1. Networking of Sensorial Embedded Systems (SES)
Guru Parulkar
Department of Computer Science and Engineering University of California, Riverside and National Science Foundation

2. Executive Summary Sensorial Embedded Systems (SES) show great promise
Will bring another IT revolution that is even bigger
Networking of SES offer tremendous opportunities
Research, technology development, education and overall impact
UCR is well positioned to lead
Expertise and track record of the faculty
Commitment of the University leadership
Funding opportunities: federal and industrial
I want to share essentially three ideas with you.
The first idea is that networking of sensorial embedded systems will bring about another bigger IT revolution.
The second idea is that this revolution offers tremendous opportunities for exciting research, technology development, education and most importantly technology transfer and impact on the society.
The third idea is that UCR is well positioned to lead this revolution and that is the mission of this center.
Finally we strongly believe that board’s participation in terms of help and advice will lead to a grand success for all parties, that is local industry and community, student population, and of course the University.
Please allow me to elaborate on each one of these ideas over the next 15 minutes or so.
28 November 2018

3. Disruptive Technologies: SES
Digital sensors and actuators
Very inexpensive and can be integrated into silicon
Wireless
Low power inexpensive RF
Silicon integration
Sensor, DSP, CPU, FPGA, wireless, actuators
Huge software on a chip
Ultimate vision of silicon integration: Smart Dust
What is an embedded system?
“Intelligence” built into physical systems
Intelligence in the form of computing
Embedded systems have been around for a long time and getting better. Examples
Weapons
Automobiles
Appliances
Scientific instruments
Toys
And more
So what is the big deal?
There are three technologies that are changing the basic rules in this space and they would bring about the revolution. The three technologies are
Digital sensors and actuators
Low power inexpensive wireless transceiver that can be also integrated into a chip.
Finally of course the silicon integration that we are all familiar with. It allows doubling silicon capability every 18 months.
The one big difference in this case is that people want to put not only processing but also sensors, DSPs, programmable logic, and actuators on the same chip.
So imagine you have a chip (10mm X 10mm) that can sense the environment around it; compute something useful; communicate its results and/or act on them by actuating something. And it costs only few cents.
UC Berkeley Motes, Crossbow Sensors
28 November 2018

4. Applications in All Aspects of Life
Intel Presentation
28 November 2018

5. Network of SES: Central Nervous System of Physical Web
Seamless Physical Web
Scientific
Precision agriculture
Structural analysis
Habitat and environment monitoring and response
Water contamination detection
Ocean monitoring
Defense
Target recognition and tracking
Asset tracking and deployment
Homeland security and protection of critical infrastructure
Industrial, Enterprise, IT
Tracking of all kinds of things
Health monitoring and alert
Instrumented buildings
Disaster recovery
Predictive equipment maintenance
Environmentally aware data center
Residential
Instrumented homes
security, lighting, entertainment
Tracking of people (kids) & goods
Network of SES: Central Nervous System of Physical Web
28 November 2018

6. Challenges and Opportunities
New Machines
New Environments Applications
New Scale
Billion to trillion devices!
28 November 2018

7. Network Technology Gap
Monitoring & Managing Spaces and Things
applications
Networking of SES
Net Prog Env
protocols
system
architecture
Management
Store
Comm.
uRobots
actuate
MEMS
sensing
Proc
Power
technology
Miniature, low-power connections to the physical world
28 November 2018

8. SES and Networking SES are not useful unless they are networked
“Network is the computer” is more true for SES
Networking of SES different from networking that we know
Solutions must lead to programmable, manageable, durable, secured networks of SES
28 November 2018

9. Realities of Network of SES Environment
Difficult interfaces to physical world
Highly resource constrained
Extreme operational environments
Highly heterogeneous
Unusually long life time
Unprecedented scale
Highly critical security and privacy
28 November 2018

10. Challenges System design Networking of SES Privacy and security
Networked programming
28 November 2018

11. System Design Challenge
More challenging trade-offs space
Energy vs duty cycle
Failure rates and redundancy
Computing vs communications
RF power vs range
Time
Space
Need new design and verification models, methodologies and tools for hardware and software
28 November 2018

12. Potential Hardware Platforms
Intel Stargate
BT Node from ETH Zurich
And there are more
Berkeley Motes
28 November 2018

13. Networking of SES Challenge
Given
New types of devices with differing capabilities
A variety of wireless and some wired links
Different connectivity & failure modes
Applications with very different requirements
Need to invent new network and protocol architectures, algorithms, and implementations
Devices: New type of devices with different capabilities
E.g. dust, motes, Imotes, stargate
Battery, solar, line power
Concept of life time and duty cycle
A variety of sensors
Links: A variety of wired and wireless links
Link reliability and bandwidth limited
Adaptive range and bandwidth of links
Networks: Very different connectivity
Dense connectivity
High failure rates but with redundancy of nodes
Need to invent new network and protocol architectures, algorithms, and implementations
28 November 2018

14. Network Architecture Need for a tiered architecture
Functions and capabilities of each tier
Trade computing and communications to suite the environment/app
An example architecture
First tier
Motes ==> smart dust
Battery powered
Numerous
Interfaced to physical world
Second tier
More powerful and bigger
Interfaced to numerous first tier SES
Battery, other sources of power, possibly line power
Functions: Coordination of first tier nodes, data aggregation, computing-in-network, data integrity, etc
Third tier
Gateway to the Internet
Sharing among multiple applications and user classes
Heterogeneous platforms in each tier
Focus of research on network and protocol architectures that explore and validate different partitioning of functions
28 November 2018

15. Protocol Architecture
Addressing, naming
What needs to have an address/name?
Name and address binding/resolution
Topology discovery and localization
Routing: data driven, different metrics, secured
Congestion control: different notion of QoS
Time synchronization
Protocol architecture that facilitates
Multi layer abstractions (divide and conquer) AND
Applications aware optimization and integrated layer processing
Security, privacy, and integrity as its fundamental capabilities
Power Aware Numerous SES Tiered Network Arch
28 November 2018

16. QoS and Capacity Network of SES capacity
Trade computing and communications, especially with in-network computing
QoS specification and guarantees
Application specific QoS semantics
Latency estimates and measurements
Real time or synchronous communication
Flexible channel isolation
28 November 2018

17. Network of SES Management
Self aware and evolvable
Dynamic topology
Identification and isolation of failed SES
Measurement
While conserving precious resources
28 November 2018

18. Security and Privacy Challenge
Very different and challenging
Physical environment cannot be protected
Traditional firewall & key distribution approaches do not work
A compromised SES or application can easily disrupt, carefully alter, or snoop operation
Damage would be very serious
Need new thinking and solutions
Implications on hardware, OS, networked programming tools
Haven’t even started
28 November 2018

19. Network Programming Challenge
Need to program a group of potentially 1000s of SES with a set of unique constraints
Existing models and tools
Presume mostly client-server model
Designed for a small number of nodes
Not designed to address other unique constraints of SES
Need new programming models and tools
Increase ease of programming and reusability
New abstractions and automatic mapping on an array of SES
Assure privacy and security
28 November 2018

20. In-network Computing Arch
How to program sensornets?
Stream processing
Distributed objects
Control loop
Declarative “SQL”
Distributed Runtime Services
Data aggregation & correlation, clustering, and adaptation
Mapping app to resources
Network
Path selection / routing Bandwidth mgmt Channel access
SensorOS
SensorOS
SensorOS
Security, app isolation
Tamper-resistance Location-aware sensors
Hardware
Hardware
Hardware
From
28 November 2018

21. UCR Project: Tight Coupling of Network, Operating System & Application Layers
Applications Layer
Type 2 Node: Mobile Robot C B A D 1 2 3
sensor
events
Type 1 Node:
Sensor
on-line
data analysis
QoS to
distributed
tasks
Distributed on-line analysis of sensor data
Deviation detection in asynchronous streams
Layer 2: Distributed
Computing Layer
Distribution of application services
Dynamic real-time scheduling of sensor events
Layer 1:
Network Layer
Anycasting and synchronization of fusion operations
Trajectory control of the mobile robots
Our Goal: How to build self-organizing and self-managing dynamic sensor network systems
(From Vana)
28 November 2018

22. Approach
Foundations Research
Experimental Systems
Reusable Systems & Science
Deployed Infrastructure Applications
Commercial Solutions
Close Coupling Must
Programmable Manageable Durable Secured Networks of SES to Enable Seamless Physical Webs
28 November 2018

23. Expected Results
Foundations research: theoretical, algorithms and systems
Systems
Several NSES platforms: range of capabilities for tiered architecture
OS and protocols stacks
Networked programming environments
Many local testbeds with applications around the country
Education and Training
New graduate and undergraduate courses with experimental projects
Many graduate students with hands-on experience with platforms, implementation of algorithms and applications
28 November 2018

24. UCR Core Expertise Strong foundation in place Long track record in
Embedded, networking, wireless, DSP, & software sys
Several research projects in progress
Strong collaborations in place
And willingness to build more
28 November 2018

25. UCR Target Applications
Lot of opportunities
to build interesting applications
Biosensors and Bioremediation Group
28 November 2018

26. Executive Summary Sensorial Embedded Systems (SES) show great promise
Will bring another IT revolution that is even bigger
Networking of SES offer tremendous opportunities
Research, technology development, education and overall impact
UCR is well positioned to lead
Expertise and track record of the faculty
Commitment of the University leadership
Funding opportunities: federal and industrial
I want to share essentially three ideas with you.
The first idea is that networking of sensorial embedded systems will bring about another bigger IT revolution.
The second idea is that this revolution offers tremendous opportunities for exciting research, technology development, education and most importantly technology transfer and impact on the society.
The third idea is that UCR is well positioned to lead this revolution and that is the mission of this center.
Finally we strongly believe that board’s participation in terms of help and advice will lead to a grand success for all parties, that is local industry and community, student population, and of course the University.
Please allow me to elaborate on each one of these ideas over the next 15 minutes or so.
28 November 2018