Routing for Sensor Networks - PowerPoint PPT Presentation

Loading...

PPT – Routing for Sensor Networks PowerPoint presentation | free to download - id: 68329d-YjZhM



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

Routing for Sensor Networks

Description:

Routing for Low Power and Lossy Networks (RL2N) BOF IETF-70 - Vancouver - December 2007 BOF Chairs: JP Vasseur/David Culler ADs: Dave Ward/Ross Callon – PowerPoint PPT presentation

Number of Views:40
Avg rating:3.0/5.0
Date added: 11 June 2020
Slides: 41
Provided by: JPVas4
Learn more at: http://www.ietf.org
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Routing for Sensor Networks


1
Routing for Low Power and Lossy Networks
(RL2N) BOF IETF-70 - Vancouver - December
2007 BOF Chairs JP Vasseur/David
Culler ADs Dave Ward/Ross Callon
2
Agenda
  • Administrativia (Chairs, 5 min)
  • Notes takers
  • Agenda bashing
  • Scoping the BOF (Chairs/ADs, 10 min)
  • Motivation and problem statement presentation
  • RL2N Routing Requirements summary (several
    chairs, 15 mn)
  • Protocol survey (Levis 5 mn)
  • Consensus and Charter discussion (15 mn)
  • Consensus
  • Work Items/Milestones
  • Interaction with other WG (6lowpan)
  • Conclusion and next steps (10mn, chairs and ADs)

3
Why are we here?
  • This is a Working Group forming BOF
  • Desired outcome is to form a new working group
  • This is not a technology tutorial or opportunity
    to discuss solutions
  • Presentations will demonstrate
  • Existence of a problem that needs to be solved
  • Approach to understand and solve problem
  • Existence of large community of interest
  • Listing of proposed charter deliverables and
    timeline
  • Please hold discussion to end of presentations

4
Agenda
  • Administrativia (Chairs, 5 min)
  • Notes takers
  • Agenda bashing
  • Scoping the BOF (Chairs/ADs, 10 min)
  • Motivation and problem statement presentation
  • RL2N Routing Requirements summary (several
    chairs, 15 mn)
  • Protocol survey (Levis 5 mn)
  • Consensus and Charter discussion (15 mn)
  • Consensus
  • Work Items/Milestones
  • Interaction with other WG (6lowpan)
  • Conclusion and next steps (10mn, chairs and ADs)

5
BoF scoping Background
  • Wireless embedded networks are widely deployed
    today
  • Industrial monitoring, process control, Automated
    Metering, Condition Based Maintenance, Building
    HVAC, Lighting control, home automation, Cold
    Chain, energy usage, agriculture, urban
    infrastructure,
  • Well-understood usage models with body of
    practical experience
  • Enabled by CMOS radios over past decade
  • Numerous link types IEEE 802.15.4, Bluetooth,
    Low Power WiFi, WiBree
  • many gateways, proxies and middleware adapters
  • Low Cost, Low Power, Low Bandwidth
  • Several industrial forums created to fill the
    standards gap
  • Zigbee, ZWave, Wireless HART, ISA SP100.11a,
    SP100,
  • plus many proprietary protocols, even over
    standard links
  • Large industry desire to move to interoperable
    devices running over IP
  • IETF has not actively been working on low power
    networks until recently
  • 6lowpan WG RFC4944 (IPv6 over IEEE 802.15.4)
  • Still no IP Routing Solution for such networks.

6
BOF scoping Domain
  • Routing over Low Power and Lossy Networks L2Ns
  • Networks comprising a large number of highly
    constrained devices interconnected by wireless
    links of unpredictable quality.
  • Web page www.employees.org/jvasseur
  • Technology and problem space discussed _at_ Chicago
    (IETF-69) Routing Area Meeting
  • Low Power ? Low Transmission Power, Modest
    Receive Sensitivity, Short Range, Multi-Hop,
  • Lossy ? BER, Small MTU, Embedded in changing,
    often harsh, environment.
  • Typically connected to small foot-print hosts
    (microcontrollers)
  • Slides can be found at http//www.employees.org/
    jvasseur/RL2N-Routing-Area-meeting-IETf-69.ppt

7
BOF scoping Domain.2
  • Productized, working implementations have been
    developed
  • Industrial routing solutions over lossy links at
    low power available today
  • Each defines its format, network, transport,
    gateway,
  • IETF 6LoWPAN Internet Area WG produced IPv6
    header compression RFC over IEEE 802.15.4 making
    IP practical for this class of networks.
  • IP Routing solution for L2Ns is needed
  • Consensus with 6lowpan and other WGs that an IP
    routing solution should be developed in the RTG
    area.
  • LoWPANs are multi-hops and interconnected using
    various link types e.g., IEEE 802.15.4, LP
    802.11, WiBree, but also some wired links.

8
BoF Scoping Problem Statement
  • Define routing requirements for a representative
    subset of the application scenarios that utilize
    L2Ns in large scale environments
  • Formulate a routing framework for these scenarios
    that provides
  • high reliability in the presence of time varying
    loss characteristics
  • connectivity while permitting low-power operation
    with very modest memory and CPU pressure
  • Paying particular attention to routing security
    and manageability (self configuration)

9
BoF Scoping Approach
  • Produce Routing Requirements documents for select
    representative use cases in collaboration with
    INT 6lowpan WG
  • Industrial application networks
  • Connected Home, Building
  • Urban infrastructure networks
  • Survey applicability of existing protocols to
    L2Ns.
  • Ability to carry new link and node attributes,
    scaling characteristics and overhead
  • Existing IGPs, MANET, NEMO, DTN
  • Provide architectural framework for routing in
    L2Ns
  • Routing metrics used in path calculation that
    include static and dynamic link/nodes attributes
  • Distributed vs. centralized path computation
  • Hierarchy
  • gtRecharter with suggestion of protocol
    direction and on to specification

10
Agenda
  • Administrativia (Chairs, 5 min)
  • Notes takers
  • Agenda bashing
  • Scoping the BOF (Chairs/ADs, 10 min)
  • Motivation and problem statement presentation
  • RL2N Routing Requirements summary (several
    chairs, 15 mn)
  • Protocol survey (Levis 5 mn)
  • Consensus and Charter discussion (15 mn)
  • Consensus
  • Work Items/Milestones
  • Interaction with other WG (6lowpan)
  • Conclusion and next steps (10mn, chairs and ADs)

11
Home Automation Routing Requirement in Low Power
and Lossy Networksdraft-brandt-rl2n-home-routing
-reqs-02
(very high level overview)
  • Anders Brandt (Zensys) - Anders Brandt
    abr_at_zen-sys.com
  • IETF 70, Vancouver, Dec. 2007

12
Devices in the home
  • Remote control
  • Only wakes up when operated...
  • Movement sensor Smoke alarm
  • Routing should be avoided to save battery gt Node
    attributes taken into account by constrained
    routing
  • Lamp module Switch module
  • Stable routing resources

13
Home scenario routing issues ..
  • Short distances but Multi-hop routing needed
    because of signal distortion
  • Reinforced concrete, refrigerator doors and
    other metal objects
  • Support for multiple paths nodes may fail or be
    powered off
  • Ability to locate a working path within
    250ms if operational and used before
  • Neighbor discovery on a frequent
    basis Consumers move nodes at will ...
  • Routing self-configuration is a MUST.

14
Industrial Routing Requirements in Low Power and
Lossy Networks draft-pister-rl2n-indus-routing-r
eqs-00
(very high level overview)
  • Kris Pister (Dust Networks) - Kris Pister
    kpister_at_dustnetworks.com
  • IETF 70, Vancouver, Dec. 2007

15
Industrial Automation Background
  • Very important functionality
  • 60 million installed process control sensors
  • 4 million shipping per year
  • 50 are smart today wired networks
  • HART
  • Most popular wired sensor network protocol
  • HART 1 1,200 baud digital comm over 4-20mA loops
  • Wireless HART
  • Ratified as a part of HART7 September 2007
  • 802.15.4 based
  • Announced vendors ABB, Emerson, Siemens,
  • Multi-hop Mesh networking
  • SP100 wireless
  • Draft standard in 2008
  • Adopted 6LoWPAN, but defining own routing,
    transport
  • Wireless HART and SP100 are a hybrid of circuit
    and packet switched

16
Examples of Data flows
  • Low frequency data collection
  • 1/s to 1/hour typically lt 1/min
  • Latency comparable to sample interval
  • Typically lt50B
  • Some time series gt10kB
  • Alarms
  • lt50B
  • Log file upload
  • 1/day, 1/year
  • 10kB ..1MB
  • Human diagnostic query/response
  • Mean latency important
  • Feedback control
  • Max latency important
  • Latency from minutes to lt1ms (infeasible w/ 15.4
    radios)
  • Often all of these will be operating in different
    parts of the network

17
Industrial ScenarioRouting Requirements
  • Link Attribute aware routing
  • The usual Latency, BW, reliability,
  • New constraints node lifetime ? traffic ?
    routing
  • Average power energy storage (batteries,
    ultra-caps, )
  • Peak power energy scavenging (solar, vibration,
    current loops, )
  • High scalability is a MUST
  • Security
  • Extremely important - risk of lives and mega-
  • Part of WG charter produce a security framework
    for routing in L2Ns

18
Routing Requirements in Urban Low Power and Lossy
Networksdraft-dohler-rl2n-urban-routing-reqs-01
(very high level overview)
  • Hassnaa Moustafa, Mischa Dohler, G. Madhusudan,
    G. Chegaray, T. Watteyne, C. Jacquenet
  • France Telecom

19
Urban L2Ns Background
  • Trial roll-outs in Voiron and other French cities
  • Serious business projections in near future
  • Network elements
  • sensors (sensing of environment)
  • actuators (control of environment)
  • repeaters (infrastructure, coverage extension,
    etc)
  • access points (gateway, information sink and
    source)
  • Peculiarities of Urban L2Ns
  • huge amount of field nodes
  • highly energy constrained nodes
  • batteries (ir)regularly recharged
  • correlated data readings
  • highly directed info flow

20
Urban RL2Ns Requirements
  • Critical importance
  • scalability (mainly w.r.t. path discovery,
    energy)
  • parameter constrained routing (mainly w.r.t.
    energy)
  • security (mainly w.r.t. data integrity and
    authenticity)
  • alien and autonomous functioning/configuration
  • Lesser importance
  • bandwidth
  • latency
  • Must support
  • highly directed information flows
  • heterogeneous field-devices (different MACs)
  • multi/group/geo cast

21
Applying Routing Protocols in L2Ns
draft-levis-rl2n-overview-protocols-02(very
high level overview)
  • Phil Levis (Stanford) - pal_at_cs.stanford.edu
  • JP Vasseur (Cisco) - jpv_at_cisco.com
  • David Culler (arch Rock) - dculler_at_archrock.com

22
Goal
  • Large spectrum of existing routing protocols
  • L2Ns introduce new and specific constraints
  • Can we use existing protocols?
  • If not, what can we learn? What can we borrow?
  • RL2N charter work item

Survey the applicability of exiting protocols to
L2Ns. The aim of this document will be to analyze
the scaling and characteristics of existing
protocols and identify whether or not they meet
the routing requirements of the L2N applications
identified above. Existing IGP, MANET, NEMO, and
DTN routing protocols will be part of the
evaluation.
23
Requirements(derived from applications IDs)
  • Small footprint
  • Flooding control and density awareness
  • Multi-path routing
  • Resource awareness (link/nodes)
  • Small MTU
  • Multi-topology routing

24
Requirements(derived from application IDs)
  • Small footprint
  • Flooding control and density awareness
  • Multi-path routing
  • Resource awareness (link/nodes)
  • Small MTU
  • Multi-topology routing

25
Illustrative preliminary study
  • Draft examines all requirements
  • Cover two here control traffic and route state

Protocol Medium Class Traffic State
OSPF wired link-state O(N2dc) O(Nd)
OLSR wireless link-state O(N2) O(Nd)
TBRPF wireless link-state O(N2) O(Nd2)
RIP wired vector O(ND) O(D)
AODV wireless vector O(NCc) O(Cd)
DSDV wireless vector O(ND) O(D)
DYMO wireless vector O(NCch) O(Ch d)
DYMO-low wireless vector O(NCc) O(C d)
DSR wireless vector O(NDh) O(Dh)
N Number of nodes C Communicating nodes D
Destinations d density h hopcount c link churn
26
Agenda
  • Administrativia (Chairs, 5 min)
  • Notes takers
  • Agenda bashing
  • Scoping the BOF (Chairs/ADs, 10 min)
  • Motivation and problem statement presentation
  • RL2N Routing Requirements summary (several
    chairs, 15 mn)
  • Protocol survey (Levis 5 mn)
  • Consensus and Charter discussion (15 mn)
  • Consensus
  • Work Items/Milestones
  • Interaction with other WG (6lowpan)
  • Conclusion and next steps (10mn, chairs and ADs)

27
Consensus reached so far
  • Informal work started about a year ago
  • Creation of a non-WG mailing list after IETF 68
    (rsn_at_ietf.org - 400 subscribers)
  • Generic requirements ID draft-culler-rl2n-routing-
    reqs to focus discussion
  • Application driven routing requirements approach
  • Industrial draft-pister-rl2n-indus-routing-reqs
  • Connected Home draft-brandt-rl2n-home-routing-req
    s
  • Urban draft-dohler-rl2n-urban-routing-reqs

28
BOF scoping - Summary of discussions
  • Pre-WG meeting was held in Boston on November 15
  • Participants K. Pister (Dust), R. King
    (Crossbow), A. Brandt (Zensys), M. A. Mc Lachlan
    (BT), P. Levis (Stanford), J. Oliveira (Drexel),
    J. Butler (Cimetrics), D. Ward (IESG), M. Dohler
    (FT) M. Murphy-hoye (Intel), Ed Butler (Intel),
    Pete St Pierre (Sun), David Culler (Arch Roc), G.
    Mulligan (6lowpan co-chair), Pierre Colle
    (Schneider Electric), Todd Snide (Schneider
    Electric), Nabil Bitar (Verizon), JP Vasseur
    (Cisco).
  • Agenda
  • Minutes posted to the RSN Mailing list

Introduction and Background JP/David -
15mn Review of the current IDs (15mn per ID)
- draft-brandt-rl2n-home-routing-reqs-01-
Anders - draft-pister-rl2n-industrial-routin
g-reqs-01- Kris - draft-levis-rl2n-overview-
protocols-01 - Phil Charter Discussion of
the proposed charter Interaction with other
WGs 6lowpan and SDOs Open issues DTN,
MANET BOF organization
29
Large Community of Interest is apparent
  • Collaboration of Industry, End User, Vendors,
    Service Providers, Academia (Schneider Electric,
    France Telecom, Dust, Arch Rock, Zensys, Cisco,
    Intel, Stanford, Berkeley, Drexel, ).
  • Active support on the list from many
    participants Telecom Italia, Sensinode,
    Crossbow, Cimetrics, Silver Spring,
  • In addition other collaborating WGs have helped
    scope charter to solve the problem and work
    through any apparent overlaps and remove any
    conflict

30
Strong support from community
20 positive feed-backs on the list, no negative
comment.
  • Many positive feed-backs supporting the formation
    of a new WG
  • .. We see a strong interest in the creation of
    this working group about routing over lossy low
    power network Sharing common approaches would
    also ease implementation of interoperable
    solutions. Schneider Electric
  • I would like to express my support for the
    R2LN effort. At present Zigbee seems to be the
    most popular open technology for wireless
    field-level networks, but I believe that IP-based
    wireless solutions that meet the requirements of
    HVAC control and lighting control would be
    welcomed by the building controls industry. Jim
    Butler - Cimetrics - CTO
  • Telecom Italia welcomes a solution of the
    issues listed in the scope of this initiative.
    Telecom Italia
  • After some FT/Orange internal discussions and
    having had you on the phone today, we have
    decided to actively participate in RL2N. France
    Telecom
  • Crossbow Technology wants to express its
    support for this working group as we believe that
    IP based communication will be an important
    standard for the markets we serve. Ralph Kling -
    Crossbow
  • There is a huge need for standardized IP WSN
    routing in well-scoped domains now, today, to
    solve real problems But the biggest concern
    they still have is a lack of a standard routing
    technique. So rl2n is solving an acute problem
    for deploying IP-based wireless sensor networks.
    I think the charter text is good, and scoping to
    these 3 areas really keeps it focused. Zach Zelby
    Sensinode

31
RL2N WG Charter Overview Work Items
  • Produce use cases documents for Industrial,
    Connected Home, Building and urban application
    networks.
  • Describe the use case and the associated routing
    protocol requirements.
  • The documents will progress in collaboration with
    the 6lowpan Working Group (INT area). ?
  • Survey the applicability of existing protocols to
    L2Ns analyze the scaling and characteristics of
    existing protocols and identify whether or not
    they meet the routing requirements of the L2Ns
    applications.
  • Existing IGPs, MANET, NEMO, DTN routing protocols
    will be part of evaluation.

32
RL2N WG Charter Overview Work Items (2)
  • 3. Specification of routing metrics used in
    path calculation.
  • This includes static and dynamic link/nodes
    attributes required for routing in L2Ns.
  • 4. Provide an architectural framework for
    routing and path selection at Layer 3 (Routing
    for L2N Architecture)?
  • Decide whether the L2Ns routing protocol require
    a distributed, centralized path computation
    models or both.?
  • Decide whether the L2N routing protocol requires
    a hierarchical routing approach.
  • 5. Produce a security framework for routing in
    L2Ns.

33
Goals and Milestones
  • April 2008 Submit Use case/Routing requirements
    for Industrial, Connected Home, Building and
    Urban networks applications to the IESG to be
    considered as an Informational RFC.
  • August 2008 Submit Routing Metrics and
    Attrributes for L2Ns document to the IESG to be
    considered as an Informational RFC.
  • November 2008 Submit Protocol Survey to the IESG
    to be considered as an Informational RFC.?
  • January 2009 Submit Security Framework for L2Ns
    to the IESG to be considered as an Informational
    RFC?
  • February 2009 Submit the Routing for L2Ns
    Architecture document (summary of requirements,
    metrics and attributes, path selection model) to
    the IESG as an Informational RFC.
  • March 2009 Recharter.

34
Interaction with other WGs
  • 6lowpan working on L2Ns over 802.15.4
  • MANET we may be end up using some (adapted)
    MANET protocols if the WG think that they satisfy
    the requirements
  • Other industry forums and SDOs.
  • Zigbee,
  • ITU,
  • Bluetooth,
  • Wosa,
  • What is out of scope?
  • Use cases not listed in the charter
    agricultural, healthcare, wild life, (aka DTN)

35
Agenda
  • Administrativia (Chairs, 5 min)
  • Notes takers
  • Agenda bashing
  • Scoping the BOF (Chairs/ADs, 10 min)
  • Motivation and problem statement presentation
  • RL2N Routing Requirements summary (several
    chairs, 15 mn)
  • Protocol survey (Levis 5 mn)
  • Consensus and Charter discussion (15 mn)
  • Consensus
  • Work Items/Milestones
  • Interaction with other WG (6lowpan)
  • Conclusion and next steps (10mn, chairs and ADs)

36
Discuss now
  • Is there consensus to form a Working Group?

37
Back-up Slides
38
RL2N Routing Requirements summary
  • 10-point General Requirements served as framework
    for building consensus, prioritization, and more
    detailed used-case requirements analysis
  • Spatially-Driven Multihop
  • Light Footprint
  • Small MTU
  • Deep power management
  • Heterogeneous Capabilities (node and link)
  • Highly Variable Connectivity
  • Structured Workload and Traffic Pattern
  • Operation with Partial Information
  • Quality of Service Capable Routing
  • Data Aware Routing

39
Good support on the list
We see a strong interest in the creation of this
working group about routing over lossy low
power networks. Schneider Electric is already
proposing wireless solutions for both Home,
Building and Industrial markets. We are very
interested to provide in this group our
application requirements. Providing a standard IP
based wireless routing solution would be useful
for our customers already using an IP
infrastructure as it simplifies gateway
implementation. We would also like to create
links between ZigBee and this working group. We
think that a lot of work done by ZigBee might be
re-used. Sharing common approaches would also
ease implementation of interoperable solutions.
Schneider Electric
This e-mail just to state that Telecom Italia
welcomes a solution of theissues listed in the
scope of this initiative. In particular
TelecomItalia would like to see a smooth and
fruitful integration of the IPlayer with existing
layer 2 protocols to enrich the IP
devices'ecosystem once a suitable routing
protocol has been developed.Telecom Italia
welcomes a solution which is well harmonized
withexisting standards for Wireless Sensor
Networks, in particular ZigBee.The solution
should also envisage for a network comprising
sleepingrouters, i.e. nodes having routing
facility and comprising a radiotransceiver that
is intermittently operating (typically
batterypowered).
Telecom Italia
I would like to express my support for the R2LN
effort. In the commercial building controls
industry, wireless communication is a very hot
topic. Major manufacturers are actively
marketing control products and systems that
support wireless communication At present
Zigbee seems to be the most popular open
technology for wireless field-level networks, but
I believe that IP-based wireless solutions that
meet the requirements of HVAC control and
lighting control would be welcomed by the
building controls industry.
Jim Butler - Cimetrics - CTO
40
Good support on the list
Geoff Mulligan - 6lowpan co-chair
In my opinion the proposed RL2N charter is
already pretty good, here is some support from
our experience. There is a huge need for
standardized IP WSN routing in well-scoped
domains now, today, to solve real problems.
For example here in Finland we are already
rolling out fairly large 6lowpan (over 1000 node)
networks in the industrial and building
automation domains. Our direct 6lowpan node
volume, will be well over 25k nodes in 2008.
ISA100 helps next year, but for many applications
a simpler 6lowpan rl2n routing solution would
be more suitable without full ISA100. Of course
I assume ISA100 would apply the routing solution
from rl2n eventually? Almost all customers love
the IP 802.15.4 concept. But the biggest
concern they still have is a lack of a standard
routing technique. So rl2n is solving an acute
problem for deploying IP-based wireless sensor
networks. I think the charter text is good, and
scoping to these 3 areas really keeps it focused.
Zach Shelby - CTO - Sensinode
Dear JP, dear all, Thanks for your efforts in
getting all this together. After some FT/Orange
internal discussions and having had you on the
phone today, we have decided to actively
participate in RL2N.
Mischa Dohler France Teleocm
I think that the ROLL WG is critically necessary
to look at the issues surrounding routing over
precisely the type of networks that 6lowpan
anticipates using/building. I'll be at the BOF
on Thursday and I support us moving forward with
this working group.
Geoff Mulligan - 6lowpan co-chair
thanks again for driving the R2LN effort.
Crossbow Technology wants to express its support
for this working group as we believe that IP
based communication will be an important standard
for the markets we serve.
Ralph Kling - Chief Architect Crossbow
Please register my support for forming a working
group, and my support for the proposed charter.
Charles Perkins - Nokia
J. De Oliveira
I would also like to reiterate my support for the
R2LN/ROLL effort.
About PowerShow.com