1. By:Vidushi Chaudhary (MT-11016)
The Low Power Energy Aware Processing (LEAP) Embedded Networked Sensor System
By:Vidushi Chaudhary (MT-11016)
Tarun Bansal (MT-11014)

2. AIM OF THE PAPER
LEAP:- To meet the requirements like minimizing high peak power dissipation, on- demand high performance computing and high bandwidth communication.
A new distributed node testbed demonstrating that by exploiting high energy efficiency components and enabling proper on-demand scheduling, LEAP architecture may meet both sensing performance and energy dissipation objectives.
Previous microcontroller based applications does not matched to high power dissipation sensors that must be scheduled for on-demand use.

3. LEAP design Course platforms
LEAP Approach:-
Diverse sensor system.
Fine grained platform instrumentation.
Highly configurable testbeds.
1. Lots of sensors 2. to make the instrumentation fine, where to put the platform

4. Some Notations ENS:- Embedded networked sensor.
EMAP: - Energy Management and Accounting Preprocessor.
MSP:- Multi client service provider.

5. Introduction
Prior development of ENS platforms has resulted in low power systems well matched to the requirements for supporting low power sensor devices.
The computing demands for such systems were matched to low data rate and low complexity sensors.
As performance need for new application have increased, it is critical to minimize system energy dissipation.
Example from net

6. Contd…
So to address these diverse applications, we need to optimize sampling to minimize the energy.

7. Environmental monitoring
Eco system monitoring
Health monitoring
Large Sensor
Power dissipation
Performance
Communication
Complex information processing
So balance of above need and energy dissipation should be maintained for fine work.
Solution is
Combine hardware and software changes(By new architecture).
Study environmental phenomenon.
Environmental monitoring:-temp
Eco system Monitoring:-
Health monitoring:-

8. New Achitecture: New architecture must :
Focus on minimizing energy required for each sensing, computing and communication task.
LEAP is the solution in that response.
LEAP is based on hardware and software partitioning specifically adopted to new requirements.

9. LEAP architecture LEAP processor LEAP preprocessor
High efficiency and high power components
( Used on Demand) are assigned to a LEAP processor partition
Continuously vigilant micropower components are assigned to a LEAP preprocessor partition.
EMAP provides fine grained monitoring…

10. EMAP
Provides fine grained monitoring and control of energy dissipation in all ENS subsystems.
Scheduling of power delivery.
Provides event detection and triggering capability.
Separate power domains supporting individual components.
It creates a range of power modes.
Jis sensor ko jo power mode chahiye usi me work krega…

11. Design Requirements Study environmental phenomenon.
Categorize sampling at low rate or event triggered fashion.
It covers a large domain of environmental monitoring applications with proper scheduling, sensing, computing and communication.
Different environmental phenomenon's like imaging system, aquatic systems have different sampling rates and event trigger system, which may be frequently or infrequently used according to time domains. So our ENS shud be adjustable to the sampling rate, so that according to time, it shud change the sampling rate scheduling and processing.

12. Design Approach: Computing platform
Activity
Interested Events
Lack of Events
Low
High
Communication
Computation
Energy
Time Constraints
Communication, computation , energy low in lack of events while in case of interested events , they became high.
In case of lack of events, when events are inactive, duty cycle is low.
We have to choose PLATFORM, WHICH CAN FIT IN BOTH CATEGORIES WITH HIGH EFFICIENCY AND MINIMIZE THE ENERGY.

13. Design approach: Platform Selection
Benchmark characterization .
1. CRC:- Test efficiency for executing Error detection and correction algorithm tasks.
2. Finite Impulse Response(FIR) Requirement for ENS for digital filtering of sensor data stream.
3. Fast Fourier Transform Test efficiency for Data transformation on sampled data.
Platform comparison
Operational measurements on each processor using C code.
Current is measured.
Benchmark execution start and completion by toggeling I/O pins.
FFT- analog to digital and vice versa

14. After applying LEAP design Above table shows that selection of high performance processor option result in reduced energy usage. It confirms the idea of LEAP design where a processor is used on demand for execution of specific task, otherwise it will be in low power mode.

15. Design approach: Communication
Selection of wireless interface
Energy efficient.
Internode communication.
Integration with existing deployed infrastructure.
LEAP uses dual radio approach so system have capability of both low power operation and high energy bulk data transfer.
802.11g is 9 times more efficient than how????????

16. Design approach: Sensing Systems
Sensor Interfaces:
RS-232,
I2C,
SPI serial,
High speed USB and Ethernet.
LEAP provide data access to PXA processor platform along with sensor interfaces, and then finally to high speed interface including USB and Ethernet.
Reduce processing overhead.
Eliminate bus protocol controller or extraneous data copies.

17. Design Approach: Storage
Local storage is a critical design requirement.
Through addition of storage, high energy cost communication episodes may be scheduled to occur at times optimal for data transport.
LEAP design allow application developers to select data allocation strategy of various memory types and to directly measure resulting energy and performance.

18. Energy Management Challenges: LEAP provides:
ENS systems also have temporal dependence for a given application.
Allocation of shared resources(including computing, communication, storage).
Fine- grained device level monitoring and control must be in system.
LEAP provides:
Partitioning devices into power domains.
Monitor each domain.
Respond to events or conditions.
LEAP includes Remote access processing
Time dependent kbhi zayada chahiye kbhi kam
Jo event ho rhe h unke bases pr power management bnayega and EMAP ko bhej dega.
EMAP partitions power domains and send account info (logging) to host processor
Power management schedule
Host processor
EMAP
Account Information

19. LEAP Hardware Architecture
PXA Mhz processor SDcard:- for external memory PCMCIA:Personal Computer Memory Card International Association

20. Slauson Processor Module
SPM is based on the Sensoria Slauson Platform.
Contains PXA Mhz processor, SDRAM and an Intel K3 Strata flash of up to 128MB and 64 MB respectively.
SPM has dual PCMCIA interfaces configurable for either 3.3V or 5V devices.
Each of the PCMCIA slots may be independently isolated and powered down.
Communication is provided by on- board 10baseT ethernet controller.
The SPM includes an extensive set of interfaces via a 180pin inter- board header including serial buses such as two RS232 ports, two SPI ports, I2C, and AC97 audio sampling ports.
Memory space of full parallel memory bus is 192MB.
PCMCIA?? *6
10/100 requires more ideal power operate at high clock rate. Therefore 10baseT is used.

21. EMAP module
The EMAP allows the LEAP system to be subdivided into five power domains and each domain is independently powered and isolated.
Power is supplied to each domain through a low- resistance current sensing resistor.

22. EMAP: Energy management and sensor interfaces
EMAP power domains are allocated to
EMAP module
SPM
Up to 3 external sensor systems.
SPM communicate to EMAP for recent charge accumulation values. EMAP responds with each domain’s value. Than host processor accurately compute integrated charge and energy for each power domain.
EMAP provides power management scheduling capability.
Each power domains are electrically isolated to avoid current leakage path when power off.

23. EMAP: processor communication
Support I2C, SPI, UART
I2C – Chosen for Interboard communication, multimaster capability( multiple EMAP modules and multiple high performance processors).
SPI are configured as master for access to dual wireless radio CC2420.
EMAP: Low power operation

24. EMAP: low power operation
For purpose of energy and performance control, MSP CPU frequency controlled from 100KHz to 8 MHz.
MSP enter into LPM3(current sensing task and IDEAL tasks), when running operating system in ideal thread.
LPM4 state(suspend state) is achieved through software, The MSP processor and all EMAP peripherals are disabled.

25. EMAP: Remote debugging
MSP processor JTAG interface has been provided to SPM’s PXA processor through inter board connection.
SPM controls MSP processor’s execution, program internal flash, control MSP’s I/O pins.
Host processor acts as proxy agent.
Debugging commands given by remote debugger are routed to processor and convert to JTAG command sequence.

26. LEAP Software Architecture
Software Tier
System Boot loader
Flash memory manipulation
Remote file retrieval
Loading and execution of other OS.
Can be updated or completely replaced over Remote Links.
First Tier
Second Tier
Linux operating system
Modules for device drivers
Network protocols and power management.
Third Tier
At boot time, validate integrity of Read/ write file
system(located on JFFS2)(Journaling Flash File System version 2).
Upon validation, transfer control to fourth tier which
start init program.

27. Contd…
4th Tier JFFS2 has standard Linux directory structure, Boot scripts, Large file system utility, glibc library
4th Tier include SPM to EMAP comm. System
SPM access EMAP utility Functions like for interaction with EMAP via I2C comm.

28. MSP client- Utility
msp-client give a good interface with interactive model for development and testing.
Access to sensor data, energy data, power control for all domains, controlled interface to set, modify power management schedule per domain.

29. EMAP MSP controller
Software architecture includes a set of uC/OS (RTOS) objects that are compatible with small MSP memory.
Main software task that has to be controlled
Host processor communication using I2C messages.
Power management scheduling.
Sensor interface monitoring.
Threshold triggering
Radio communication.

30. Blocking nonblocking DMA control Messaging task
I2C perform blocking non blocking operations used DMA controlled, utilize CPU during blocking task. Responsible for messaging task(with prpcessor and MSP client) , master slave read/write
Blocking nonblocking
DMA control
Messaging task
Host processor communication using I2C messages.
Power management scheduling.
Sensor interface monitoring.
Threshold triggering
Radio communication.

31. Host processor communication using I2C messages.
I2C perform blocking non blocking operations used DMA controlled, utilize CPU during blocking task. Responsible for messaging task(with prpcessor and MSP client) , master slave read/write
Host processor communication using I2C messages.
Power management scheduling.
Sensor interface monitoring.
Threshold triggering
Radio communication.

32. Host processor communication using I2C messages.
I2C perform blocking non blocking operations used DMA controlled, utilize CPU during blocking task. Responsible for messaging task(with prpcessor and MSP client) , master slave read/write
Host processor communication using I2C messages.
Power management scheduling.
Sensor interface monitoring.
Threshold triggering
Radio communication.

33. Host processor communication using I2C messages.
I2C perform blocking non blocking operations used DMA controlled, utilize CPU during blocking task. Responsible for messaging task(with prpcessor and MSP client) , master slave read/write
Host processor communication using I2C messages.
Power management scheduling.
Sensor interface monitoring.
Threshold triggering
Radio communication.

34. Experimental Results LEAP demonstrate different algorithms:
Reactive Algorithm: Respond to external event captured by sensor.
Proactive Algorithm: Estimate arrival of event in advance.
Hybrid Algorithm: it is combination of Reactive algorithm and proactive algorithm.

35. Contd…
Reactive algorithm suited where sensors and sensor platforms are highly agile in time and energy usage or Sensed phenomenon are poorly understood.
Proactive approaches may suit less agile sensors and sensor platforms or where the sensed phenomenon are well understood.

36. Testbeds and Algorithm
LEAP Testbeds:- deploying many LEAP based nodes in a distributed network, each supporting multiple sensor inputs for environmental event detection.

37. LEAP Testbed
A new testbed has been developed that provides accurately reproducible physical events that may be detected both by the micropower, as well as by the on- demand use of high performance imaging devices supported by each LEAP based node.

38. Experimental Results: Event fenerator
Consist two horizontal linear arrays of 32 individually controlled lamps distributed over an 8 m length.
Both Red and Green lamps are attached to the rigid assembly at fixed intervals .
Power for each lamp is sequenced by an independent relay control.

39. Algorithm design and implementataion
Iski theory dekh liyo likhni ho to

40. Conclusion
Achieve desired performance goals while simultaneously meeting energy dissipation requirements
Design approach focuses on exploiting high energy efficiency components that are scheduled for operation on demand operation.