Dreams for Profoundly Implanted Systems.


53 views
Uploaded on:
Category: Funny / Jokes
Description
Current chip fabricating innovation will soon achieve its ... Nano-scale Computing. DNA control can sort out cells into decisively designed ...
Transcripts
Slide 1

Dreams for Deeply Embedded Networks Challenges for the following decade

Slide 2

The Trends in Computing Technology 1970s 1990s tomorrow

Slide 3

Future Computing Applications Smart paint Smart homes and universal registering Wearable figuring Ingestible gadget organizes Computationally-increased situations Common vision: Massively dispersed systems of minor preparing components

Slide 4

Important Projects MIT – Oxygen, Amorphous processing Berkeley – Smart Dust UCLA, Xerox – Sensor Networks AT&T, … – Smart situations

Slide 5

Rethinking the Fundamentals Hardware Miniaturization Software New programming Abstractions Communication New Protocol Stacks Theory Scalable new Behavior

Slide 6

A Future Hardware Technology Towards nano-scale gadgets

Slide 7

The Hardware Challenge Miniature equipment gadgets must be produced financially in vast numbers Current microchip fabricating innovation will soon achieve its lithographic size points of confinement What are conceivable option future advancements?

Slide 8

Cellular Computing Cells as rationale doors Basic inverter: Concentration of protein Z is conversely corresponding to grouping of protein A. NAND entryway: Production of protein Z is hindered by nearness of proteins An and B.

Slide 9

Tools and Challenges in Cellular Computing Tools: BioSpice: A test system and verifier of hereditary computerized circuits Plasmid compiler: Converts a rationale outline into a hereditary execution Challenges: Understanding of protein/repressor sets Biological obstructions between circuits Mismatch of natural response energy

Slide 10

Nano-scale Computing DNA control can sort out cells into decisively built examples This innovation could be the establishment for development of complex sub-nano-scale additional cell circuits: Biological framework – machine shop Proteins – machine apparatuses DNA – control tapes Circuits are created in expansive numbers by shoddy organic procedures

Slide 11

Smart Dust Current innovation: 5mm bits Goal: 1mm

Slide 12

Macro Motes Stable innovation: inch-sized bits

Slide 13

Future Programming Paradigms Distributed inserted registering

Slide 14

Software Challenges in Sensor Networks Support for substantial quantities of unattended gadget Sensor nets versus Web? Versatile conduct in a capricious situation Sensor nets versus mechanized assembling? Information driven correspondence

Slide 15

Programming Massively Distributed Systems Individual gadgets are not imperative Program must endure gadget disappointments and anomaly Program does not know definite gadget areas Program must give the craved general conduct

Slide 16

Growing Point Language reflections: Growing focuses Pheromones Tropisms Sequential conceptualization of a parallel calculation Any planar outline can be ordered into a GPL program

Slide 17

Programming in a Physical Environment In a sensor-rich environment, the human and machine can have the same physical model of the world Physical items can have a product representation Physical article area can be a piece of programming state New applications? New interfaces?

Slide 18

Active Bat System A keen office framework actualized by AT@T "Bats" are connected to followed questions and individuals Bats transmit ultrasonic signs A variety of sensors in the building finds the bats by going Physical elements in the earth are spoken to as programming CORBA objects

Slide 19

Applications Spatial screen Browsing the physical environment Follow-me frameworks Personalization of shared assets (e.g., advanced cameras) Novel client interfaces Augmented reality A computerized sensory system

Slide 20

Rethinking the Protocol Stack Towards new correspondence standards

Slide 21

The Communication Protocol Stack Internet Sensor Net Transport (TCP, UDP) Transport (Area-to-Area) Network (IP) Network (Diffusion; ID-less hubs) Link (Ethernet,… ) Link (Low Power Wireless)

Slide 22

Low Power Communication Radio correspondence Consumes an excess of force Requires an extensive reception apparatus Optical correspondence Base-band correspondence No tweak, dynamic separating, demodulation Beam can be gone for destination

Slide 23

Passive Optical Communication Base station points pillar at a variety of bits Motes reflect bar utilizing a corner-3D square retro-reflector Motes adjust reflected shaft force (a few kilobits for every second)

Slide 24

Challenges Line of sight necessity Aiming prerequisites Matching the field of perspective Link directionality Bit-rate, separation, vitality tradeoffs (for given SNR) Multi-jump optical steering Mobility

Slide 25

Directed Diffusion Data-driven and area driven tending to Sinks express enthusiasm for a few information quality Interest is spread (diffused) to coveted area Data coordinating this credit is converse engendered to sinks Passage of information revives angles

Slide 26

Towards a New Theory of Computing Local calculations, scale and rising conduct

Slide 27

Aggregate Behavior Theory Current circulated calculations ordinarily depict associations of a limited number of effective machines Such disseminated calculations have adaptability issues How to create calculation models for a "boundless" number of straightforward gadgets? Will we create calculations perform better with expanded scale?

Slide 28

Analogy In numerous physical and natural frameworks boisterous nearby segment communications produce vigorous total conduct by uprightness of scale Emergence of mass properties of matter from neighborhood nuclear associations Formation of complex natural living beings from neighborhood cell connections Phase moves coming about because of nearby sub-atomic interactions

Slide 29

Main Challenge How can one specialist pre-indicated conduct from collaboration of huge quantities of untrustworthy parts connected in obscure unpredictable ways? New ways to deal with adaptation to internal failure and total conduct How to plan neighborhood associations to create a total conduct of decision?

Slide 30

Scalable Coordination and Self-Organization Algorithms are expected to self-sort out huge quantities of hubs Clustering calculations Team arrangement Approximate accord Triangulation Routing and correspondence

Recommended
View more...