From IEEE 802.15.4 to IEEE 802.15.

4e
Another Step towards the
Internet of (Important) Things

Giuseppe Anastasi
Pervasive Computing & Networking Lab (PerLab)
Dept. of Information Engineering, University of Pisa
E-mail: giuseppe.anastasi@iet.unipi.it
Website: www.iet.unipi.it/~anastasi/
PerLab

Sun Yat-Sen University

Guangzhou, China, May 16, 2014

Internet of Things
PerLab

The next logical step in the technological revolution connecting people

anytime, anywhere is to connect inanimate objects. This is the vision
underlying the Internet of things: anytime, anywhere, by anyone and
anything
(ITU, Nov. 2005)
More than 26 billions devices
will be wirelessly connected to
the Internet of Things by 2020
computers and communication devices
cars, robots, machine tools
persons, animals, and plants
garments, food, drugs, etc.

From IEEE 802.15.4 to IEEE 802.15.4e

Wireless Sensor & Actuator Networks (WSANS)

PerLab

WSANs will behave as a digital skin for the IoT

providing a virtual layer through which any
computational system can interact with the physical
world

From IEEE 802.15.4 to IEEE 802.15.4e

Standards for WSANs

PerLab

IEEE 802.15.4/802.15.4e
ZigBee
Bluetooth
WirelessHART
ISA-100.11a
6LoWPAN
IPv6 over Low power WPAN

RPL
Routing Protocol for Low power
and Lossy networks

CoAP
Constrained Application Protocol
From IEEE 802.15.4 to IEEE 802.15.4e

IEEE 802.15.4 Standard

Reference technology for WSANs
Expected to be a major enabling
technology also for IoT

Overview
PerLab

IEEE 802.15.4
MAC protocols

Limits of IEEE 802.15.4 MAC

Performance and reliability

Possible enhancements
IEEE 802.15.4e
Time-Slotted Channel Hopping (TSCH) mode

Open Issues

From IEEE 802.15.4 to IEEE 802.15.4e

IEEE 802.15.4 standard

PerLab

Standard for Personal Area Networks (PANs)

low-rate and low-power
PHY and MAC layers

Main features
transceiver management
channel access
PAN management
From IEEE 802.15.4 to IEEE 802.15.4e

802.15.4 Network Topologies

PerLab

Ed Callaway, Paul Gorday, Lance Hester, Jose A. Gutierrez, Marco Naeve, Bob Heile, Home Networking with IEEE 802.15.4:
Developing Standard for Low-Rate Wireless Personal Area Networks, IEEE Communications Magazine, August 2002.
From IEEE 802.15.4 to IEEE 802.15.4e

Channel frequencies
PerLab

Europe

USA

(20 Kbps)

(40 Kbps)

(250 Kbps)

Ed Callaway, Paul Gorday, Lance Hester, Jose A. Gutierrez, Marco Naeve, Bob Heile, Home Networking with IEEE 802.15.4:
Developing Standard for Low-Rate Wireless Personal Area Networks, IEEE Communications Magazine, August 2002.
From IEEE 802.15.4 to IEEE 802.15.4e

802.15.4 MAC protocols

PerLab

Two different channel access methods

Beacon-Enabled duty-cycled mode
Non-Beacon Enabled mode (aka Beacon Disabled mode)
MAC

Non-beacon enabled

Beacon enabled
Superframe Structure

Contention based

Contention based

Contention free

Unslotted CSMA-CA

Slotted CSMA-CA

Reserved time slot

From IEEE 802.15.4 to IEEE 802.15.4e

10

802.15.4 Beacon Enabled mode

PerLab

Active
Beacon

Beacon

GTS

CAP

GTS

Inactive

CFP

SD = aBaseSuperFrameDuration*2SO sym
BI = aBaseSuperFrameDuration*2BO sym

From IEEE 802.15.4 to IEEE 802.15.4e

11

CSMA/CA: Beacon-Enabled mode

PerLab

At each trial the backoffwindow size is doubled

Yes

Only a limited number of

attempts is permitted
(macMaxCSMABackoffs)

No

No

No

Yes

From IEEE 802.15.4 to IEEE 802.15.4e

Yes

12

CSMA/CA: Non-Beacon Enabled mode

PerLab

Un-slotted CSMA

NB = 0,
BE = macMinBE

Delay for
random(2BE - 1) unit
backoff periods

At each trial the backoffwindow size is doubled

Perform CCA

Only a limited number of

attempts is permitted
(macMaxCSMABackoffs)

Y
Channel idle?
N
NB = NB+1,
BE = min(BE+1, aMaxBE)

NB>
macMaxCSMABackoffs
?

Y
Failure

From IEEE 802.15.4 to IEEE 802.15.4e

Success

13

Acknowledgement Mechanism
PerLab

Optional mechanism
Destination Side
ACK sent upon successful reception of a data frame

Sender side
Retransmission if ACK not (correctly) received within the
timeout
At each retransmission attempt the backoff window size
is re-initialized
Only a maximum number of retransmissions allowed
(macMaxFrameRetries)

From IEEE 802.15.4 to IEEE 802.15.4e

14

Limits of IEEE 802.15.4 MAC

PerLab

MAC Unreliability
Unbounded latency
No guaranteed bandwidth
Unless GTS is used
GTS only provides a limited service (7 slots)

No built-in frequency hopping technique

Prone to failures due to interferences and multi-path
fading

G. Anastasi, D. De Guglielmo, A Seghetti, From IEEE 802.15.4 to IEEE 802.15.4e: a Step towards the Internet of Things,
Chapter 10 in Advances onto the Internet of Things (S. Gaglio, G. Lo Re, Editors), pp. 135-152. January 2014. Springer
From IEEE 802.15.4 to IEEE 802.15.4e

15

802.15.4 CSMA/CA Performance

PerLab

G. Anastasi, M. Conti, M. Di Francesco, A Comprehensive Analysis of the MAC Unreliability Problem in IEEE 802.15.4
Wireless Sensor Networks, IEEE Transactions in Industrial Informatics, Vol. 7, N. 1, Feb 2011.
From IEEE 802.15.4 to IEEE 802.15.4e

16

Key Question
Why the 802.15.4 MAC
Unreliability Problem?

Possible Answers
PerLab

The access method is CSMA-CA

Contention increases with the # of active nodes

The periodic Beacon synchronizes nodes

accesses
All sensor nodes contend for channel access upon
receiving a beacon

How about other CSMA-based MAC protocols

operating in same conditions?

From IEEE 802.15.4 to IEEE 802.15.4e

18

802.15.4 MAC vs. S-MAC

PerLab

G. Anastasi, M. Conti, M. Di Francesco, A Comprehensive Analysis of the MAC Unreliability Problem in IEEE 802.15.4
Wireless Sensor Networks, IEEE Transactions in Industrial Informatics, Vol. 7, N. 1, Feb 2011.
From IEEE 802.15.4 to IEEE 802.15.4e

19

CSMA/CA parameters
PerLab

Parameter

2003 release

2006 release

Notes

07

Max number of retransmissions

macMaxFrameRetries

(aMaxFrameRetries)

macMaxCSMABackoff

05
Default: 4

05
Default: 4

Max number of
backoff stages

38

macMaxBE

5
(aMaxBE)

Maximum Backoff
Window Exp.

macMinBE

03
Default: 3

From IEEE 802.15.4 to IEEE 802.15.4e

Default: 3

Default: 5
07
Default: 3

Minimum Backoff
Window Exp.

20

CSMA parameter values

PerLab

DPS (Default Parameter Set)

Default parameter values defined in the standard

SPS (Standard Parameter Set)

Larger parameter values allowed by the standard

NPS (Non-standard Parameter Set):

Larger parameter values not allowed by the standard

macMinBE

macMaxBE

macMaxCSMABackoff

macMaxFrameRetries

DPS

SPS

NPS

10

10

10

From IEEE 802.15.4 to IEEE 802.15.4e

21

Delivery Ratio vs. Packet Error Rate

PerLab

G. Anastasi, M. Conti, M. Di Francesco, A Comprehensive Analysis of the MAC Unreliability Problem in IEEE 802.15.4
Wireless Sensor Networks, IEEE Transactions in Industrial Informatics, Vol. 7, N. 1, Feb 2011.
From IEEE 802.15.4 to IEEE 802.15.4e

22

Latency and Energy vs. Packet Error Rate

PerLab

Avg. Latency

Energy/msg

G. Anastasi, M. Conti, M. Di Francesco, A Comprehensive Analysis of the MAC Unreliability Problem in IEEE 802.15.4
Wireless Sensor Networks, IEEE Transactions in Industrial Informatics, Vol. 7, N. 1, Feb 2011.
From IEEE 802.15.4 to IEEE 802.15.4e

23

Learned Lesson
PerLab

MAC unreliability mainly due to CSMA/CA

Periodic Beacons synchronize channel accesses
thus maximizing contention and increasing the collision
probability

Default MAC parameter values not appropriate

More appropriate settings can alleviate the problem
However the optimal setting depends on
Applications requirements
Operating conditions

From IEEE 802.15.4 to IEEE 802.15.4e

24

Enhancements
How to achieve the desired reliability
level (e.g., 80%) with the minimum energy
expenditure?

Strategies for Optimal Settings

PerLab

Model-based offline computation

Model-based online adaptation
Measurement-based adaptation
Learning-based adaptation

[1]
[2]
[3]
[4]

[1]

P. Park, P. Di Marco, P. Soldati, C. Fischione, K. H. Johansson, A Generalized Markov Chain Model for Effective
Analysis of Slotted IEEE 802.15.4, Proc. IEEE International Conference on Mobile Ad Hoc and Sensor Systems
(MASS 2009). Macao, China, October 2009

[2]

P. Park, P. Di Marco, C. Fischione, K. H. Johansson, Modeling and Optimization of the IEEE 802.15.4 Protocol for
Reliable and Timely Communications, IEEE Transactions on Parallel and Distributed Systems, 2012.

[3]

M. Di Francesco, G. Anastasi, M. Conti, S. K. Das, V. Neri, Reliability and Energy-efficiency in IEEE 802.15.4/ZigBee
Sensor Networks: An Adaptive and Cross-layer Approach, IEEE Journal on Selected Areas in Communications,
2011.

[4]

S. Brienza, D. De Guglielmo, C. Alippi, G. Anastasi, G. Anastasi, M. Roveri, A Learning-based Algorithm for Optimal
MAC Parameters Setting in IEEE 802.15.4 Wireless Sensor Networks, Proc. ACM International Symposium on
Performance Evaluation of Wireless Ad Hoc, Sensor, and Ubiquitous Networks (PE-WASUN 2013), Barcelona, Spain,
November 3-7, 2013

From IEEE 802.15.4 to IEEE 802.15.4e

26

Parks Model
PerLab

Markov Chain Model

Assumptions
Star network
Ideal channel
Fixed and known # of nodes

Reliability (R)
Latency (D)
Energy per packet (Etot)
As functions of the 802.15.4 MAC
parameters
P. Park, P. Di Marco, P. Soldati, C. Fischione, and K. Johansson, A Generalized Markov Chain Model for Effective Analysis of Slotted
IEEE 802.15.4, Proc. IEEE International Conference on Mobile Ad hoc and Sensor Systems (MASS 2009), October 2009, pp. 130 139.
From IEEE 802.15.4 to IEEE 802.15.4e

27

Model-based Offline Computation

PerLab

Offline Algorithm
Provides the MAC
parameters set that
guarantees
(R>Rmin ) & (D<Dmax)

while minimizing
Etot

From IEEE 802.15.4 to IEEE 802.15.4e

28

Limits of Model-based Computation

PerLab

Offline approach
Requires to know the operating conditions of the
sensor network
unsuitable for real environments

Requires a significant amount of time

to solve the Parks model and derive the optimal MAC
parameter setting
About 400s (i.e., around 7 minutes) with a quad-core PC

From IEEE 802.15.4 to IEEE 802.15.4e

29

Model-based Adaptation
PerLab

Online Algorithm
Estimates , , and through direct measurements
: probability to find the channel busy during the first CCA
: probability to find the channel busy during the second CCA
: probability that the node will attempt the first CCA in a generic
backoff slot

Introduces , , and in the Parks model

Derives R, D and Etot
for any MAC parameters set

Selects the optimal parameter setting

P. Park, P. Di Marco, C. Fischione, K. H. Johansson, Modeling and Optimization of the IEEE 802.15.4 Protocol for Reliable and
Timely Communications, IEEE Transactions on Parallel and Distributed Systems, 2012.
From IEEE 802.15.4 to IEEE 802.15.4e

30

Limits of model-based algorithms

PerLab

Effectiveness bounds to the model accuracy

Assumptions on packet arrival process typically does
not fit real scenarios
The communication channel is never ideal

Requires to know configuration parameters

They may change over time (e.g., # of sensor nodes)

Unsuitable for multi-hop WSNs

From IEEE 802.15.4 to IEEE 802.15.4e

31

Measurement-based Adaptation
PerLab

ADAPT: Adaptive Access Parameters Tuning

Completely heuristic approach
Provides the desired reliability with minimum energy
consumption (and latency)
Based on local measurements of unreliability
No a-priori knowledge about operating conditions
Suitable for real-life scenarios

M. Di Francesco, G. Anastasi, M. Conti, S. K. Das, V. Neri, Reliability and Energy-efficiency in IEEE 802.15.4/ZigBee Sensor
Networks: An Adaptive and Cross-layer Approach, IEEE Journal on Selected Areas in Communications, 2011.
From IEEE 802.15.4 to IEEE 802.15.4e

32

AdAPT: Unreliability Measurements

PerLab

Sources of message loss

Congestion (Contention)
Noisy channel (fading, interferences, )

Unreliability Measurements
Messages discarded due to exceeded # of backoffs
congestion

Messages discarded due to exceeded # of rtx

Channel errors
Collisions (can be neglected in the considered scenario)

From IEEE 802.15.4 to IEEE 802.15.4e

33

AdAPT: Adaptation Strategy

PerLab

From IEEE 802.15.4 to IEEE 802.15.4e

34

Limits of ADAPT
PerLab

Oscillations
ADAPT tends to oscillate between adjacent sets
Thus consuming more energy than necessary

Memory-less
ADAPT does not take advantage of a recurrent states

From IEEE 802.15.4 to IEEE 802.15.4e

35

LEAP
PerLab

LEAP: LEarning-based Adaptive Parameter Tuning

based on local measurements, like ADAPT
does not require any input parameter
it can be used in a real scenario

learns the past history

it can take advantage of recurrent states in the
system

behaves very close to an ideal algorithm

S. Brienza, D. De Guglielmo, C. Alippi, G. Anastasi, M. Roveri, A Learning-based Algorithm for Optimal MAC Parameters
Setting in IEEE 802.15.4 Wireless Sensor Networks, Proceedings of ACM International Symposium on Performance
Evaluation of Wireless Ad Hoc, Sensor, and Ubiquitous Networks (PE-WASUN 2013), Barcelona, Spain, November 3-7, 2013.
From IEEE 802.15.4 to IEEE 802.15.4e

36

LEAP Algorithm
PerLab

LEAP selects the optimal parameters setting

using the knowledge learned in the past history.

LEAP algorithm
Exploration Phase
An adaptive approach is used to set CSMA-CA parameters
Sensor node builds a Learning Table using the knowledge
learned during the Exploration phase.
Then it enters the Exploitation phase

Exploitation Phase
Sensor node checks the Learning Table to derive the optimal
setting, given the current operating conditions.

From IEEE 802.15.4 to IEEE 802.15.4e

37

Performance Comparison
PerLab

PHY layer
Bit Rate
Sensor nodes

2.4 GHz
250 Kbps
from 1 to 50

Distance from Coordinator Node

10m

CS range

30m

RX range

15m

Traffic Generation
Message Size
Messages per Period
Message Loss Rate

Periodic
100 bytes
10
0-%

Coordinator always ON

From IEEE 802.15.4 to IEEE 802.15.4e

38

Performance Measures
PerLab

Packet Delivery Ratio

ratio between # of packets correctly received by the sink, and the total # of
packets generated by all sensor nodes.

Miss ratio
fraction of times when the packet delivery probability calculated over the current
Beacon Interval drops below the threshold required by the application

Average Energy per Packet

total energy consumed by each sensor node divided by the # of data packets
correctly delivered to the sink.

Average Latency
average time from when the packet transmission starts at the source node to
when the packet is correctly received by the sink

From IEEE 802.15.4 to IEEE 802.15.4e

39

Performance in Stationary Conditions

PerLab

Requirements
Delivery Ratio: > 80%
Miss Ratio:
< 15%
Delivery ratio (%)

From IEEE 802.15.4 to IEEE 802.15.4e

Miss Ratio (%)

40

Performance in Stationary Conditions

PerLab

Energy per packet (mJ)

From IEEE 802.15.4 to IEEE 802.15.4e

Latency (ms)

41

Summary
PerLab

The desired reliability can be achieved

Through an appropriate CSMA/CA parameter setting

At the cost of
Increased energy consumption
Increased latency

From IEEE 802.15.4 to IEEE 802.15.4e

42

Other limits of 802.15.4

PerLab

Unbounded latency
Due to contention-based CSMA/CA algorithm

No guaranteed bandwidth
Unless GTS is used
GTS only provides a limited service (7 slots)

No built-in frequency hopping technique

Prone to failure due to interferences and multi-path
fading

From IEEE 802.15.4 to IEEE 802.15.4e

43

Limits of 802.15.4/ZigBee
PerLab

Unbounded latency
Due to contention-based CSMA/CA algorithm
No guaranteed bandwidth
Unless GTS is used
GTS only provides a limited service (7 slots)
No built-in frequency hopping technique
Prone to failure due to interferences and multi-path
fading

From IEEE 802.15.4 to IEEE 802.15.4e

44

Link Reliability vs. Channel

PerLab

Connectivity traces were collected by J. Ortiz and D. Culler in a UC Berkeley office space (traces are made available
at http://wsn.eecs.berkeley.edu/connectivity/). 46 IEEE802.15.4-compliant TelosB motes are deployed in a 50m by 50m
indoor environment, and are constantly listening for packets. One after the other, each mote transmits a burst of 100
packets, with a 20ms inter-packet time and a transmission power of 0dBm, on each of the 16 frequency channels which
span the 2.4-2.485GHz band.
From IEEE 802.15.4 to IEEE 802.15.4e

45

Channel Hopping
PerLab

Subsequent packets are sent at a different frequency

following a pseudo-random hopping pattern.

If a transmission fails, retransmission will happen on

a different frequency
Greater chance of successful transmission
that if the retransmission happened on the same channel
Since a different frequency means different effects of multi-path
fading and interference

Increased security
Against Selective Jamming attack

Channel Hopping used in

Bluetooth
WirelessHart
From IEEE 802.15.4 to IEEE 802.15.4e

46

Key Question
What to do in critical
scenarios?

IEEE 802.15.4e
PerLab

IEEE 802.15 Task Group 4e

chartered to define a MAC amendment to the existing
standard 802.15.4-2006.
The intent of this amendment was to enhance and
add functionality to the 802.15.4-2006 MAC
better support the industrial markets
permit compatibility with modifications being proposed
within the Chinese WPAN.

On February 6, 2012 the IEEE Standards Association

Board approved the IEEE 802.15.4e MAC
Enhancement Standard document for publication.
http://www.ieee802.org/15/pub/TG4e.html
From IEEE 802.15.4 to IEEE 802.15.4e

48

Major Changes
PerLab

MAC Behavior Modes

support of specific application domains

General functional improvements

not tied to any specific application domain.

Comment:
Many ideas taken from previous industrial standards
WirelessHART and ISA 100.11.a
slotted access, shared and dedicated slots, multichannel communication, and frequency hopping.

From IEEE 802.15.4 to IEEE 802.15.4e

49

General functional improvements

PerLab

Low Energy (LE)

Information Elements (IE)
Enhanced Beacons (EB)
Multipurpose Frame
MAC Performance Metrics
Fast Association (FastA)

From IEEE 802.15.4 to IEEE 802.15.4e

50

MAC Behavior Modes

PerLab

Radio Frequency Identification Blink (BLINK)

item and people identification, location, and tracking

Asynchronous multi-channel adaptation (AMCA):

application domains where large deployments are required (e.g., process
automation/control, infrastructure monitoring, etc.);

Deterministic & Synchronous Multi-channel Extension (DSME)

industrial and commercial applications with stringent timeliness and
reliability requirements;

Low Latency Deterministic Network (LLDN)

applications requiring very low latency requirement (e.g., factory
automation, robot control)

Time Slotted Channel Hopping (TSCH)

application domains such as process automation

From IEEE 802.15.4 to IEEE 802.15.4e

51

Deterministic & Synchronous Multi-channel Extension (DSME)

PerLab

DSME networks run on Beacon-enabled PAN

All devices synchronize to multi-superframe via
beacon frames
Multi-superframe enhances GTS in 802.15.4
By grouping multiple superframes
and extending single channel operation to multi-channel
operation
either by adapting (switching) channels or by hopping channels.

A pair of peer devices wakes up at a reserved GTS

slot to exchange a data frame and an ACK frame.

From IEEE 802.15.4 to IEEE 802.15.4e

52

Low Latency Deterministic Network (LLDN)

PerLab

Target: factory and process automation

automotive robots, machine tools, milling/turning/ robot
Very low and deterministic latency (e.g., 5-50 ms)

Designed for small networks and small frames

Operates in star topology

LLDN MAC uses a superframe, based on time slots,

to provide determinism
beacon timeslot
management timeslots (if present)
base transmission timeslots (of equal length)
The base timeslots include uplink timeslots and
bidirectional timeslots.
From IEEE 802.15.4 to IEEE 802.15.4e

53

Low Latency Deterministic Network (LLDN)

PerLab

Group Acknowledgement (GACK) slots

Configured to promote the retransmission of failed tx
in uplink timeslots.

Two categories of base timeslots

Dedicated timeslots
The slot owner has access privileges in dedicated
timeslots

Shared timeslots
devices assigned to a shared timeslot use the contentionbased access method.

From IEEE 802.15.4 to IEEE 802.15.4e

54

Time Slotted Channel Hopping (TSCH)

PerLab

Combines time slotted access, multi-channel

communication and channel hopping
Particularly suitable for multi-hop networks

Time-slotted access
Predictable and bounded latency
Guaranteed bandwidth

Multi-channel communication
More nodes can communicate at the same time (i.e., same slot) using
different channels (identified by different channel offsets)
increased network capacity

Channel hopping
mitigates the effects of interference and multipath fading
improves reliability
From IEEE 802.15.4 to IEEE 802.15.4e

55

Periodic Slotframe
PerLab
During a timeslot, one node typically sends a frame, and another sends back an acknowledgement if it
successfully receives that frame. If an acknowledgement is not received within the timeout period,
retransmission of the frame waits until the next assigned transmit timeslot (in any active slotframe) to that
address occurs.

From IEEE 802.15.4 to IEEE 802.15.4e

56

Frequency Translation
PerLab

The channel offset is translated in an operating

frequency f

ASN: total # of slots that elapsed since the

network was deployed
ASN=(kS+t) where S is the slotframe size, k the slotframe
cycle

: number of used channels

F is implemented as a look-up-table containing
the sets of available channels
nch

From IEEE 802.15.4 to IEEE 802.15.4e

57

Frequency Translation
PerLab

channelOffset

ASN 1

3 4 5

6 7

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

15

1
0
slotOffset
Slotframe cycle
Channel

0 1

2 3

4 5 6

t
2

1
ch15

From IEEE 802.15.4 to IEEE 802.15.4e

ch22

3
ch13

ch20

58

TSCH Link
PerLab

Link=Pairwise assignment of a directed communication

between devices in a specific slot, with a given channel
offset
t
chOf

From IEEE 802.15.4 to IEEE 802.15.4e

slot in the slotframe

channel offset

59

TSCH Scheduling
PerLab

The standard only explains how the MAC layer

executes a schedule
it does not specify how such a schedule is built

Centralized Scheduling
a manager node is responsible for building and maintaining the
network schedule

Distributed Scheduling
No central entity
Each node decides autonomously

From IEEE 802.15.4 to IEEE 802.15.4e

60

Centralized Scheduling
PerLab

Every node regularly updates the manager with

the list of other nodes it can hear
the amount of data it is generating

The manager draws the connectivity graph,

assigns slots to different links in the graph
Based on data generation demands

and informs each node about links in the schedule it

is participating in.
If the connectivity graph changes (i.e., a node lost a neighbor), the
manager updates its schedule and informs affected nodes

Palattella, M. R., Accettura, N., Dohler, M., Grieco, L. A., & Boggia, G., Traffic Aware Scheduling
Algorithm for reliable low-power multi-hop IEEE 802.15. 4e networks. IEEE International Symposium
on Personal Indoor and Mobile Radio Communications (PIMRC 2012), pp. 327-332, 2012
From IEEE 802.15.4 to IEEE 802.15.4e

61

Distributed Scheduling
PerLab

Each node decides locally on which links to

schedule with which neighbors.
Suitable for highly dynamic networks
e.g., networks with mobile nodes
or when the network has many gateway nodes
centralized schedules are known to be superior to distributed ones for
fairly static networks
P. Zand, A. Dilo, and P. Havinga, D-MSR: A distributed network management scheme for real-time
monitoring and process control applications in wireless industrial automation, Sensors, vol. 13, no. 7,
pp. 82398284, 2013.
Accettura, N., Palattella, M. R., Boggia, G., Grieco, L. A., & Dohler, M., Decentralized Traffic Aware
Scheduling for multi-hop Low power Lossy Networks in the Internet of Things. In IEEE International
Symposium and Workshops on a World of Wireless, Mobile and Multimedia Networks (WoWMoM 2013),
2013.
Tinka, A., Watteyne, T., & Pister, K., A decentralized scheduling algorithm for time synchronized
channel hopping. Ad Hoc Networks, pp. 201-216, Springer Berlin Heidelberg (2010)
From IEEE 802.15.4 to IEEE 802.15.4e

62

Network Formation Protocol

PerLab

Advertiser nodes emits Enhanced Beacons

(EBs) at regular times
EBs are special TSCH frame containing
Synchronization information
allows new devices to synchronize to the network

Channel hopping information

allows new devices to learn the channel hopping sequence

Timeslot information
describes when to expect a frame transmission and when to send an
acknowledgment

Initial link and slotframe information

allows new devices to know:
when to listen for transmissions from the advertising device
when to transmit to the advertising device

Network Formation Protocol

PerLab

A joining device starts scanning for possible EBs.

Upon receiving an EB from an advertiser node
the MAC layer notifies the higher layer
The higher layer initializes the slotframe and links
Using information in the received EB message

and switches the device into TSCH mode.

At this point the device is connected to the network.

Then, the device allocates communication resources

(i.e., slotframes and links)

And can start advertising, on its turn

The EB advertising policy is not part of the 802.15.4e
standard

Advertising Policy
PerLab

Random Advertisement Protocol

Each node is assigned a link
,
for
EB advertisement
Timeslots devoted to EB advertisement repeat periodically
with period
(
and
are assumed to be relatively prime)
All nodes use the same slot, possibly with different channel offset

Each advertiser node transmits its EB, using the scheduled

link, with a probability
is derived autonomously by each node
so as to minimize the collision probability
= 1
)
!(
D. De Guglielmo, A. Seghetti, G. Anastasi, M. Conti, A Performance Analysis of the Network Formation
Process in IEEE 802.15.4e TSCH Wireless Sensor/Actuator Networks, Proc. IEEE International
Symposium on Computers and Communications (ISCC 2014), Madeira, Portugal, June 23-26, 2014.

Some results
PerLab

Average Joining Time vs. Number of Links used for EBs

Conclusions
PerLab

IEEE 802.15.4 suffers many performance

limitations
in terms of both delay and packet dropping probability
due to the CSMA-CA access method and default parameter
values

and is vulnerable to fading and interferences

IEEE 802.15.4e overcomes the previous limitations
To meet the requirements of industrial and time-constrained
applications
TSCH combines
Time-slotted access
Multi-channel communication
Frequency hopping

Open Research Issues

PerLab

Link Scheduling Algorithm

Centralized vs. distributed

EB Advertisement protocol
for network formation

Interaction between TSCH and IoT protocols

Interactions between TSCH and RPL
IETF 6TiSCH standardizes mechanisms for running an
IPv6-enabled protocol stack on top of TSCH.

Additional Info on IEEE 802.15.4e and 6TiSCH

PerLab

IEEE 802.15 WPAN Task Group 4e (TG4e), [Online],

http://www.ieee802.org/15/pub/TG4e.html
http://datatracker.ietf.org/wg/6tisch/charter/
G. Anastasi, D. De Guglielmo, A Seghetti, From IEEE 802.15.4 to IEEE
802.15.4e: a Step towards the Internet of Things, Chapter 10 in Advances
onto the Internet of Things (S. Gaglio, G. Lo Re, Editors), pp. Pages 135152. January 2014. Springer. Also available at:
http://info.iet.unipi.it/~anastasi/papers/book-lo-re-10.pdf

From IEEE 802.15.4 to IEEE 802.15.4e

69

Questions