4
 160 2
 0
 165
 400 200 0 200 400
 170 Time [s] & Elevation Angle (top)
 400 200 0 200 400 Fig. 4. Gain of using IRS over LOS only communication (legend as in Fig. 3).
 Time [s] & Elevation Angle (top)
 shown that the phase shifts can be chosen such that they
 Fig. 3. Channel gain over time for one satellite pass. The plots of the scenarios 14 incurring any Doppler
 without IRS and with reflectors are congruent. maximize the received power without
 spread. Moreover, they can be kept within the interval [0, 2⇡]
 and as a diffuse reflector with phase shifts chosen by Snell’s which is in the feasible range of recent IRS prototypes [7]. It
 § predictive satellite mobility used to adjust the reflective surface
 law and uniformly in the interval [0, 2⇡], respectively. These
 configurations emulate the behavior of a planar obstacle that
 also results in the minimum delay spread under meaningful IRS
 operation. In a numerical study, we demonstrate the benefits
 reflects the signal in place of the IRS. Further, to obtain an of IRS-assisted LEO satellite communication and show that
 § multi-objective optimization:
 upper bound that serves as a best case deployment, we consider the SNR is increased by 3 dB to 6 dB for an IRS the size of
 an IRS with isotropic elements, i.e., G(✓, ') = 1 for ✓ 2 [0, ⇡2 ] two billboards. This requires rotating the IRS such that it has
 received power, Doppler spread and delay spread
 and zero otherwise. The number of elements in this case is a favorable orientation to the transmitter and receiver. Finding
 chosen such that the effective antenna area matches the size of this optimal orientation is left open for future work. Other
 2
 the IRS. With an effective area per element of 4⇡ this amounts open topics are the inclusion of statistical channel model to
 to 433 ⇥ 288 elements. account for atmospheric effects and LOS outages.
B. Matthiesen, E. Björnson, E. De Carvalho and P. Popovski, "Intelligent Reflecting Surface
 It can be observed from Fig. 3 that the gain of the IRS
 R EFERENCES
Operation under Predictable Receiver Mobility: A Continuous Time Propagation Model," in IEEE
 with isotropic elements over the baseline is 7.9 dB. This is
 also directly displayed in Fig. 4. In contrast, the gain of the [1] C. Liaskos et al., “A new wireless communication paradigm through
Wireless Communications Letters, 2020.
 IRS with planar elements is negligible. This is due to the software-controlled metasurfaces,” IEEE Commun. Mag., vol. 56, no. 9,
 pp. 162–169, Sep. 2018.
 unfavorable angle of the receiver towards the IRS which reduces [2] M. Di Renzo et al., “Smart radio environments empowered by recon-
 the effective area of the IRS to nearly zero. This issue can figurable AI meta-surfaces: an idea whose time has come,” EURASIP J.
 WiLab
 be avoided- by
 Huawei
 physicallyJIC Workshop,
 rotating the IRS towardsJanuary
 the sky. In11, 2021
 Wireless Commun. Netw., vol. 2019, no. 1, May 2019. 30
 [3] E. Basar et al., “Wireless communications through reconfigurable
 particular, by rotating the x-axis by 45° and keeping the IRS intelligent surfaces,” IEEE Access, vol. 7, pp. 116 753–116 773, 2019.

massive satellite IoT aided by RIS

 Sensing

 Collection of closely spaced tiny antenna elements over a larg
sely spaced tiny antenna elements over a large surface
 Provides a high resolution image of the propagation enviro
 resolution image of the propagation environment

 14

 WiLab - Huawei JIC Workshop, January 11, 2021 31

outline

connectivity types

 IoT and time

 IoT and space

 transforming the IoT traffic

 WiLab - Huawei JIC Workshop, January 11, 2021 32

interactions in massive IoT systems

 how interactions in massive IoT systems
 were envisioned:
 • mostly uplink transmissions
 • assumption reflected in a lot of
 mainstream mMTC systems,
 such as LoRa and SigFox

 at least two drivers to change
 the patterns of interaction
 • smart contracts and blockchains
 • distributed learning

 WiLab - Huawei JIC Workshop, January 11, 2021 33

smart contract
mple Use Case II: Transportation Rental
 § a digital contract that combines
 The smart The smart
Jane uses a – programmed computer
 Upon deposit,
 the smart
 Janefunctions
 returns
 contract
 monitors
 bike to a
 contract
 transfers funds
mart contract contract
 different rental
 speed and to RentCo and
 – logic for legal documents
o rent a ride. unlocks bike
 for use.
 point.
 distance with relocks the
 ride tracking. bike.

 § example (from labCFTC)
 Smart
 RentCo

 RTS

 Smart Contract Running
 Tracks and transfers fees, fines, payments and refunds. Alerts RentCo if Jane Ride Tracking
 strays outside of service area. Locks/unlocks ride. Service
 Jane Oracle tracks location,
 speed, and accidents.
 Jane can see records
 of her rides on the
 Smart Contract
 Blockchain.
 14

 LabCFTC, A Primer on Smart Contracts, https://www.cftc.gov/sites/default/files/2018-
 11/LabCFTC_PrimerSmartContracts112718.pdf

 WiLab - Huawei JIC Workshop, January 11, 2021

IoT and smart contracts
 IoT smart contracts

 distributed
 IoT IoT
 decentralized
 apps ledger
 connectivity

 connectivity
 security
§ massive amount of transactions and contracts
 among autonomous connected devices

 WiLab - Huawei JIC Workshop, January 11, 2021 35

example: pollution monitoring with IoT and blockchain

 Marketplace
 Connectivity

 Distributed
Distributed Trust
Transparency Blockchain Pollution
Tamper-Proof Smart Contract
 Monitoring
 Sensing

 Data Analytics

 Connectivity Sensing Data Analytic Micro-Transactions Marketplace

 WiLab - Huawei JIC Workshop, January 11, 2021 36

example: pollution monitoring with IoT and blockchain

A sequence of message exchanges between DLT with UEs and eNB

 L. D. Nguyen, A. E. Kalor, I. Leyva-Mayorga and P. Popovski, "Trusted
 Wireless Monitoring Based on Distributed Ledgers over NB-IoT Connectivity,"
 in IEEE Communications Magazine, vol. 58, no. 6, pp. 77-83, June 2020

 WiLab - Huawei JIC Workshop, January 11, 2021 37

example: smart data trading

 three trading protocols
 § general trading
 § buying on demand
 § selling on demand

L. D. Nguyen, I. Leyva-Mayorga, A. Lewis, and P. Popovski, "Modeling and
Analysis of Market Blockchain-based Data Trading in NB-IoT Networks,"
available at https://arxiv.org/abs/2010.06003, October 2020

 WiLab - Huawei JIC Workshop, January 11, 2021 38

outlook
§ interaction among different generic traffic types

§ spectrum usage to be redefined by heterogeneity

§ better understanding of timing requirements

§ new opportunities in spatial processing

§ distributed ledgers and learning may fundamentally change
 the interactions and traffic patterns in massive IoT

 WiLab - Huawei JIC Workshop, January 11, 2021 39