Context-awareness, cloudlets and the case
for AP-embedded, anonymous computing
Anthony LaMarca
Associate Director
Intel Labs Seattle
Context Awareness
•
Is driving the next generation of applications and services
– Health & assisted living, advertising, gaming, travel, social networking,
task assistance and education, etc…
•
Is currently driven by low data rate sensors
– Accelerometers, RFID and radio base station IDs,
thermocouples, barometers, and capacitive sensors
2
The Future of Context Awareness
•
SOA is not accurate or robust enough for many applications
– Location: 5-100 meters error, no pose or direction info
– Activity: 20-40% error rates unless drastically limited in scope
•
Next gen context aware solutions
– High data rate sensors (Cameras and microphones)
– Compute intensive (real time classification & online learning)
– Interactive
•
3
Puts huge pressure on mobile devices in terms
of compute capacity, communication, and power budget
Supporting Mobile Context Awareness
The Case for VM-based Cloudlets in Mobile Computing
Satyanarayanan, Bahl, Caceres, Davies
• The cloud isn’t the solution
– Too much latency for real-time responsiveness
(Internet2 mean RTT 50-250 ms)
• Bring the cloud closer
– Cloudlet: “data center in a box”
• One network hop from the client
• Shared with other nearby clients
4
Put the cloudlet into the Access Point
802.11n AP
with a n-core
CPU
Low latency,
high bandwidth
5
–
Lots of cores with no power constraints one
network hop away
–
Forms a natural rendezvous point between the
cloud and the client
• Trusted by both parties and lies within
trusted boundary of the client (AP in the
home or coffee shop)
–
Leverage existing 802.11 protocols for service
discovery, encryption, billing and
authentication with no extra overhead
–
Provides tight coupling between the network
and the computation
–
Incremental deployment
• Today APs are added as business, home
usage grow
• Cloudlet capabilities could be incrementally
added with cloudlet-APs
Challenges
•
Providing mechanisms to guarantee execution integrity
– Privacy: Client should trust that no sensitive data is retained
– Correctness of execution: Clients and service provider should trust the
correctness of the computation
– Approach: Leverage trusted hardware primitives of Dynamic Root of Trust
for Measurement (DRTM) and attestation
•
Developing applications that span the client, AP and cloud
– Seamless migration of execution between the client, AP and the cloud
•
Supporting flexible business models
– Give users access to proprietary algorithms in exchange for context
– Provide micropayments to cloudlet owners in exchange for cycles
– Broker connections between advertisers and customers based on conetxt
6