VIEWS: 3 PAGES: 30 POSTED ON: 12/30/2011
Engineering at Aberdeen Communications and Imaging Research Group 7 academic staff 15 research staff and students •Electronic Engineering •Communications Engineering •Optical Engineering •Parallel and Image Processing Communications and Imaging Research Group: WARMER personnel Directly involved: Dr Alastair Allen Prof Tim Spracklen Dr Oliver Faust Mr Bernhard Sputh Mr Golam Murshed + other research students Other staff related to Wireless Sensor Networks: Prof Anne Glover (Medical Sciences) Dr A Manivannan (Biomedical Physics) Dr Norval Strachan (Physics) + research staff/students Experience & expertise Embedded systems Communications Wireless Sensor Networks Embedded systems Development of optimized algorithms and high performance computer architectures for embedding imaging and other computational intelligence into small devices Processing of Using •Signals •Multiprocessor systems •Images •Digital Signal Processors •Sensor data •Microcontrollers •Field Programmable Gate Arrays •Artificial Neural Network hardware Embedded systems • Concurrency – Parallel compiler technology – Formal methods/tools for secure and provable systems: CSP/FDR – Java and concurrency • Operating Systems and Device Drivers • Artificial Neural Network hardware Field Programmable Gate Arrays Reconfigurable systems •Algorithmic optimisation •Multi-core (eg. MicroBlaze) Software Defined Radio •Partial reconfigurability •Formal methods for hardware / firmware / software interfacing Communications Professor Spracklen has been the UK representative on the UN Comprehensive Nuclear Test Ban Treaty Organisation (CTBTO) since 1998, where he has been responsible for – Satellite Network QoS – Network SLA (service level agreement) – NMS (Network Management System) – Independent Subnetworks … Communications • A comprehensive simulation of the CTBTO Satellite network was undertaken • This included aspects such as QoS, SLA issues, NMS coverage • UN member states’ private networks (the so-called Independent subnetworks) were examined. Engineering Research at Aberdeen Communications • Adaptive link layer communications protocols incorporated in the first robust reconfigurable satellite modem • Satellite communications for Road Traffic Management • Piloting the use of digital satellite TV for high speed direct-to-home Internet services Wireless Sensor Networks Modelling of WSN Wireless sensor networks offer a great deal of flexibility. Sensor nodes might be added, dislocated or removed. That means the network topology is subject to constant change. •Use of CSP in the development of reliable communications protocols in a changing network topology •Power minimisation techniques in processing and communication •Modelling of network connectivity Percolation theory for modelling network connectivity in WSN Percolation theory for modelling network connectivity in WSN Percolation theory for modelling network connectivity in WSN Wireless Sensor Networks Working prototypes - Physiological monitoring using: – RFID – ZigBee Data RFID Reader Clock Tag Energy University of Aberdeen role in WARMER WP2: Development of modular algorithms and firmware for data processing and instrument control Assistance with selection of the best software/hardware platform for implementation of the developed algorithms, taking into account flexibility, possibility of integration with other parts of the system and market-related concerns. WP3: Technology of remote data collection Assistance with networking, data fusion, image processing. UNIABDN role WP4: Networking data of water risk management The overall objective of this work package is to achieve a robust, flexible computational and data networking architecture to support water risk management. WP 4.1: Development and verification of networking protocols for distributed data processing systems. WP 4.2: Review and integration in the processing platform of the networking technology. WP 4.3: Design of a system capable of communicating via different standards at different times. WP 4.4: Integration of in situ data with satellite-derived data. UNIABDN role WP5: Hardware preparation and industrialisation of the in-situ monitoring system Assistance with integration of computation and communication algorithms developed in WP2 and WP4 inside the new in-situ monitoring system. WP7: Field experiments and satellite remote sensing Assistance with field demonstration University of Aberdeen role in WARMER WP2: Development of modular algorithms and firmware for data processing and instrument control Assistance with selection of the best software/hardware platform for implementation of the developed algorithms, taking into account flexibility, possibility of integration with other parts of the system and market-related concerns. Designing the next generation in- situ monitoring system (IMS) Processing Platform considerations Questionnaire Results • Many different electrical interfaces (RS-232, RS-422, RS-485, SDI-12, USB, Analogue) • Different communication standards (GSM, UMTS, Bluetooth) • Long service intervals (min 3, max 12 months) • Many different processor architectures (x86, XScale, ARM, 8051 derivatives, MSP430, CPLD, FPGA) • Many different programming languages in use (Assembler, C, C++, Java, Fortran, VHDL) Current Design of In-Situ Monitoring Systems Resulting Constraints for the Processing Platform • Energy Efficiency • Flexible Electrical Interfaces Increasing energy efficiency The ideal energy efficient solution are Systems on Chip (SoCs). These are very energy efficient, because: • Components are connected directly • Less components • Avoidance of unnecessary abstraction layers However, SoCs are inflexible, therefore not applicable to the Processing Platform. Refined Design Constraints SoCs are power efficient, because they avoid unnecessary abstraction layers! Therefore, our refined design constraints are: • Removal of unnecessary abstraction layers • Flexible Electrical Interfaces Overview of proposed IMS setup Storage Communication Sensors FPGA Module CPU Processing Platform Within the Processing Platform Sensor Sensor Hardware Controller 1 Accelerator 1 Sensor Sensor Controller 2 CPU Comms Comms 2 . Core Controller Module . . . . . Fast Duplex Link Sensor Storage Sensor Component Controller Controller N Specific Interface N Processing Platform Storage Module Inside the CPU Hardware Accelerator Sensor Hardware Sensor Controller Accelerator Process 1 1 Process Sensor Sensor Controller IMS Process 2 Comms Comms 2 Control . Process Controller . Processs . . . . Fast Duplex Link Sensor Storage Sensor libCSP2 Controller Process Process N Duplex Channel N CPU Storage Controller Possible Areas of Collaboration • Protocol Design – IMS to Data Centre – Data Centre to Applications • In-situ Monitoring System – What is inside your box? • Energy saving Operating System – What power saving techniques does your OS use?