page 9total16

479 Sony 480 Sony 481 Sony 482 Sony 483 Sony 484 Sony 485 Sony 486 Sony 487 Sony 488 Sony 489 Sony 490 Sony 491 Sony 492 Sony 493 Sony 494 Sony 495 Sony 496 Sony 497 Sony 498 Sony 499 Sony 500 Sony 501 Sony 502 Sony 503 Sony 504 Sony 505 Sony 506 Sony 507 Sony 508 Sony 509 Sony 510 Sony 511 Sony 512 Sony 513 Sony 514 Sony 515 Sony 516 Sony 517 Sony 518 Sony 519 Sony 520 Sony 521 Sony 522 Sony 523 Sony 524 Sony 525 Sony 526 Sony 527 Sony 528 Sony 529 Sony 530 Sony 531 Sony 532 Sony 533 Sony 534 Sony PCG-FX77G/BP PCG-FX77S/BP PCG-FX77V/BP PCG-FX77Z/BP PCG-FX950H PCG-FX990 PCG-FX990L PCG-FX99V/BP PCG-FXA10 PCG-FXA10H PCG-GR18C PCG-GR5E/BP PCG-GR5F/BP PCG-GR5N/BP PCG-GR7E PCG-GR7F PCG-GR9 PCG-GR9/K PCG-GR9E PCG-GR9F/P PCG-GRS50/B PCG-GRS55/B PCG-GRS70/P PCG-GRT25 PCG-GRT25L PCG-GRT30LP PCG-GRT30P PCG-GRT55/B PCG-GRT55E/B PCG-GRT77/B PCG-GRT77E/P PCG-GRT99/P PCG-GRT99S/P PCG-GRT99V/P PCG-GRV7LP PCG-GRV7P PCG-GRV7TP PCG-GRV88G PCG-GRV99G PCG-GRX1P PCG-GRX3HP PCG-GRX3LP PCG-GRX3P PCG-GRX52G/B PCG-GRX71 PCG-GRX81G/P PCG-GRX90/P PCG-GRX92G/P PCG-GRZ10 PCG-GRZ20 PCG-GRZ20L PCG-GRZ77/B PCG-GT3/K PCG-NV77M/BP PCG-NV7EL PCG-NV7L Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook ME XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home ME W2K XP Home XP PRO XP Home XP Home SP1 XP PRO XP Home XP Home XP PRO XP PRO XP Home SP1 XP Home SP1 XP Home SP1 XP Pro SP1 XP Pro SP1 XP PRO XP Pro SP1a XP PRO XP PRO XP PRO XP Home SP1 XP Pro SP1 XP PRO XP PRO XP PRO XP PRO XP Home XP Home XP PRO XP PRO XP PRO XP Home XP Home XP Home XP Home SP1 W2K XP Home XP Home XP Home

page 10total16

535 Sony 536 Sony 537 Sony 538 Sony 539 Sony 540 Sony 541 Sony 542 Sony 543 Sony 544 Sony 545 Sony 546 Sony 547 Sony 548 Sony 549 Sony 550 Sony 551 Sony 552 Sony 553 Sony 554 Sony 555 Sony 556 Sony 557 Sony 558 Sony 559 Sony 560 Sony 561 Sony 562 Sony 563 Sony 564 Sony 565 Sony 566 Sony 567 Sony 568 Sony 569 Sony 570 Sony 571 Sony 572 Sony 573 Sony 574 Sony 575 Sony 576 Sony 577 Sony 578 Sony 579 Sony 580 Sony 581 Sony 582 Sony 583 Sony 584 Sony 585 Sony 586 Sony 587 Sony 588 Sony 589 Sony 590 Sony PCG-NV99E/B PCG-NV99M/BP PCG-QR1S/BP PCG-QR3/BP PCG-QR3E/BP PCG-QR3S/BP PCG-R505/ABW PCG-R505AFT PCG-R505AGT PCG-R505B/P PCG-R505BFL/P PCG-R505F/BD PCG-R505FR/D PCG-R505J/BD PCG-R505MFL PCG-R505MGH PCG-R505MP PCG-R505Q/BD PCG-R505R/AK PCG-R505S/PD PCG-R505TFP PCG-R505V/BD PCG-R505VF/K PCG-R505VJ/K PCG-R505VM/K PCG-R505W/PD PCG-R505X/PD PCG-SR1M/BP PCG-SR9M/K PCG-SRX3/BD PCG-SRX3E/BD PCG-SRX3F/BD PCG-SRX3S/BD PCG-SRX55H PCG-SRX55L PCG-SRX55TH PCG-SRX55TL PCG-SRX7 PCG-SRX7E/P PCG-SRX7F/PB PCG-TR1/B PCG-TR1L PCG-TR1T PCG-TR2/B PCG-TR3E/B PCG-TR5B PCG-U1 PCG-U101 PCG-U101/P PCG-U3 PCG-U3/P PCG-U70P PCG-V505/B PCG-V505E/B PCG-V505F/B PCG-V505G/B Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook XP Home XP Home ME XP Home XP Home XP Home ME XP Home XP Home XP Pro XP PRO XP Home XP Home XP Home XP Home XP Home XP PRO XP Home W2K XP PRO XP PRO XP Home W2K W2k SP2 W2k SP3 XP PRO XP PRO ME W2K XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP Home XP PRO XP PRO XP Home SP1 XP Home SP1 XP Home XP Home SP1 XP Home SP1a XP Home SP1a XP Home XP Home SP1 XP PRO XP Home XP PRO XP Pro SP1 XP Home SP1 XP Home SP1 XP Home SP1 XP Home SP1a

page 13total16

757 Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony VGC-M50B/S VGC-RA50 VGC-RA51 VGC-RA53L7 VGC-RA60 VGC-RA70P VGC-RA71PL9 VGC-RC72DP VGC-RM53D VGC-RM91S1 VGC-RM93US VGC-V171 VGC-V201 VGC-V202RB VGC-VA200B VGN-A60B VGN-A61B VGN-A62B VGN-A63 VGN-A70P VGN-AR50B VGN-AR73DB VGN-AR91S VGN-AS33B VGN-B90PS VGN-BX4KANB VGN-BX90S VGN-BZAANS VGN-C240E/B VGN-C50HB/W VGN-E50B/B VGN-E91B/B VGN-FJ21B/L VGN-FS51B VGN-FS70B VGN-FT50B VGN-FT53DB VGN-FW30B VGN-FZ50B VGN-G1KBN VGN-K70B VGN-K71B VGN-N50HB VGN-S38TP VGN-S51B VGN-S53B/S VGN-S54B/S VGN-S70B VGN-S71PB VGN-S72PB/B VGN-SR70B/S VGN-SZ23TP/C VGN-SZ70B/B VGN-SZ750N/C VGN-SZ94US Desktop Desktop Desktop Desktop Desktop Desktop Desktop Desktop Media Center PC Desktop Desktop Desktop Desktop Desktop Desktop Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook XP Home SP2 XP Home SP1a XP Home SP2 XP Home SP2 XP Home SP1 XP Pro SP1a XP Pro SP2 XP Pro SP2 Vista Home Premium Windows Vista Home Premium Windows Vista Ultimate XP Home SP1a XP Home SP1a XP Home SP2 XP Home SP2 XP Home SP1a XP Home SP2 XP Home SP2 XP Home SP2 XP Pro SP1a XP Home SP2 Windows Vista Home Premium XP Home SP2 XP Home SP2 XP Pro SP2 Windows Vista Business XP Home SP2 Vista Business SP1 Vista Home Premium Windows Vista Home Basic XP Home SP1a XP Home SP2 XP Home SP2 XP Home SP2 XP Home SP2 XP Home SP2 Vista Home Premium Vista Home Premium SP1 Windows Vista Home Premium Windows Vista Business XP Home SP1a XP Pro SP2 Windows Vista Home Basic XP Pro SP2 XP Home SP2 XP Home SP2 XP Home SP2 XP Home SP1a XP Pro SP2 XP Pro SP2 Vista Home Premium SP1 XP Pro SP2 XP Home SP2 Vista Business SP1 Windows Vista Ultimate

page 14total16

Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony Sony SOTEC SOTEC SOTEC SOTEC SOTEC SOTEC SOTEC SOTEC SOTEC SOTEC TG TG Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba Toshiba VGN-T70B VGN-TX16TP/W VGN-TX50B/B VGN-TZ73B VGN-TZ90S VGN-U50 VGN-UX50 VGN-Y70P VGN-Z70B VGX-TP1 VGX-X90P VGX-XL1D M355V PC Station BJ3316P PC STATION E4200XP PX7513P (PC Station) Station BA9715P WinBook DN3000 WinBook DN5020 WinBook WD311 WinBook WH5514P WJ4160R TG Averatec 2500 TG Averatec 8300 A100 dynabook AX/940LS dynabook AX1/424CME dynabook Qosmio G40/97D Dynabook Satellite 2210X dynabook Satellite WXW/79CW DynaBook SS S5/280PNLN dynabook SS SX/15A dynabook TX/68D DynaBook V9/W14LDEW DynaBook WX/3727CDS EQUIUM 5080 EQUIUM 5190 PE51918CNY111 M5 Protg 3505 Qosmio E10/2JCDT Qosmio F50 PQF5088GLR Qosmio G20 PQG20390LS Satellite A100-VA9 Satellite A25-S279 Satellite L40 PSL48T-02D00J Satellite Pro 6100 Satellite X205-S9349 Toshiba Dynabook AX/53C Toshiba Dynabook CX/45C Toshiba Dynabook SS RX1/STA TOSHIBA PORTEGE M803 Toshiba Satellita L30-107 Toshiba Satellite A135 Toshiba Satellite A200 TOSHIBA Satellite M306 Toshiba Satellite P105 Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Desktop Desktop Desktop Desktop Desktop Desktop Desktop Desktop Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Desktop Desktop Notebook Tablet PC Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook Notebook XP Home SP2 XP Pro SP2 XP Home SP2 Vista Home Premium SP1 Windows Vista Home Premium XP Home SP1a XP Home SP2 XP Pro SP2 Vista Home Premium SP1 Vista Home Premium XP Pro SP2 Windows Vista Home Premium 98 SE Vista Home Premium SP1 XP PRO Windows Vista Home Premium Windows Vista Home Premium Windows Vista Home Basic Vista Home Premium XP Home SP2 Vista Home Premium XP Pro SP2 Vista Home Premium(32bit) Vista Home Premium(32bit) XP VISTA32 XP Home SP2 XP Home SP1a Vista Home Premium 98 SE Windows Vista Home Premium XP Pro SP2 Vista Business Windows Vista Home Premium XP Home SP1 XP Home SP2 XP Pro SP1a Windows Vista Business XP 98 1st XP Home SP1a Vista Home Premium SP1 XP Home SP2 Vista Home Premium XP Home SP1 Vista Home Basic XP PRO Vista Home Premium Vista Home Premium(32bit) Vista Home Premium(32bit) Vista Business(32bit) Vista Home Basic(32bit) Vista home Basic(32bit) Vista Home premium(32bit) Vista Home premium(32bit) Vista Home Basic(32bit) Vista Home premium(32bit)

This section describes the key contributions that each testbed makes to validating e-SENSE innovations. Body Sensor Network Testbed to be the basis for Show Case 1, demonstrating key features of the personal application scenarios that require capturing users physiological and psychological data in order to provide the user with security, comfort or entertainment applications. to validate the e-SENSE System Architecture, developed in WP2, by o demonstrating how context information, captured via WSN/BSNs, can be used by B3G/IMS services/applications in an end-end system implementation
Date: 19/10/2007 Security: Public Page 5/53
demonstrating how context information can be used by multiple B3G services/applications to validate the e-Stack protocol stack architecture for WSNs, by o developing an implementation that conforms to the e-Stack specification at sub-system, Service Access Point and service primitive level o implementing e-Stack on both sensor nodes and WSN gateway o implementing key e-Stack functions including communications protocol stack, service discovery, node discovery and publish/subscribe middleware. to be a portable testbed so that public demonstrations may be provided at various locations/events. o
Ambient Smart Signs to be the basis for Show Case 2, demonstrating key features that are required in several application scenarios of the e-SENSE portfolio personalised guidance in public spaces, emergency evacuations, etc. Integration of a context-aware application (Smart Signs) and WSN in an office/enterprise application environment o Transfer of dynamic context information (e.g. position) for multiple users o Large-scale deployment (entire building) Validates a number of e-SENSE technical concepts in a WSN-based environment sensor network o 3D Localisation and proximity detection mechanisms o In-field data processing.
Letibee Testbed Addresses the e-SENSE low-power air interface objective by validating the power consumption of the Letibee IEEE 802.15.4 compliant PHY implementation (target is 20 nJ/bit) Integrates the Letibee chip into a Starwatch, a wristwatch containing a 3-D attitude sensor, to produce a realistic context-capture scenario
Mood Detection Testbed Captures physiological data from human subjects during mood experiments and verifies the e-SENSE mood detection algorithms A portable BSN-based testbed, incorporating physiological sensors, and an offline algorithm development/execution environment

1.3 None

Deviation from objectives
If relevant: corrective actions

Not relevant 1.5 IPR

ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 10/53
2.5 Hardware Components The body sensor network test bed is comprised of several hardware components. This includes the actual sensor node platform, the gateway platform as well the B3G service platform. Each of the hardware platforms is described in the following in more detail. 2.5.1 Sensor node platform The sensor node platform forms the core of the body sensor network test bed. It is based on wireless sensor nodes with processing and communication capabilities, physical sensor probes directly sensing the physical phenomena and custom designed sensor-conditioning units for some of the sensor probes. 2.5.1.1 Sensor Probes The BSN provides six different sources of sensory information to extract a variety of different physiological parameters of the mobile user and information of its environment. The six different sensor probes attached to the sensor nodes of the BSN are: ECG electrodes with snap leads: used as a source to derive the heart rate and its variability of a person Piezo Respiratory Belt Transducer: used as a source to derive the breathing rate of a person Finger electrodes: used as a source to measure the electro dermal response or activity of a person Skin temperature sensor: used as a source to measure skin temperature. Accelerometer sensor: used to measure acceleration hence the pattern of motion of a person Environmental temperature sensor: used to measure the temperature of the environment a person is located in Table 1 provides a specification of the type and supplier for utilised sensor probes. Sensed Signal
Breathing rate ECG EDA/GSR Skin temperature Accelerometer Environmental temperature

Sensor probe/ model

Piezo Respiratory Belt Transducer MLT1132 ECG electrodes H92SG ECG snap on cables Z93-630 Various connectors GSR Finger Electrodes MLT116F Skin Temperature Probe MLT 491A Nano.accel Temp resistor

Supplier

AD Instruments MedCat MedCat MedCat AD Instruments AD Instruments SensiNode SensiNode
Table 1: Specification of sensor probes for the BSN system. 2.5.1.2 Signal conditioning units Two custom signal-conditioning boards have been developed for the different sensor probes. The first signal conditioning board hosts necessary circuitry for ECG and breathing rate sensor. For the ECG signal, a ASICs is used that is specifically designed for the amplification and conditioning of bio-potential signals by one of our partners [2]. The respiratory belt requires additional amplification due to the low output voltage of the piezo-electric signal source.

ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 11/53
The second signal conditioning board hosts a measurement bridge for the skin resistance of a person and an input for the skin temperature sensor. Both signal conditioning boards provide the conditioned signals as analogue output to the AD converters of the micro controller on the sensor nodes. Adequate drivers that allow a configuration and interaction with the signal conditioning boards were developed. The amplitude of the electrical signal delivered by the breathing strap sensor has been found very low and may require amplification before proper breathing rate detection can be realised. In order to avoid the additional hardware design, an off-the-shelf data acquisition board offering a programmable PSoC (Programmable System-on-Chip) has been utilised. The data acquisition board is the SensiNode Micro.data, and is compatible with the sensor node platform. In addition specific PSoC software had to be developed to perform the desired amplification. Table 2 provides a summary of utilised signal conditioning units. Data sheets for their design can be found in the appendix. Model ECG conditioning unit EDA conditioning unit Micro.data Purpose ECG, BR EDA, Skin temp Breathing rate amplification Manufacturer SensiNode SensiNode signal SensiNode QTY 1
Table 2: Signal conditioning units. 2.5.1.3 Sensor nodes The body sensor network is comprised of four sensor nodes. Two sensor nodes are utilised to host physiological sensors, a third one captures environmental data with respect to a wearer of the BSN and the fourth one used for the communication interface to the gateway device. Two different sensor node platforms have been utilised for the realisation of the BSN, as the hardware platform of the BSN has been designed in two development stages. In the first stage the physiological sensors of the body sensor network have been created, based on the state of the art SensiNode Micro.2420 research platform [1], which is shown in Figure 2-2. One of these two nodes hosts ECG and breathing rate sensors and respective signal conditioning unit, while the other hosts EDA and temperature sensor. The logical grouping has been made based on the proximity of the body positions, where the sensor probes are usually attached (chest, hand).
Figure 2-2: Utilised sensor node platforms: Micro.2420.
ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 12/53
In the second stage the BSN was extended by a node hosting both acceleration and environmental temperature probes. At this stage a node with a smaller form factor, namely the SensiNode Nano.2430 [1], became available and was selected for the BSN extensions. In addition a SensiNode NanoUSB stick, which is based on the Nano.2430 is used as the gateway device interface. Figure 2-3

Figure 2-3: Utilised sensor node platforms: Nano.2430. A summary of the sensor node platforms and their respective purpose in the BSN is summarised in Table 3. The key features of the different sensor platforms are summarised in Table 4. Model Micro.2420 Nano.2430 Micro.2420 Purpose Manufacturer Physiological sensors SensiNode Accelerometer, Env. SensiNode Temperature GW node SensiNode QTY 1
Table 3: Sensor nodes and signal conditioning units required for a BSN system. Micro.2420 Texas Instruments MSP430 10kB 256kB+ 4Mbit external Chipcon CC2420 IEEE 802.15.4 2.4 GHz 250 kbps ~ 25mA nA 40 x 40 mm Nano.integrated 8kB 128kB Chipcon CC2430 IEEE 802.15.4 2.4 GHz 250 kbps 27-50 mA 0.9 uA 18 x 40 mm
MCU RAM Flash memory Radio chip Phy/MAC standard Frequency Max data rate Rx mode current Sleep mode current Dimensions
Table 4: Key characteristics of sensor node platforms. Figure 2-4 shows the completed physiological BSN node for ECG/Respiration. Besides the sensor node casing, the left figure shows the utilised ECG snap leads/electrodes and respiratory belt. The right side of the figure shows the internals of the BSN node, including the sensor node board, signal conditioning unit and batteries.
Date: 19/10/2007 Security: Public Page 13/53
Figure 2-4: Outside and inside view of the BSN node including signal conditional units and sensor probes for ECG and respiration. Analogously Figure 2-5 depicts the physiological BSN node for EDA/Skin Temperature. On the left side of the figure, EDA finger electrodes and the skin temperature sensor are depicted. The right hand side shows again the internals including the sensor node board, signal conditioning unit and batteries.
Figure 2-5: Outside and inside view of the BSN node including signal conditional units and sensor probes for EDA and Skin temperature.
2.5.2 Gateway platform As a mobile gateway device, a Nokia 770 Internet Tablet [3], which is based on an ARM-9 micro controller core @252MHz. There are two major motivations for the selection of the device. The N770 runs a Linux 2.6.16 kernel. This allows us to easily port Linux based tools that are used for the communication with the sensor network. The second motivation is the configurable USB host functionality, to connect the bridging sensor node directly via USB.

Figure 2-8: Protocol stack architecture of the e-SENSE gateway. 2.6.2.1 Mood recognition entity The mood recognition entity computes various features of the physiological signals and a simple emotional state vector, which are made available as additional high-level contextual building blocks to the service platform. At the time of the BSN testbed implementation, only a simplified version of the mood recognition algorithm has been available. Its capabilities include only the detection of the level of arousal, interpreted as two emotional states, namely relaxed and activated.
ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 18/53
Figure 2-9 gives an overview of the signal processing chains utilized for the extraction of desired features for the mood recognition algorithm. The measured analogue signals at the sensor probes are conditioned in the sensor processing boards at the sensor nodes, before being sampled at different rates. The ECG signal is sampled at 250Hz, while the other remaining signals are sampled at 20Hz. Digital pre-filtering takes place on the sensor node before transmission to the Nokia N770. Further processing steps for the extraction of required features of the physiological signals take place in the Nokia N770.
Physiological system Analogue signal conditioning and sampling

ECG raw data

EDA raw data

Respiratory raw data

Temperature raw data

Digital Prefiltering

Peak recognition
SRL Analysis and Parameter Extraction
SRR Analysis and Parameter Extraction

Spectral analysis

HRV Parameter Extraction
Analysis and Parameter Extraction

Statistical Analysis

ECG processing chain

EDA processing chain

Breathing rate processing chain
Temperature processing chain
Analysis and data fusion for mood recognition

Emotional state vector

Figure 2-9: Signal-processing chains for physiological data. The most complex processing chain in terms of algorithmic and computational complexity is the ECG processing chain. A peak detection algorithm has been implemented with automatic threshold detection, in order to extract the RR peak distances of the heartbeat. Additional mechanisms increased the reliability of peak detection in presence of noise artefacts in the signal or packet loss. Several features are computed from the RR peak distances: the heart rate, RMSSD and PNN-50. Furthermore a spectral analysis is performed to identify high and low frequency components and their ratio. For the spectral analysis a 256 FFT is utilized. The EDA processing chain considers both the short term (ST) phasic components in form of skin resistance response (SRR), and the long term (LT) tonic components in form of the skin resistance level (SRL). The skin temperature processing chain computes short and long term development of the skin temperature over time in form of the mean and temperature rise. All physiological data is continuously provided to the processing chains and buffered for further processing. The processing chain is invoked every 5 seconds on the currently buffered data set and necessary features are extracted and provided to the mood detection

Date: 19/10/2007 Security: Public Page 23/53
2.8 Relationship to other e-SENSE WPs
The BSN Test Bed forms the technological platform to realize Show Case 1. On top of this platform two example IMS application are implemented that make use of the sensor based context information. The BSN test bed also reflects several e-SENSE concepts that have been developed in the other e-SENSE work packages. This section briefly describes the research results that have been integrated into this test bed. Implemented WP1 Concepts The test bed reflects the user expectations and requirements that are derived from several scenarios in each application space of the e-SENSE scenario portfolio. In particular, the components of this test bed address key features of the personal application scenarios that require capturing users physiological and psychological data in order to provide the user with security, comfort or entertainment applications. This test bed is particularly in line with the required functions of the community application space (healthcare scenarios) since the monitoring of patients requires context data that are captured by BSN nodes. In the Industrial applications space, the e-SENSE scenarios have also demonstrated the needs for such BSN system, for instance to monitor the workers condition and security. Implemented WP2 concepts The test bed components reflect the e-SENSE system architecture developed in WP2. This includes the e-SENSE protocol stack architecture with all its subsystems that is implemented on the e-Stack on the WSN nodes including the gateway node. The e-Stack realizes the different e-SENSE subsystem as separate tasks on top of the operating system. Function calls for the invocation of services that have been implemented in different subsystem also follows the semantics defined by the service primitives at the respective service access points of the architecture. In addition prototypes of the gateway extensions and the e-SENSE service enabler have been implemented to provide the necessary integration of the e-SENSE system into IMS. Figure 2-13 provides an overview of the protocol stack configurations of the test bed components. As can be seen from the figure the service enabler resembles all specified service functions, except knowledge repository, task graph data base and authentication and authorisation. The basic service functions for the gateway extensions are completely implemented. Each subsystem of the e-SENSE system architecture implements a key set of service functions that is required for the basic operation of the WSN. These service functions are realized by protocol elements developed in WP3 and WP4 and are described in the following.

Date: 19/10/2007 Security: Public Page 24/53

APMA-SAP

Figure 2-13: Overview of the protocol stack architecture of the test bed components. Implemented WP3 concepts WP3 mainly concentrates on the development of protocol elements for the connectivity subsystem. The test bed stack implements an 802.15.4 PHY and MAC as well as a 6LowPAN compliant network stack. Implemented WP4 concepts Several key protocol elements developed by WP4 are implemented as part of the e-Stack. This includes the service discovery and node discovery functions in the management subsystem and a light-weight publish subscribe middleware.
Date: 19/10/2007 Security: Public Page 25/53
Contract: 027227 Deliverable report WP5 / D5.2.1 Section 3 - Ambient Smart Signs Testbed

3.1 Objectives

The objectives of the Ambient Smart Signs testbed are to be the basis for Show Case 2, demonstrating key features that are required in several application scenarios of the e-SENSE portfolio personalised guidance in public spaces, emergency evacuations, etc. Integration of a context-aware application (Smart Signs) and WSN in an office/enterprise application environment o Transfer of dynamic context information (e.g. position) for multiple users o Large-scale deployment (entire building) Validates a number of e-SENSE technical concepts in a WSN-based environment sensor network o 3D Localisation and proximity detection mechanisms o In-field data processing.
Relationship to Show Cases
The primary purpose of the e-SENSE Show Case 2 is to demonstrate how environmental context information can be integrated with a context-aware intelligent signage application in a large-scale office/enterprise scenario. In Show Case 2, the environmental context is captured by environmental sensors (temperature, humidity, smoke, etc.), deployed in a five-storey building, processed and communicated by a wireless sensor network to the Smart Signs system. The location of occupants and visitors is tracked through the localisation capability of the wireless sensor network. Show Case 2 demonstrates two scenarios: Normal Situation, where employees and visitors are guided, via wall-mounted displays, to their requested destinations. The messages displayed are personalised as the signage system can identify the screen nearest to the individual. Emergency Situation, where the environmental sensor network detects a potentially serious situation such as a fire, by fusing multiple sensor outputs, and passes the alarm to the smart signage system. The signage system coordinates the evacuation of the building, taking account of the location of potential hazards and occupants.

The location where test bed 2 is deployed is the Faculty of Computer Science of the University of Twente in Enschede, The Netherlands. The building has a total of five (5) floors and the entire building is subject to the testbed. Smart Signs are placed throughout the building at key points and the Ambient Network is also deployed throughout the network but only provides location information on a single floor.
ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 27/53
To cover the Faculty of Computer Science and its side building, 25 Smart Signs will be deployed at key points. Each Smart Sign consists of a LCD and an e-box attached at the back running Linux. To each Smart Sign, an Ambee Module is attached that is used for proximity detection of users in the vicinity of the Smart Sign to show each user relevant information only when he is close to the Smart Sign. An Ambient WSN is deployed throughout the building. An Ambient WSN consists of a single Gateway and additionally up to 31 MicroRouters. This should be sufficient to cover the entire building. For future research and as a continuously operating demo and testbed, the network is deployed for permanent usage and thus powered by mains. Coverage of the floors is required to be sufficient for deploying various sensors. Effectively, the aim is to have at least a number of temperature and humidity sensors, as well as smoke detectors hooked up to the network. All sensors (temperature, humidity, smoke) are combined into a single component and accessed by an EndPoint. The sensors are required to report their sensor readings once every few minutes (simultaneously, an alive message). The smoke detector is sampled on a regular basis and only in case of an emergency should it report itself. Integration of these sensors should shed some light into potential for sensor fusion techniques. Ambient will look also look into potentially adding additional sensors for detecting fires, such as CO2 and or CO, if time allows. At least every corridor a floor is fitted with EndPoints that function as localization beacons (10 at each floor). This should give a coarse-grained location estimate. Additionally, the fourth floor is fitted with additional beacons and should provide a fine-grained location estimate. Additionally, the major staircases are fitted with beacons as well. These beacons can be configured with the backbone Ambient Network infrastructure and transmit a beacon-signal at certain intervals. Effectively, the system provides a 3D position indication. The SmartTag carried by the employees and visitors checks every half minute for beacons and if beacons are detected calculates its position and reports this to the Gateway using the backbone Ambient Network infrastructure. The beacons provide a simple adaptive localization approach to environmental influences which are made available to the SmartTags via the beacon signal. 3.3.3 System Design

Date: 19/10/2007 Security: Public Page 37/53
Any application task interested in that data can subscribe to it through the publish/subscribe mechanism provided by the AmbientRT operating system. Outgoing Pushes The application or driver can send a push message by using the Send Push function. The message is then sent using either the serial line protocol or the network protocol stack based on the destination of the message.
3.6.1.2 Ambient Studio A known problem of WSN technology is that the visualization and deployment of the networks is still cumbersome and time consuming. The plug-and-play factor is not yet to the level where we want it to be. Errors are often hard to spot and for laymen hard to solve. For this reason Ambient has been developing on a software application that will help its customers with the deployment and interaction of its networks. Ambient Studio is effectively a component-based software application that is primarily used to visualize and interact with the network and additionally provide help with the deployment of WSNs. The architecture allows for easily extending the application framework with additional components and plug-ins. The other reason for its creation was to streamline and ease internal development of WSN applications. A screen-shot of the network visualization tool is shown in Figure 3-8.
Figure 3-8; Ambient Studio Snapshot
Date: 19/10/2007 Security: Public Page 38/53

3.6.2 Smart Signs

Smart Signs deployment includes preparation of (i) Central Services, (ii) Smart Signs, and (iii) Kiosk client.
Central Services: Central Services provide an API for making guidance and message requests. They run on a central server. Clients running on kiosk-terminals, PDAs, Smart Phones etc. can access these services via a Web interface or Java RMI. Although, not shown in the picture, the central services also handle the user and group administration and provide an appropriate API.
The central server can run on Linux, UNIX, Windows or any other operating system supporting Java 1.6. Additionally, it requires at least Postgres 8.2.4. Preferably, the server should also have an FTP Server to provide software updates to the Smart Signs. The server needs a known IP address and communication to the server is done through any Internet connection. The software package running on the server also includes the Smart Signs system Simulator, which allows simulating the functions of the deployed Smart Signs network and the context received via the context infrastructure. It visualizes the Smart Signs network on a floorplan of the building(s) and the operator can drag user-figures around the floorplan and immediately see which guidance and messages these users would get (see Figure 3-9).

Relationship to other e-SENSE WPs
The test bed supports key features that are required in several application scenarios of the eSENSE portfolio: personalized and guidance information via ubiquitous displays in public spaces, rescue guidance for the healthcare personnel in hospitals (or even for the hospital visitors), guidance for workers in offices, in warehouse and factories as well as emergency guidance for rescue teams in industrial contexts, guidance of consumers in shopping centres, theme parks, ski resorts etc. The testbed validates a number of e-SENSE technical concepts, including Localisation and proximity detection mechanisms (WP4) In-network data processing & event detection (WP4) Context-awareness (WP4)
Date: 19/10/2007 Security: Public Page 42/53
Contract: 027227 Deliverable report WP5 / D5.2.1 Section 4 - Letibee Testbed

4.1 Objectives

The Letibee testbed in development at the CEA-LETI consists in a RF transmission system using the Letibee device, designed in WP3 of e-Sense. It is based on a previous RF system developed at Leti named Starwatch. It is composed of a set of wristwatches which are equipped with some accelerometers and magnetometers and which can transmit their information over a low power radio link. Therefore, the watches can transmit the movement and position of the person who is wearing them. The original Starwatch RF link was developed with a proprietary protocol and an off-theshelf RF device. The Letibee testbed will use this system, removing the off-the-shelf RF device and replacing it by the Letibee. Therefore, the movement and position of a person who is wearing them is captured and the information centralized onto a PC unit. The objective of this Letibee testbed is to demonstrate the low power consumption ability of the Letibee in a real demonstrator environment. The proposed testbed also conforms to the e-SENSE objective of gathering sensor information to a concentrator (the PC), which can act as a gateway. 4.2 Relationship to Show Cases
The test bed is not used for the e-SENSE show cases. It is used to prove the e-SENSE concept regarding low-power PHY communications.

Testbed Architecture

The Letibee testbed is intended to demonstrate the advance in power consumption of the Letibee device. The communication link, which is currently being developed, has thereby been simplified. So, it will not include any e-Sense protocol or stack for the purpose of this testbed. The architecture that has been considered mainly enables the active mode of the chipset in order to make possible current measurements. The testbed is composed of a master unit connected to a PC through a serial link and two slave units which embed the sensors (3 x magnetometers + 3 x accelerometers). The protocol between the master and the slaves is using two addresses to select the correct sensor unit to receive back the information of interest. The PC will then be used as the terminal exhibiting the information acquired by the sensor units, i.e. movement operated by the person wearing the watch. As the main purpose of this testbed is to be able to demonstrate the low power consumption, the communication link with only 2 slaves and 1 master unit is intended to allow simple time schemes. This way, the communication functionality is at the same a tool for the measurement of the energy efficiency of this 802.15.4 PHY layer. 4.4 Functional & Technical Requirements

ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 44/53
accessible to measurements thanks to the repetitive programming of the communication between the master and the slave. 4.5 Hardware Components
The Letibee testbed boards will present a form factor compatible with the Starwatch. They are populated with the Letibee device (QFN 56) and some application components required to operate the testbed. The Letibee device can operate the reception and the transmission of the data, including the frequency synthesis, base band processing and Controler. Thus, the testbed has to include some other components like the sensors and their converters. The sensor node boards are populated with 3 accelerometers and 3 magnetometers. Also, one or several ADCs, connected to a multiplexer, will operate the conversion of the analogue signals to digital data. Despite the embedded C on the Letibee SoC, another C or FPGA has to be populated on the testbed board. The main reason is the fact that the Letibee doesnt contain any EPROM or Flash memory, thereby cannot maintain a program once it has been switched off. The external C will do this by loading the program into the Letibee memory at start up. It is to be noticed that the Letibee SoC design purpose was mainly to demonstrate the low power consumption in active mode of such a 802.15.4 PHY layer. The EPROM memory embedding was not an axis of research for this work, as it doesnt impact the power consumption, except in power down mode. But it has a major impact on the cost of the technology and the overall layout considerations. So, the Letibee SoC, due to the time constraints of the e-Sense project, has been developed without such memory. The external C can also be used for other process in case the full-functionality of the embedded C is not completed. The envisioned device is to be chosen out of the MSP430 series, a QFN 4x4mm with 8kBytes Flash memory. As a first step, the processing of the Letibee programming will be done by an FPGA supporting a 40 MHz clock. The reason is that the full-functionality of the Letibee has not been demonstrated; thereby the embedded test mode on the chipset has to be used. This test mode cannot be addressed by the embedded C, so the external processor has to program some extended registers, in addition to loading the program into the RAM of the Letibee. Nevertheless, the communication layer of the demonstration will be managed by the integrated C 8051, only the initial programming of the chipset will require the use of the external devices. The testbed also requires a reference oscillator as it has not been implemented in the chipset. In the proof-of-concept phase (T3.3 in WP3 ), the measurements have been carried out using an external 40 MHz frequency synthesizer. A crystal oscillator will be implemented on board. The typical form factor of this device is 3x3 mm with a current consumption in the range of 1 mA. Indeed, if this function had been integrated on chip, the current consumption would have been rather lower, in the range of some 100 A. The demonstration board also manages the battery power source, as the Letibee SoC has to be supplied by VCC=1.2 V. 4.6 Software Components

ID: e-SENSE-WP5-D5.2.1 Revision: Final Date: 19/10/2007 Security: Public Page 51/53
5.8 Relationship to other e-SENSE WPs
The design is a first step towards mood detection and therewith a preparation of the moodbased applications envisioned in the project (D1.2.1). The mood detection algorithms are developed and validated in Task 4.2 of WP4.

Results

A series of experiments have been conducted at the University of Surrey in September 2006 and in Berlin at HFC in July 2007 in order to obtain more insights on the relationship of emotional states and physiological parameters. Based on a detailed analysis of data recorded at those experiments, HFC developed algorithms that are capable of inferring some emotional states of a person (Figure 5-3).
Figure 5-3 Classification of the two states happiness and sadness via the two normalized parameters EDA-sd and zygomatic EMG activity. Blue circles represent data vectors taken from the happy film; the red crosses the data from the sad one. The wrong classified data points are framed in bold.
Date: 19/10/2007 Security: Public Page 52/53
Contract: 027227 Deliverable report WP5 / D5.2.1 Section 6 - Abbreviations
ADC AS B3G BSN ECG EDA EMG ESN GSR IMS ODBC PSoC SoC WSN Analogue to Digital Conversion Application Server Beyond 3rd Generation Body Sensor Network Electrocardiogram Electrodermal Activity Electromyography Environmental Sensor Network Galvanic Skin Response IP Multimedia Subsystem Open Database Connectivity Programmable System on Chip System on Chip Wireless Sensor Network

Section 7 - References

[1] SensiNode Ltd: http://www.sensinode.com/ [2] R. F. Yazicioglu, P. Merken, R. Puers, and C. Van Hoof, A 60W 60 nV/Hz Readout Front-End for Portable Biopotential Acquisition Systems, IEEE International Solid-State Circuit Conference 2006 (ISSCC 2006), 4-9 February 2006, San Francisco Marriott, CA, USA [3] http://europe.nokia.com/770 [4] IPv6 over Low power WPAN (6LowPAN) working group, www.ietf.org/html.charters/6lowpan-charter.html, last access on 31.8.2007. [5] D2.2.1: Initial e-SENSE system architecture, e-SENSE project deliverable, November 2006 [6] IR2.2.2: Report on work item service platform, reconfiguration and energy management, e-SENSE project internal report, September 2007 [7] J. Wu, L. Chen, J. Zhou and H. Jiang, A New Reliable Routing Method Based on Probabilistic Forwarding in Wireless Sensor Network, Proceedings of the 5th International Conference on Computer and Information Technology, IEEE, Washington, DC, USA, September 21 - 23, 2005 [8] libXml2: The XML C parser and toolkit of Gnome, http://xmlsoft.org/, last accessed 31.08.07 [9] SOFIA SIP library, Sofia-SIP is an open-source SIP User-Agent library, http://sofiasip.sourceforge.net/, last accessed 31.08.07 [10] FreeRTOS, A portable open source mini Real Time Kernel, http://www.freertos.org/, last accessed 31.08.07 [11] Koralewski Industrie-Elektronik oHG, http://www.koralewski.de/