1. Old Dominion University
Towards Autonomous Vehicular Clouds
M. Eltoweissy
S. Olariu
M. Younis
AdHocNets 2010, Victoria, BC, August 2010 1 9/16/2010
2. Old Dominion University
Outline
• What is VANET?
• What is cloud computing?
• Why vehicular clouds?
• Potential applications
• Research challenges
• A call to action
AdHocNets 2010, Victoria, BC, August 2010 2 9/16/2010
3. Old Dominion University
VANET – the dream
F. Doetzer, Privacy Issues in Vehicular Ad Hoc Networks
AdHocNets 2010, Victoria, BC, August 2010
3
4. Old Dominion University
VANET – the killer app
• Huge potential societal impact
• An overnight success with
automotive industry
various government agencies (USDOT, FCC, etc)
standardization bodies: ASTM, IEEE, SAE, ISO
• Emergence of projects and initiatives
Car-2-Car communication consortium
Vehicle Safety Consortium
Vehicle Infrastructure Integration
Networks on Wheels
AdHocNets 2010, Victoria, BC, August 2010
4
5. Old Dominion University
Smart vehicles (1)
• Vehicles are becoming more sophisticated
powerful on-board computing capabilities
tons of on-board of storage
significant communication capabilities
no power limitations
• Computations capabilities supported by
hosts of sensors and actuators
on-board radar and GPS
AdHocNets 2010, Victoria, BC, August 2010 5 9/16/2010
7. Old Dominion University
Cloud computing
• A paradigm shift suggested by
low cost high-speed Internet
virtualization
advances in parallel and distributed databases
• Basic idea: why buy when you can rent
exactly what you need
exactly when you need it
• Appealing to startups and other players
no upfront investment
no maintenance costs
AdHocNets 2010, Victoria, BC, August 2010 7 9/16/2010
8. Old Dominion University
Cloud services
• Software as a Service (SaaS)
customers rent software hosted by the vendor
• Platform as a Service (PaaS)
customers rent infrastructure and programming tools hosted by the
vendor to create their own applications
• Infrastructure as a Service (IaaS)
customers rent processing, storage, networking and other
fundamental computing resources for all purposes
AdHocNets 2010, Victoria, BC, August 2010 8 9/16/2010
9. Old Dominion University
Vehicular clouds ‒ motivation
• Consider the parking lot of a typical company on a typical
workday
• hundreds/thousands cars go
unused for hours on end
• Why rent computational/storage
resources elsewhere?
• you have them in your own backyard
• they are yours to waste!
AdHocNets 2010, Victoria, BC, August 2010 9 9/16/2010
10. Old Dominion University
AVC – autonomous vehicular cloud
• Vehicles are ideal candidates for nodes in clouds of various
resources
• Autonomous Vehicular Cloud (AVC)
a group of autonomous vehicles whose corporate computing,
sensing, communication and physical resources can be
coordinated and dynamically allocated to authorized users
• How are AVCs different?
mobility
agility
autonomy
AdHocNets 2010, Victoria, BC, August 2010 10 9/16/2010
11. Old Dominion University
How are AVCs different?
• Mobility: the presence of vehicles in close proximity to an event
is often un-planned
pooling of the resources in support of mitigating the event must occur
spontaneously
• Agility: refers to the ability of AVCs to tailor the amount of
shared resources to the actual needs of the situation in support
of which the AVC was constituted
agility does not exist in conventional clouds and is an important defining
characteristic of AVCs
• Autonomy: refers to the decision of each vehicle to participate
in the AVC
AdHocNets 2010, Victoria, BC, August 2010 11 9/16/2010
12. Old Dominion University
A TLSP application
• Traffic Light Synchronization Project (TLSP) a major US-DOT
initiative
• Consider a city block where a traffic-related event has occurred
• Once the traffic event has been cleared, relying on the existing
scheduling of the traffic lights will not help dissipate the traffic
backlog in an efficient way
• Our solution
the vehicles themselves will pool their computational resources together
creating the effect of a powerful super-computer
the AVC will recommend to a higher authority a way of rescheduling the
traffic lights that will serve the purpose of de-congesting the afflicted
area as fast as possible
AdHocNets 2010, Victoria, BC, August 2010 12 9/16/2010
13. Old Dominion University
AVC for planned evacuations (1)
• In cases of predicted disasters, massive planned evacuations are
often necessary in order to minimize the impact of the predicted
disaster on human lives
• Once an evacuation is underway, finding available gasoline,
drinking water, shelter and medical facilities quickly becomes an
issue
• The 2005 hurricane evacuations in New Orleans and Houston
have confirmed that there is no room for mistakes or
misjudgments here
AdHocNets 2010, Victoria, BC, August 2010 13 9/16/2010
14. Old Dominion University
AVC for planned evacuations (2)
• How can an AVC help?
• We anticipate that the vehicles
involved in the evacuation will
self-organize into one or several
inter-operating AVC that will
work hand in hand with the
emergency management center
• In the course of this interaction,
the emergency managers can
upload information about open
shelters to the central server
AdHocNets 2010, Victoria, BC, August 2010 14 9/16/2010
15. Old Dominion University
How is an AVC set up?
• The formation of an AVC involves the following:
a broker elected spontaneously will attempt to form an AVC
the broker will secure a preliminary authorization from a higher (city) forum
the broker will inform the vehicles in the area of the received authorization and will
invite participation in the AVC.
the cars will/or will not respond to the invitation on a purely autonomous basis
the broker decides if a sufficient number of vehicles have volunteered and will then
announce the formation of the AVC
the cars in the AVC will pool their computational resources to form a powerful
supercomputer that, using a digital map of the area, will produce a proposal
schedule to the higher (city) forum for approval and implementation
once the proposal has been accepted and implemented, the AVC is dissolved
AdHocNets 2010, Victoria, BC, August 2010 15 9/16/2010
16. Old Dominion University
AVC research challenges
• To make AVCs reality research is needed along the following
three engineering dimensions
architectural
functional
operational and policy
AdHocNets 2010, Victoria, BC, August 2010
17. Old Dominion University
AVC architectural challenges
• Elastic mobile architecture
AVC networking and protocol architecture must be developed to accommodate
changing application demands and resource availability on the move
• Resilient architecture
AVC basic structural and composed building blocks must be designed and
engineered to withstand structural stresses induced by the inherent instability in
the operating environment.
• Service-oriented network architecture
contemporary layered network architectures, (e.g. the TCP/IP stack) have
proven inadequate in face of evolving applications and technologies
we envision the adoption of service-oriented component-based network
architectures with intrinsic monitoring and learning capabilities
AdHocNets 2010, Victoria, BC, August 2010
18. Old Dominion University
AVC functional challenges
• Enabling AVC autonomy
research is needed on developing a trustworthy base, negotiation and strategy,
efficient communication protocols, data processing and decision support
systems, etc.
• Managing highly dynamic cloud membership
there is a critical need to efficiently manage mobility, resource heterogeneity,
trust, and vehicle membership
• Cyber-Physical control
AVCs can be defined by their aggregated cyber and physical resources. Their
aggregation, coordination and control are non-trivial research issues.
• Cooperation between AVCs
AdHocNets 2010, Victoria, BC, August 2010
19. Old Dominion University
AVC policy challenges
• Trust and trust assurance
research is needed on developing a trustworthy base, negotiation and strategy
formulation methodology, efficient communication protocols, data processing
• Contract-driven versus ad hoc AVC
we anticipate that the bulk of AVCs will be contract-driven, where the owner of
the vehicle or fleet consents to renting out some form of excess computational
or storage capacity
in addition to the contract-based form of AVC, there should be possible to form
a AVC in an ad hoc manner as necessitated by dynamically changing situations
• Effective operational policies
In order for the AVCs to operate and inter-operate seamlessly, issues related to
authority establishment and management, decision support and control
structure, the establishment of accountability metrics, assessment and
intervention strategies, rules and regulations, standardization, etc.
AdHocNets 2010, Victoria, BC, August 2010
20. Old Dominion University
A call to action
• We expect cloud computing to see a phenomenal adoption
rate and penetration of the IT market
• As cloud computing takes root it will be emulated by other
areas
• It is only a matter of time before it will be extended to
• vehicular assets from individual vehicles to entire fleets
• cell phones and other commodity consumer products
• Are we ready for this paradigm shift?
AdHocNets 2010, Victoria, BC, August 2010