Physical layer and application layer issues of wireless sensor networks

The potential for collaborative, robust networks of wireless sensors has attracted a great deal of research attention. For the most part, this is due to the compelling applications that wireless sensor networks will enable. Location sensing, environmental observation and surveillance, medical monitoring and a lot other applications are all gaining interest. However, wireless sensor network poses a large number of challenges. Among all, one of the most important challenges is design sensor networks that have long network lifetime, which will become especially difficult due to the energy-constrained nature of the sensor nodes. In this thesis, we focus on physical layer issues and approaches to design algorithms.; Virtual Multiple-Input Multiple-Output (MIMO) structure is very attractive to wireless sensor networks due to its huge diversity gain and potential ability to save energy. With channel state information at transmitter side, water-filling algorithm can be applied to optimize the power allocation in each sub-channel of the MIMO system based on the estimated channel matrix H, so as to maximize the channel capacity of the system even in deep fading scenario. In reality, however, both the estimation error and the delay will iv be introduced when estimate the channel. We will investigate how the estimation error will impact the optimal water-filling strategy in wireless sensor networks.; Interferences due to the hostile environment and the Multi-User Access are critical factors affecting performance of the Wireless Sensor Networks. There is clearly a need of a system that can survive from the severe interference. A hybrid Frequency Hopping/Time Hopping-Pulse Position Modulated Ultra Wide Band system is proposed for Wireless Sensor Networks to confront the hostile environment. Frequency-hopping and time-hopping are both used to get as much diversity gain as possible. An exact analysis is also derived to precisely calculate the bit error rates for both Additive White Gaussian Noise channel and path-loss channel in the presence of multitone/pulse (tone in frequency domain and pulse in time domain) interference and Multi-User Interference.; For event-centric wireless sensor networks, event detection is one of the key issues. Two methods are proposed to detect the event, one is Double Sliding Window Detection, the other one uses Fuzzy Logic approach. We evaluate the event-detection approaches based on the acoustic data collected by the test-bed in different experiments.; To further extend our work, we demonstrate that real-world sensed acoustic signals are self-similar, which means they are forecastable. We showed that a type-2 fuzzy membership function (MF), i.e., a Gaussian MF with uncertain mean is appropriate to model the sensed signal strength of wireless sensors. Two fuzzy logic systems (FLS), a type-1 FLS and an interval type-2 FLS are designed to forecast signal strength. Therefore, we can not only detect the event, but also forecast the event based on the forecasted signals. Simulation results show that the interval type-2 FLS outperforms the type-1 FLS in signal strength forecasting and the performance of event detection based on the forecasted signal from type-2 FLS is much better than that based on type-1 FLS.; Another issue in wireless sensor networks is redundancy. Not only does the data of one sensor node have self-similarity, but the data from adjacent sensor nodes also have cross-similarity. Clearly, there exists highly redundancy in the collected data from sensor nodes in the neighborhood. Due to the intrinsic properties wireless sensor networks have, e.g., energy constraint, bandwidth limitation, this kind of information redundancy will impact the whole networks in a negative way. We propose to use Singular-Value-QR Decomposition(SVD-QR) to reduce the redundancy in wireless sensor networks.