An Optimal Design and Deployment of Wireless and Optical Sensor Networks

无线传感器网络由许多密集部署的传感器节点组成，这些节点通常随机分布在一个区域内。传感器节点体积小，计算能力低，功率有限。由于其在监视、军事、栖息地监测等方面的广泛应用，在过去的十年里，它们在研究中受到了相当大的关注。无线传感器网络的目的是通过许多传感器节点收集关于现象的数据，并将其转发到sinkode或融合中心进行进一步处理，这可以通过多跳中继进行。传感器可以是在光学频率区域上操作的光学传感器，或者是射频传感器。在无线传感器网络中，传感器的部署策略对网络的性能和有效寿命有着巨大的影响。在这个建议中，我们首先调查的最佳传感器部署问题在三个维度。我们的目标是找到方便的封闭形式的解决方案，这样的部署计划，往往是由非凸优化问题建模。这一部分的研究结果对真实的应用具有重要意义，并可推广到非无线传感器网络领域，如蜂窝通信系统中的最优发射机和接收机定位。第二，我们的目标是最大限度地减少在无线传感器网络中的检测时间，所有传感器在同一频带上工作，并在同一时间传输数据。这里的结果是非常重要的即时multitargetdetection技术在极端延迟和频率的关键应用。换句话说，使用我们提出的技术，在网络中的多个目标检测是即时的，而频分复用不涉及。值得一提的是，我们提出的方法也不同于传统的码分复用方案。我们的方法的实用性将通过实验验证。

A WSN consists of many densely deployed sensor nodes that are often randomly distributed in an area. The sensor nodes are small in size with low computational capacity and limited power. Due to their wide range of applications in surveillance, military, habitat monitoring,etc., they have received considerable attention in research in the last decade. The purpose of a WSN is to collect data about a phenomenon by many sensor nodes and forward them to the sinknode or a fusion center for further processing, which could be conducted through multi-hop relays. The sensors can be either optical sensors, which operate on the optical frequency regions, or Radio frequency sensors. In a WSN, the sensor deployment strategy has ahuge impact on the network’s performance as well as its effective lifetime. In this proposal, we first investigate the optimal sensor deployment problem in three dimensions. The objective is to find convenient closed form solutions for such deployment schemes that are often modeled by non-convex optimization problems. The results of this part are of substantial importance for the real world applications and are extendable to non-WSN fields, such as optimaltransmitter and receiver positioning in cellular communication systems. Second, we aim at minimizing the detection time in a WSN where all sensors operate on the same frequency band and transmit data at the same time. The results here are important for instant multitargetdetection techniques in extremely delay and frequency critical applications. In other words, using our proposed technique, the multiple target detection in the network is instant, whereas the frequency division multiplexing is not involved. It is worth mentioning that our proposed approach is also different from the conventional code division multiplexing scheme. The practicability of our approaches will be verified by experimental validation.