Search for a command to run...
6 15-322-3 162 Abstr-act-The design of reliable, dynamic, fault-tolerant services in wireless sensor networks is a big challenge and a hot research topic. In this paper an optimization method is proposed that can be used to tune parameters of the middleware services and applications to provide optimal performance. The optimization method is based on simulation, and is capable of handling 'noisy' error surfaces. The proposed optimization algorithm is illustrated by a new spanning-tree formation algorithm, which can effectively operate even if links between nodes are asymmetrical. In the near future large-scale sensor networks will be the key elements of embedded systems used in space and aviationrelated challenges, e.g. monitoring and control of safety critical systems [l], Smart Surfaces, Smart Dust [2], or can be used to make everyday life more comfortable, e.g. Intelligent Spaces [3]. These sensor networks often use distributed operating system-like services (called middleware) over wireless communication protocols, which must be fault tolerant and adaptive because of the dynamic network topology and changing mission objectives. The design of such middleware services is not straightforward, since the sensors have limited resources, and thus the used protocols are usually very simple compared to ones used in wired communication schemes. The nondeterministic nature of the environment is another factor making the design more difficult. This paper presents a simulation-based optimization method that can be used to tune the algorithms used in the middleware layer. Also some results are presented that were gained by the proposed method. The hardware structure of the wireless sensors may vary greatly, but invariably each of the intelligent sensors is a compact device with its own power source, it contains a processing unit (a small microprocessor), a communication unit and the sensor itself. The widely used Berkeley fieldnodes (or motes) have similar structure containing an 8-bit microcontroller, a 916.5 MHz radio and several interchangeable sensors. These tiny units have a simple local operating system called TinyOS. Application-specific middleware services can be added to provide an interface between the application and the primitive services of the local operating system. The middleware can also be considered as a distributed operating system that establishes network-wide resources and functions that the applications can utilize, e.g. leader election, spanning tree formation, distributed consensus and mutual exclusion, distributed transactions, group communication services, clock synchronization, etc. Typical applications may include hundreds or thousands of motes with often unknown or random distribution (e.g. motes dropped from an airplane to a hostile environment). The communication services must be reliably established to achieve the overall goal of the distributed sensor system. During the operation of the sensor network different metrics for the quality of service (QoS) are required, a dynamic tradeoff is necessary between accuracy, response time, power consumption, and other qualities of interest. Thus, the middleware services must be prepared to adapt to the actual circumstances and the QoS metric. To design such middleware services, the highly random nature of the environment (wireless communication with possible disturbances, random layout, possibly damaged motes, etc.) must be taken into consideration. The proposed design method is a probabilistic simulationbased optimization that can help the designer choose the right algorithm with an optimal parameter set. The MATLAB-based simulator is capable of simulating the important aspects of the communication scheme: local OS services including the network protocol stack, and also the radio transmission phenomena (signal power vs. distance, fading, collision, disturbances). In the simulation