Instrumentation and Communication (IaC)

Testdata lagras i fordonet för senare analys. Dynamiskt valda delmängder av testdata skickas vidare i realtid via trådlös överföring för omedelbar analys. Den trådlösa överföringen användas även som trådlös Internetuppkoppling (har provats med framgång säsongen 2011 på Colmis med namnet HOTSPOTS on ICE).

Systemet med trådlösa sensorer för instrumentering och kommunikation bygger till stor del på befintlig teknik. För att åstadkomma bra funktion och prestanda bör systemet byggas med den senaste tekniken som förbättras med forskning kopplat till utmaningar med trådlösa sensorer för mätning.

Systemet ska använda infrastrukturlösningen HOTSPOTS on ICE. Arbetsnamnet på systemet är ”Fast Instrumentation for Testing (FIT) for ICE, FIT4ICE”. 

Projektledare: Ulf Bodin

Research Engineer: Krzysztof Wolosz

The first part of the project aims at developing a system inside the vehicle with modern wireless components. The Mulle sensor is used as wireless node. It collects and processes the data from different kind of sensors e.g. thermometers, accelerometers, vibration sensors, gyroskope, compass, Road Eye sensor, CAN on-board diagnostics (OBD) etc.

The ZigBee standard is used for the communication between nodes. It provides low data rates and support low battery consumption. The additional Bluetooth 2.0 wireless technology enables communicating the data to mobile phones, tablets or laptops.

The Constrained Application Protocol (CoAP) is used for the system to achieve easy configuration. All of the data collected can be presented in real-time on a smartphone or tablet equipped with the Android operating system. For further analyze the data is stored on a CoAP server.

The second part of the project concerns the distributed control system, which explores the wireless transmission parameter behavior. This is to find the most efficient parameter setting for the varying traffic and demand in the wireless network.

The physical parameters of the transmission e.g. RSSI and LQI as well as data link parameters e.g. throughput and end-to-end delay are analyzed to predict the communication capacity. The goal is to control and balance the load and find the best parameter settings to achieve the highest possible throughput and reliable transmission. All of the experiments are executed in real benchmark application and in the discrete-event network simulator NS3.

One of the wireless sensor networks (WSN) common solution is ZigBee, especially when low power consumption is demanded. ZigBee may however provide unpredictable throughput although transmission distances are short. This is especially evident in difficult environments with complicated reflections and various materials through which radio signals need to pass through.

Distributed scheduling based on cognitive networking principles may improve both network predictability and overall throughput. We intend to design a dynamic system that changes configuration based on the analyzed parameters to automatically adapt to its current usage and location of nodes.

The following results were achieved until now:

• it was found transmit power, distance, message size and number of nodes impacts the link capacity and resulting throughput

• throughput is correlated with RSSI and LQI, which indicates that throughput can be predicted from those parameters and may hence prove useful in designing a cognitive and distributed scheduler

• it is designed the cognitive system using Utility Function and Fuzzy Logic to distribute the load in NS3