The Next Tier of The Internet

1
The Next Tier of The Internet

• David E. Culler
• dculler_at_archrock.com
• culler_at_cs.berkeley.edu

10-3-06
BroadNets Keynote
2
The Internet Today
3
The Internet Tomorrow
4
The Internet The World
5
Physical Information Streams

• Sensors are everywhere
• but the data is mostly dropped on the floor
• Physical gt Digital gt Information
• Each sensor becomes a network citizen

6
What are Wireless Sensor Networks?

• Network of tiny footprint computers
• Optimized for long life on low power
• Equipped to sense physical data
• Networked using low-power radio
• Function
• Sense any measurable parameter
• Light, motion, chemicals, proximity, biometrics
• Form network and communicate
• Automatic meshing and routing over the air
• Apply user-defined business logic
• Sampling, summarizing, reporting events
• Form
• Mote (Processor, Radio, Storage) Sensors
• Embedded Operating System and Networking
• Gateways towards Enterprise IT systems

Internet or Enterprise
Gateway
Gateway
Wireless Sensor Nodes
7
Why are they so Important?
Increase Comfort
High-Confidence Transport
psst. were hiring?
8
The Next Tier of the Internet ?

• More than protocols - IP/TCP/HTTP
• Open Building blocks for physical information
• platforms, systems, networks
• Open Process for Networking Excellence
• Reference implementations leading to standardized
  protocols and interfaces
• Design philosophy
• Expect Change, Failure, Uncertainty, Unexpected
• Enable broad applications
• Web services

9
Emergence of Open Platforms
Physical World
10
Emergence of Open Systems
Digital World
WSN mote platform
Sensors
Wireless
Physical World
11
Snapshot of an International Community

• 346,000 downloads from tinyos.net
• plus sourceforge, crossbow, moteiv,
  www.tinyos.or.kr,
• Open TinyOS Alliance

http//www.tinyos.net
12
Wireless Embedded Systems
TinyOS 2.0
WSN mote platform
Communication Centric Resource-Constrained Event-d
riven Execution
13
Wireless Embedded Networks
Applications and Services
Over-the-air Programming
Blocks, Logs, Files
Streaming drivers
Scheduling, Management
Digital World
Link
ADC, Sensor I/F
Physical World
14
Multi-Hop Mesh Routing
0
15
Embedded Networking Requirements

• Reliable Dissemination
• Data Collection and Aggregation
• Point-to-point Transfers
• Reliably over lossy links
• At low power
• Idle listening, management, monitoring
• Adapting to changing conditions
• Scalar and Bulk Versions

16
Confluence of Technical Developments
Embedded Networking
Tiny Operating System
WSN
CMOS Radios
Microcontroller SOCs
Integrated Sensors
17
Applications gt Network Characteristics

• Monitoring Spaces
• Env. Monitoring, Conservation biology, ...
• Precision agriculture,
• built environment comfort efficiency ...
• alarms, security, surveillance, EPA, OSHA, treaty
  verification

18
Applications gt Network Characteristics

• Monitoring Spaces
• Env. Monitoring, Conservation biology, ...
• Precision agriculture,
• built environment comfort efficiency ...
• alarms, security, surveillance, EPA, OSHA, treaty
  verification
• Monitoring Things
• condition-based maintenance
• automated meter reading
• Energy efficiency, cold chain
• disaster management
• Civil infrastructure

19
Intel Fab BP Machine Monitoring

• Goal Pre-empt equipment failures through
  non-destructive analysis
• Media Gap Majority of data is collected by hand
• Thousands of sense points
• Intel Fab and an Oil Tanker engine room
• Wireless vibration data collection
• High-speed sampling, reliable bulk transfer
• Sensor-to-Analysis App flow
• Overcome interference
• Support disconnected operation
• Loch Rannoch Network
• 150 accelerometers
• 26 motes
• 4 stargates
• 1 PC
• Efficient installation and management
• 36hr install period on tanker
• No crew intervention
• BP going to product in 06

20
Applications gt Network Characteristics

• Monitoring Spaces
• Env. Monitoring, Conservation biology, ...
• Precision agriculture,
• built environment comfort efficiency ...
• alarms, security, surveillance, EPA, OSHA, treaty
  verification
• Monitoring Things
• condition-based maintenance
• automated meter reading
• disaster management
• Civil infrastructure
• Interactions of Space and Things
• manufacturing, asset tracking, fleet franchise
• context aware computing, non-verbal communication
• Assistance - home/elder care
• Action and control
• Optimizing processes
• Automation

21
Proximity, Tracking, Compliance
Industrial Asset mgmt CBM Process control Energy
mgmt environmental
Medical Agriculture HLS
Home Lighting HVAC access
Shipping QoS Vibration threshold Temp/humidity
threshold

• Current Conditions
• In Motion, No Magnetic Fields, 67
• Humidity Range 30-35
• Temp. Range 55 - 78

22
Transportation and Mobility

• Smart Pallets/Containers equipped with WSN motes
• Shipping Warehouse
• Commission smart pallet A (PA)
• Set shipping information
• EPC Temperature, humidity, light exposure
  thresholds
• Conflicting material types
• Ship to Receiver Warehouse
• Offline Scenario
• Continuous monitoring, log exceptions to flash
  memory
• Receiver Warehouse
• Auto-detect Shipment arrival, exception reporting
• If move PA to conflicting material PB, sound
  alarm and report exception

Shipper
Receiver
PB
PA
23
Canonical SensorNet Network Architecture
Patch Network
Sensor Node
Gateway
BroadNets, hmm?
24
How will SensorNets and IP play together?
XML / RPC / REST / SOAP / OSGI
?
HTTP / FTP / SNMP
TCP / UDP
IP
802.15.4, CC,
802.11
Ethernet
Sonet
25
Beware IP hype

• Pushing HTML over a wireless connection to a
  serial port attached to a PC is NOT running IP on
  the sensorNet
• Nor is packet-scraping into XML-RPC

26
1st Internet Principle IP hour glass
Application
XML / RPC / REST / SOAP / OSGI
Session
HTTP / FTP / SNMP
Transport
TCP / UDP
Network routing
IP
Link
Ethernet
Sonet
802.11

Physical
Coax
10BT
Fiber
Serial
DSS
QPSK
27
2nd Internet Principle End-to-End
Architecture
28
Full IP stack throughout
XML / RPC / REST / SOAP / OSGI
HTTP / FTP / SNMP
TCP / UDP
IP
802.15.4, CC,
802.11
Ethernet
Sonet
29
Sensor Net Challenges for traditional IP

• Large headers ltgt small packets
• Very different protocol assumptions
• Low-power
• Intermittent connectivity (mostly sleeping)
• Loss ?gt Congestion
• Very different traffic patterns
• Not universal point-to-point file transfer
• Aggregate collection, dissemination
• Naming
• The data is important, not the machine
• Mobility

30
Enterprise-Scale Sensor Nets
31
The Internet End-to-End no more
Network Address Translation Firewall
Load Leveler Switch
32
Edge Network Approach
XML / RPC / REST / SOAP / OSGI
HTTP / FTP / SNMP
TCP / UDP
IP
802.15.4, CC,
802.11
Ethernet
Sonet
33
Hybrid approach

• Security
• No IP to the nodes, attacks have to get through
  the gateway or be physically close
• Namespace management
• Name nodes, networks, services
• Hosts, URLs,
• Mask intermittent connectivity
• Terminate IP on the powered side
• Loosely couple, energy aware protocols on the
  other
• Distillation proxies
• Small binary packets where constrained
• Expanded to full text, XML, HTML, web services
• Rich suite of networking techniques in the Patch
  unimpeded by overall internet-scale process

34
SensorNets need the Wisdom of the Internet
Architecture

• Design for change!
• Network protocols must work over a wide variety
  of links
• Links will evolve
• Network protocols must work for a variety of
  applications
• Applications will evolve
• Provide only simple primitives
• Dont confuse the networking standard with a
  programming methodology
• Dont try to lock-in your advantage in the spec
• Open process
• Rough consensus AND running code

35
Networking vs Distributed Applications

• IP, 802.15.4, Zigbee are still plumbing
• Enterprise information infrastructure means
  Integration with Applications
• The data, how it is represented, what it means
• The operations you can do on it
• How it is obtained from the physical world
• Actions based on it

36
Database View of the World

• The World is a Database
• SQL, ODBC/JDBC
• The good old days of client-server
• Data warehouses and distributed databases
• Network taken for granted

37
Web Services

• The ultimate interoperability
• unify within and between enterprises
• Service Oriented Architecture
• Widely adopted standards
• Backed by OASIS and W3C
• Open-source reference implementations, e.g.,
  Apache
• SOAP WSDL ( UDDI)
• Or XML-RPC over HTTP (REST)

38
Web Service

• Platform and implementation independent software
  component that can be
• Described using a service description language
• Published to a registry of services
• Discovered through a standard mechanism
• Invoked through a declared API
• Usually over a network
• Composed with other services

Building Web Services with Java, Graham et al,
SAMS 2002
39
Service Oriented Architecture
Service Registry
Service Provider
Service Requestor

• Service Description gt interface implementation
• Operations supported, input/output objects
• Bindings to network and data encoding schemes
• Network address where service can be invoked
• Enough that client can generate code to access
  the service well

40
Embedded Application Tier

• Embedded applications built on a rich set of node
  services.
• Timing, sensor streams, storage
• Local processing
• Reliable, low-power communication
• Platform independent extensions

client
tier1
server
tier2
tier3
SensorNet GW/Proxy
tier4
Networking Protocols
mote
Management
Common Link Abstraction
TinyOS Runtime Services
Physical World
Hardware Abstraction Layer
MCU
Radio
Sensors
41
WSNs and Web Services

• Decorate external interface points
• Attributes (shared data)
• RPCs (control points)
• Events (signals)

client
tier1
server
tier2
Embedded Application
Networking Protocols
Management
Common Link Abstraction
TinyOS Runtime Services
Hardware Abstraction Layer
MCU
Radio
Sensors
42
WSNs and Web Services

• Auto-generate Web Services
• Service description
• Service implementation

client
tier1
server
tier2
Embedded Web Service
Embedded Application
Networking Protocols
Management
Common Link Abstraction
TinyOS Runtime Services
Hardware Abstraction Layer
MCU
Radio
Sensors
43
WSNs and Web Services

• Auto-generate Web Services
• Service description
• Service implementation

client
tier1
server
tier2
Embedded Web Service
44
A new WSN world
client
tier1
server
AquaLogic
tier2
Perl
Python
Embedded Services
C
NetWeaver
Excel
45
Sensor Web Services
46
Pick your favorite client
47
Real World Webs
48
Real-World Wireless Web Today?

• LANDSAT gt Google Earth
• vast, 50-year satellite infrastructure
• massive servers
• Cellular gt M2M
• vast, 30-year tower network infrastructure
• Wireless client
• GPS
• 25-year satellite infrastructure
• Wireless 1-function client
• RFID
• extensive, intra-facility infrastructure
• Wired powered readers, wireless tags
• WiFi
• Relatively light AP infrastructure
• Rapidly advancing
• Powered server, wireless client
• Wireless Sensor Networks

49
Wireless Client vs Wireless Server

• Wireless Client
• needs to last a day or two
• has a human to keep them working properly
• mostly formats specific incoming data for display
• Wireless server
• needs to last for long periods
• must be self-managing, adaptive, robust
• Generates meaningful data for many uses
• Often mobile (!!!)

50
Whats in a battery?

• 2 AA gt 1.5 amp hours (4 watt hours)
• Cell gt 1 amp hour (3.5 watt hours)
• Cell 500-1000 mW gt few hours active
• WiFi 300-500 mW gt several hours
• GPS 50 100 mW gt couple days
• WSN 50 mW active, 20 uW passive
• 450 uW gt one year, 45 uW gt 10 years

51
The Next Tier
Clients
Servers
Internet
52
Arch Rock
Did I say were hiring?

• Connecting the physical world to the information
  enterprise
• Solid Foundation
• High-quality, Cross-Platform embedded application
  infrastructure
• Standard-based
• OS, Networking, Services
• Reliable, efficient, extensible
• Novel information in familiar forms
• IP-accessible web services
• Auto-generated from embedded application

internet
tier3
tier4
53
Uniting long-lost relatives
Thanks
Instrumentation Computers
General Purpose Computing
54
Arch Rock Technology

• Infrastructure that integrates physical
  information with business logic
• Applications on Real-World Web Services
• Java, .net, php,
• IP-based Service Oriented Architecture
• App, mgmt, conversion, persistence, data
• WiFi, GPRS
• Embedded Application Services
• Extensible attributes events
• Embedded Network Services
• IEEE 802.15.4-based
• Cross-platform industrial-quality WSN operating
  system
• TinyOS 2.0
• Broad platform and sensor choice

Client
tier1
IT Enterprise
Server
tier2
Physical World