4/4/2012
1
SensorsCon 2012 March 21, Santa Clara
1
Standards-Based Reliable Wireless Sensor Networking
Thomas Watteyne, PhD Senior Networking Design Engineer Linear Technology/Dust Networks
SensorsCon 2012 March 21, Santa Clara
2
Pioneer of reliable, ultra-low power embedded wireless sensor networks (WSN)
Low power wireless SOC and networking technology leader
Acquired by Linear Technology, Dec 2011
2
M2M 100
Dust Networks Overview
Over 10,000 networks in 120 countries Data centers, oil refineries, museums, mines, chemical plants, city streets, trains,
steel mills, pharma plants…and more
4/4/2012
2
SensorsCon 2012 March 21, Santa Clara
3
CoAP
UDP
6LoWPAN
IEEE802.15.4e
The Internet of Things Stack
IEEE802.15.4
web-like interaction
Internet Integration
Low-power reliability
simple hardware
SensorsCon 2012 March 21, Santa Clara
4
IEEE802.11 (Wi-Fi)
IEEE802.15.1 (Bluetooth)
IEEE802.15.4
First Challenge: External Interference
4/4/2012
3
SensorsCon 2012 March 21, Santa Clara
5
IEEE802.11b/g/n
IEEE802.11a/n
868 MHz
433 MHz
2.4 GHz 5 GHz
IEEE802.15.4
First Challenge: External Interference
SensorsCon 2012 March 21, Santa Clara
6
Second Challenge: Multipath Fading
4/4/2012
4
SensorsCon 2012 March 21, Santa Clara
7
ch.11
Second Challenge: Multipath Fading
SensorsCon 2012 March 21, Santa Clara
8
ch.11 ch.12
0% reliability 100% reliability
Second Challenge: Multipath Fading
4/4/2012
5
SensorsCon 2012 March 21, Santa Clara
9
ch.11
ch.13
ch.15
ch.17
ch.12
ch.14
ch.16
ch.18
ch.19
ch.21
ch.23
ch.25
ch.20
ch.22
ch.24
ch.26
Second Challenge: Multipath Fading
SensorsCon 2012 March 21, Santa Clara
10
A B
Channel Hopping
4/4/2012
6
SensorsCon 2012 March 21, Santa Clara
11
IEEE802.15 Task Group 4e
- Amendment to enhance and add functionality to the 802.15.4-2006 MAC to better support the industrial markets
Time Synchronized Channel Hopping
- Nodes are synchronized on a common sense of time
- Nodes send successive packets on different frequencies using a pseudo-random hopping pattern
IEEE802.15.4e - Time Synchronized Channel Hopping
SensorsCon 2012 March 21, Santa Clara
12
16 c
hannel offsets
e.g. 31 time slots (310ms)
A
B C
D E
F G
H
I
J
A super-frame repeats over time
- Number of slots in a superframe is tunable
- Each cell can be assigned to a pair of motes, in a given direction
IEEE802.15.4e - Slotted Structure
4/4/2012
7
SensorsCon 2012 March 21, Santa Clara
13
16 c
hannel offsets
e.g. 33 time slots (330ms)
A
B C
D E
F G
H
I
J
IEEE802.15.4e - Slotted Structure
• Cells are assigned according to application requirements
SensorsCon 2012 March 21, Santa Clara
14
2.120ms
2ms
< 4.256ms 0.800ms 0.400ms
2.400ms
9.976ms
IEEE802.15.4e - A Slot
4/4/2012
8
SensorsCon 2012 March 21, Santa Clara
15
16 c
hannel offsets
e.g. 33 time slots (330ms)
A
B C
D E
F G
H
I
J
…and energy consumption
IEEE802.15.4e - Trade-Off
• Cells are assigned according to application requirements
• Tunable trade-off between – packets/second
SensorsCon 2012 March 21, Santa Clara
16
16 c
hannel offsets
e.g. 33 time slots (330ms)
A
B C
D E
F G
H
I
J
• Cells are assigned according to application requirements
• Tunable trade-off between – packets/second
– latency …and energy consumption
IEEE802.15.4e - Trade-Off
4/4/2012
9
SensorsCon 2012 March 21, Santa Clara
17
16 c
hannel offsets
e.g. 33 time slots (330ms)
A
B C
D E
F G
H
I
J
• Cells are assigned according to application requirements
• Tunable trade-off between – packets/second
– latency
– robustness
…and energy consumption
IEEE802.15.4e - Trade-Off
SensorsCon 2012 March 21, Santa Clara
18
2.120ms
2ms
< 4.256ms 0.800ms 0.400ms
2.400ms
9.976ms
Type of slot Transmitter Receiver
“OFF” - -
transmission w. ACK 6.856ms 7.656ms
Transmission w.o. ACK 4.256ms 5.256ms
Listening w.o. reception - 2.000ms
IEEE802.15.4e - Energy Consumption
“atomic” operations
4/4/2012
10
SensorsCon 2012 March 21, Santa Clara
19
Dust Networks DN6000 – “Mote-on-chip”
TX @ +8dBm 9.7 mA
TX @ 0dBm 5.4 mA
RX 4.5 mA
Cortex-M3 @ 7.4MHz
2.4 mA
Doze mode 1.2 uA
ARM Cortex-M3 32-bit uC
IEEE802.15.4 radio
SensorsCon 2012 March 21, Santa Clara
20
Internet
Wireless
mesh
IEEE802.15.4e - Energy Consumption
gateway
Example 1: “leaf” node
Activity:
- Data publication
- 80B data payload
- Published every 30s
- 80% stability paths to parents
Consumption:
- 11uA average
- 22 years lifetime (2200mAh AA)
4/4/2012
11
SensorsCon 2012 March 21, Santa Clara
21
Internet
Wireless
mesh
IEEE802.15.4e - Energy Consumption
gateway
Example 2: router node, with one child
Activity:
- Data publication
- Provisioning for routing
- Advertisement sent every 2sec
- Listening for join requests
- Forwarding data from child
Consumption:
- 31uA average
- 8 years lifetime (2200mAh AA)
SensorsCon 2012 March 21, Santa Clara
22
IEEE802.15.4e - Reliability
Doherty, Lindsay, Simon. “Channel-Specific Wireless Sensor Network Path Data”, ICCCN 2007.
44 nodes, 26 days
Printing Facility
>99.999% end-to-end reliability
4/4/2012
12
SensorsCon 2012 March 21, Santa Clara
23
Why Standardize?
The Internet
SensorsCon 2012 March 21, Santa Clara
24
Interoperation - IPSO
“IP for Smart Objects”
Alliance of companies
Promote the use of IP in Smart Objects
Not a standardization body
White Papers, interop events…
www.ipso-alliance.org
4/4/2012
13
SensorsCon 2012 March 21, Santa Clara
25
Interoperation - IEEE802.15.4e
SensorsCon 2012 March 21, Santa Clara
26
Standards-Based Reliable Wireless Sensor Networking
Thomas Watteyne, PhD Senior Networking Design Engineer Linear Technology/Dust Networks
4/4/2012
14
SensorsCon 2012 March 21, Santa Clara
27
Abstract
The Internet of Things revolution is quietly coming, and with it an epochal turnpoint in wireless network design. Major standardization bodies have been looking at how wireless multi-hop networks should operate reliably (WirelessHART, IEEE 802.15.4E, IETF RPL) and how they can integrate within the Internet (IETF 6LoWPAN, CoAP).
This talk will highlight the challenges faced by wireless multi-hop networks, before showing how communication protocols and standards can address these. Numerous use cases, examples and lessons learnt will be taken from open-source and commercial implementations.
SensorsCon 2012 March 21, Santa Clara
28
Speaker’s Biography
Thomas Watteyne is a Senior Networking Design Engineer at Dust Networks, a company specializing in ultra-low power and highly reliable Wireless Sensor Networking. He designs networking solutions based on a variety of M2M standards and promotes the use of highly reliable standards such as IEEE802.15.4e. In 2009 and 2010, he was a postdoctoral researcher at the University of California, Berkeley, working with Prof. Kristofer Pister. He created Berkeley’s OpenWSN project, an open-source initiative to promote the use of fully standards-based protocol stacks in M2M applications. He obtained his PhD in Computer Science (2008) and MSc in Telecommunications (2005) from INSA Lyon, France.