CS-541

Wireless Sensor Networks

Lecture 6: Radio Duty Cycling for Wireless Sensor Networks

Spring Semester 2016-2017

Prof Panagiotis Tsakalides, Dr Athanasia Panousopoulou, Dr Gregory Tsagkatakis

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 1
University of Crete, Computer Science Department
Today’s Objectives

• Why do we need duty cycling

• Categorization of RDC techniques: synchronous – asynchronous

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 2
University of Crete, Computer Science Department
 Why do we need duty cycling for WSN
• Sensor Nodes states: TX, RX, Idle, Sleep

• Idle Listening: radio is on, channel scanning, but no tx-rx activity.

• Power consumed during idle listening depends on the transceiver / hardware

platform characteristics.

Casilari, Eduardo; Cano-García, Jose M.; Campos-Garrido, Gonzalo. 2010. "Modeling of Current Consumption in 802.15.4/ZigBee
CS-541 Wireless Sensor Networks
SensorSpring Semester
Motes." 2016-2017
Sensors 10, no. 6: 5443-5468. University of Crete, Computer Science Department
3
We do we need duty cycling for WSN
• On idle listening: huge amounts on energy loss especially on low-traffic
conditions.

• Network becomes dependent on the capacity of the battery and the

quality of the battery cells (depth of discharge and how fast it discharges)

• Regardless of the network traffic and sensor nodes activity (e.g. relays or
not) lifetime is the same
12 V batteries: ~3 weeks of
AA (1.5x2) batteries: lifetime (GG Bridge project).
~6 days of lifetime

vs.

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 4
University of Crete, Computer Science Department
We do we need duty cycling for WSN
• Example: 2 small-scale testbeds (10-15 nodes), same protocol stack,
operating at different environments (rural and industrial) and having
different traffic demands
Empirical characterization of discharge of AA batteries

@ 6 seconds - Industrial

High relay activity

@ 60 seconds - Rural
Low relay activity
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 5
University of Crete, Computer Science Department
We do we need duty cycling for WSN
• Example: 2 small-scale testbeds (10-15 nodes), same protocol stack,
operating at different environments (rural and industrial) and having
different traffic demands
Empirical characterization of lifetime (Available Energy / Consumed Power)

@ 60seconds - Rural @ 6 seconds - Industrial

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 6
University of Crete, Computer Science Department
Duty Cycling

• Radio Duty Cycling is an energy-saving

technique for reducing the amount of energy
consumed during idle listening.

• Radio Duty Cycling corresponds to scheduling

sensor nodes between active (tx, rx, idle) and
sleep operational modes

• Challenges:
• How to cope with extra latency (sampling on
sleep mode -> queued until active period
• Relay node wakes up to forward the packet –
packet is lost or delayed
• Communication connectivity – network “Wireless Sensor Networks”, Ian F. Akyildiz,
Mehmet Can Vuran, 2010, Willey, Ch 1
partitioning

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 7
University of Crete, Computer Science Department
 Duty Cycling Categorization
• Synchronization between sensor nodes to accommodate the
same sleep/active schedule (how do we do that?)
• All nodes share the same clock (not the same as TDMA –
they can still compete for accessing the medium)
Synchronous • Communication is always feasible during the active periods
• Link disconnections are avoided
• Not efficient for not fully connected networks or scalable
multi-hop topologies (latency)
• Duty cycle is selected before network deployment - how to
deal with traffic fluctuations

• No need for synchronization between sensor nodes

• Each node preserve its own clock and independent
active/sleep periods
• More resilient to network dynamics
• More appropriate for large-scale networks & best effort traffic
Asynchronous
• Communication might not be feasible
• Nodes may suffer from intense interference due to excessive
retransmissions
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 8
University of Crete, Computer Science Department
Synchronous Duty Cycling

• If data sampled by a source node during its sleep period have to be

queued until the active period

• An intermediate node has to wait until the next hop wakes up to

forward the packet

Approaches:

• Interference is not considered as a problem

• Interference aspects are addressed as is part of the solution – resulting to

a contention-based MAC protocol

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 9
University of Crete, Computer Science Department
 Synchronous Duty Cycling

• Interference is not considered part of the problem

• Time is slotted and interference can be resolved within each slot
(contention-based)
• Goal: small communication delay between any pair of nodes.
“Guide to Wireless Sensor Networks”, S. Misra, I. Woungang, S. C. Misra,
2009, Springer, Ch 15
Each sensor node is active in exactly one of k G(V) = (V, E)
slot, even if it does not have data to transmit

slot assignment function f(u): V(G) -> [0, k-1] :

• Each sensor node u is active only in slots i*k + f(u), i = 0,1,2,...

• If a sensor node has a packet to forward to a neighbor, it can wake up in the active slot of
that neighbor and transmit the packet.
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 10
University of Crete, Computer Science Department
 Synchronous Duty Cycling

• Interference is not considered part of the problem

• Time is slotted and interference can be resolved within each slot
(contention-based)
• Goal: small communication delay between any pair of nodes.

f(u) = 0, if u V(G) -> [0, k-1] : G(V) = (V, E)

0, 𝑖𝑓 𝑢 ∈ 𝑎, 𝑐, 𝑓
Example (a): 𝑓 𝑢 =
2 𝑖𝑓 𝑢 ∈ 𝑒, 𝑑, 𝑏

0, 𝑖𝑓 𝑢 ∈ 𝑎, 𝑏 “Guide to Wireless Sensor Networks”, S. Misra, I. Woungang, S. C. Misra,

Example (b): 𝑓 𝑢 = 1, 𝑖𝑓 𝑢 ∈ 𝑒, 𝑐 2009, Springer, Ch 15

2 𝑖𝑓 𝑢 ∈ 𝑑, 𝑓
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 11
University of Crete, Computer Science Department
 Synchronous Duty Cycling

• Interference is not considered part of the problem

• This means that: (a) time is slotted and interference can be resolved
within each slot (contention-based)
• Goal: small communication delay between any pair of nodes.
“Guide to Wireless Sensor Networks”, S. Misra, I. Woungang, S. C. Misra,
2009, Springer, Ch 15
Delay between two nodes u, v: G(V) = (V, E)

Delay along a path P:

Delay diameter of a duty cycled network is defined as the largest delay

CS-541 Wireless Sensor Networks
distance between any two nodes
Spring Semester 2016-2017
University of Crete, Computer Science Department
12
Synchronous Duty Cycling

• Interference aspects are addressed as is part of the solution – resulting

to a contention-based MAC protocol
• S-MAC: exploits the synch-based RTS/CTS scheme
Wei Ye; Heidemann, J.; Estrin, D., "An energy-efficient MAC protocol for wireless sensor networks," INFOCOM 2002.

Active period
Wakeup period

SYCH but nothing to send

Sleep period

No SYNCH & performs RTS

For SYNCH For RTS For CTS

Updates the schedule & performs RTS
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 13
University of Crete, Computer Science Department
Synchronous Duty Cycling
• Interference aspects are addressed as is part of the solution – resulting
to a contention-based MAC protocol
• S-MAC: exploits the synch-based RTS/CTS scheme

How to go from SYNCH packets to synchronized nodes: Creation of virtual

clusters w.r.t. to wakeup schedule
• First-come, first served: Nodes try to follow a schedule from 1-hop
neighbors.
• If not found, they start their own and let other nodes to follow
• Nodes having “followers” and pick up a new schedule, have to adopt both
schedules.
• Time-synchronization packet ~per 15 s

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 14
University of Crete, Computer Science Department
 Synchronous Duty Cycling
• Interference aspects are addressed as is part of the solution – resulting
to a contention-based MAC protocol
• S-MAC: exploits the synch-based RTS/CTS scheme
Holger Karl, Andreas Willig, Protocols and Architectures for Wireless Sensor Systems, 2005, Willey, Ch. 5

Scheduler

Follower
Follower

Scheduler
Follower
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 15
University of Crete, Computer Science Department
 Synchronous Duty Cycling
• Interference aspects are addressed as is part of the solution – resulting to a
contention-based MAC protocol
• S-MAC
@ large scale, multi-hop networks: nodes following more than one schedule
consume more energy
@ all nodes on the same schedule: per-hop latency equals to the sleep period

Dynamic traffic: nodes don’t change their wake-up / sleep intervals – if the
traffic becomes lighter nodes will spend time in idle mode

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 16
University of Crete, Computer Science Department
 Synchronous Duty Cycling
• Interference aspects are addressed as is part of the solution – resulting to a
contention-based MAC protocol
• T-MAC
An improvement of S-MAC - Retaining the creation of virtual clusters
Adaptive duty cycling: adaptive duration of active (& sleep) time.
Burst all messages during the active time –
Go back to sleep mode when no traffic has happened for a certain time (TA)
T. van Dam and K. Langendoen. An adaptive energy-efficient MAC protocol for wireless sensor networks. In ACM Sensys, 2003.

TA determines the minimal amount of idle listening per frame.

CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 17
University of Crete, Computer Science Department
 Synchronous Duty Cycling

T. van Dam and K. Langendoen. An adaptive energy-efficient MAC protocol for wireless sensor networks. In ACM Sensys, 2003.
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 18
University of Crete, Computer Science Department
 Synchronous Cycling

• Interference aspects are addressed as is part of the solution – resulting to a

contention-based MAC protocol
• Adaptive listening – to improve delay over hop transmissions (S-MAC)

• Node overhearing neighbors’ transmissions

stay awake for a short period of time at the end
of the transmission.
• Neighbors are able to immediately pass the
data to it instead of waiting for its scheduled
listen time.
• If the node does not receive anything during
the adaptive listening, it will go back to sleep
until its next scheduled listen time.

CS-541 Wireless Sensor Networks

Spring Semester and
W. Ye, J. Heidemann, 2016-2017
D. Estrin. Medium access control 19
Universitywith coordinated
of Crete, adaptive
Computer Science sleeping for wireless sensor networks. IEEE/ACM
Department
Transactions on Networks, 12(3):493–506, 2004.
 Synchronous Cycling

W. Ye,Spring
J. Heidemann, CS-541with
and D. Estrin. Medium access control Wireless Sensor Networks
coordinated adaptive sleeping for wireless sensor networks. IEEE/ACM
Semester 2016-2017 20
Transactions on Networks, 12(3):493–506, 2004. University of Crete, Computer Science Department
 Asynchronous Duty Cycling
• Preamble Sampling

The preamble transmission

time should be at least equal to the
duty-cycling period!!
Typical preamble sizes implemented >> Theoretical values

Long preambles decrease the effective

channel capacity & increase the receiving
and listening overhead.

Overhearing nodes will have to be awake for

half of the preamble transmission time before receiving the destination address information
in the packet header and going back to sleep if they are not the intended receivers.
Raghunathan, V.; Ganeriwal, S.; Srivastava, M., "Emerging techniques for long lived wireless
sensor networks," Communications Magazine, IEEE , vol.44, no.4, pp.108,114, April 2006
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 21
University of Crete, Computer Science Department
 Asynchronous Duty Cycling

B-MAC
• Combines Preamble sampling (Low Power
Listening) + CCA
• No synch phase, no RTS/CTS (default)
• Node goes back to sleep after successful
packet reception or time out period
• @ TX: Preamble length matches channel
checking period
• Interval between LPL samples is maximized
so that the time spent sampling the
channel is minimized.
Polastre, Joseph, Jason Hill, and David Culler. "Versatile low power
media access for wireless sensor networks." SENSYS, ACM, 2004.

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 22
University of Crete, Computer Science Department
Polastre, Joseph, Jason Hill, and David Culler. "Versatile low power media access for wireless sensor networks." SENSYS,
ACM, 2004.

BMAC is
about 4.5
faster than
SMAC-unicast

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 23
University of Crete, Computer Science Department
Polastre, Joseph, Jason Hill, and David Culler. "Versatile low power media access for wireless sensor networks."
SENSYS, ACM, 2004.

Low data rates:

SMAC is better
BMAC: larger
preambles at
low
throughput,
progressively
becoming
smaller

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 24
University of Crete, Computer Science Department
Asynchronous Duty Cycling

• ContikiMAC

Periodical wake-ups to listen for

packet transmissions from
neighbors.

If a packet transmission is detected during a

wake-up, the receiver is kept on
Dunkels, Adam. "The contikimac radio duty cycling protocol."
(2011).
Link layer acknowledgments from the receiver

Repeated transmissions until a link layer acknowledgment from the receiver.

Or within the full wake-up ensure (broadcast)
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 25
University of Crete, Computer Science Department
 Asynchronous Duty Cycling

• ContikiMAC

Timing Aspects…

Dunkels, Adam. "The contikimac radio duty cycling protocol." (2011).

𝑡𝑖 : the interval between each packet transmission.

𝑡𝑟 : the time required for a stable RSSI, needed for a stable CCA indication.
𝑡𝑐 : the interval between each CCA.
𝑡𝑎 : the time between receiving a packet and sending the
acknowledgment packet.
𝑡𝑑 : the time required for successfully detecting an ACK from the receiver.
𝑡𝑠 : transmission time of the shortest packet
CS-541 Wireless Sensor Networks
Spring Semester 2016-2017 26
University of Crete, Computer Science Department
Asynchronous Duty Cycling

• ContikiMAC

Timing Aspects…

Dunkels, Adam. "The contikimac radio duty cycling protocol." (2011).

𝑡𝑖 < 𝑡𝑐 To ensure that at least one of the two CCA will catch the transmission attempt

𝑡𝑠 > 2𝑡𝑟 + 𝑡𝑐 Data packet should be longer than 2 successive CCA detection periods
𝑡𝑎 + 𝑡𝑑 < 𝑡𝑖

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 27
University of Crete, Computer Science Department
 Asynchronous Duty Cycling

• ContikiMAC

Timing Aspects…

Dunkels, Adam. "The contikimac radio duty cycling protocol." (2011).

Based on symbol duration & channel bit rate you can calculate the min PHY payload size
E.g., IEEE 802.15.4 link layer:
𝑡𝑎 = 12 symbols (1 symbol = 4/250 ms)
An IEEE 802.15.4 receiver can reliably detect the reception of an ACK after the 4-byte long
preamble and the 1-byte start of frame delimiter, which totals to 10 symbols
𝑡𝑟 = 0.192 ms (HW dependent)

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 28
University of Crete, Computer Science Department
Practical Aspects
• Side effects of bad RDC synchronization

A single TX-RX pair, scanning different channels, and different tx power

On weak links:
Repeatedly retransmitted packets due to asynchronous policy can be
perceived as noise (multi-path effect…)

Intra-network interference

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 29
University of Crete, Computer Science Department
 Practical Aspects
No RDC ACK-based RDC ACK-based RDC – poor synch
PRR
NF

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 30
University of Crete, Computer Science Department
 Practical Aspects
• Side effects of bad RDC synchronization

Multi-hop networks with low data rate

Destabilizing network behavior, due to

low data rate (nodes that are not
relaying packets)

Keeping the control / sync messages

could alleviate this issue

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 31
University of Crete, Computer Science Department
References and Material for Reading
1. Casilari, Eduardo; Cano-García, Jose M.; Campos-Garrido, Gonzalo. 2010. "Modeling of Current
Consumption in 802.15.4/ZigBee Sensor Motes." Sensors 10, no. 6: 5443-5468.
2. “Guide to Wireless Sensor Networks”, S. Misra, I. Woungang, S. C. Misra, 2009, Springer, Ch 15
3. W. Ye, J. Heidemann, and D. Estrin. Medium access control with coordinated adaptive sleeping
for wireless sensor networks. IEEE/ACM Transactions on Networks, 12(3):493–506, 2004.
4. Dunkels, Adam. "The contikimac radio duty cycling protocol." (2011).
5. Raghunathan, V.; Ganeriwal, S.; Srivastava, M., "Emerging techniques for long lived wireless
sensor networks," Communications Magazine, IEEE , vol.44, no.4, pp.108,114, April 2006
6. T. van Dam and K. Langendoen. An adaptive energy-efficient MAC protocol for wireless sensor
networks. In ACM Sensys, 2003.
7. Holger Karl, Andreas Willig, Protocols and Architectures for Wireless Sensor Systems, 2005,
Willey, Ch. 5
8. Wei Ye; Heidemann, J.; Estrin, D., "An energy-efficient MAC protocol for wireless sensor
networks," INFOCOM 2002. Twenty-First Annual Joint Conference of the IEEE Computer and
Communications Societies. Proceedings. IEEE , vol.3, no., pp.1567,1576 vol.3, 2002
9. Polastre, Joseph, Jason Hill, and David Culler. "Versatile low power media access for wireless
sensor networks." SENSYS, ACM, 2004.

CS-541 Wireless Sensor Networks

Spring Semester 2016-2017 32
University of Crete, Computer Science Department