物理信息敏感的自适应WSN通信管理策略

Low-cost, low-power and miniaturized sensor nodes can construct wireless sensor networks (WSNs) supporting many applications, such as smart grid, environment monitoring, and wild animal tracking. The emerging Internet of Thing (IoT) applications also largely rely on WSNs. Different from some conventional networks; many sensor nodes are usually deployed in dynamic and even hostile environments, e.g., the environment monitoring or industrial field monitoring applications. The fluctuations of environment factors have a great impact on the working status of those sensor nodes and the communication links as well. However, the conventional protocol designs have little concerns about the dynamic environment, which loose the chance to further improve the system performance. ..Similar to many networked/distributed systems, many WSN functions require that all nodes have a common notion of time to facilitate data transmission, localization, sleep and transmission scheduling, information fusion, etc. However, the output value of a node’s local clock, local time, is often different from that of another node, leading to clock offset. An unbounded clock offset will degrade the network performance and even endanger the proper functioning of WSNs. Therefore, clock synchronization, the process of mitigating clock offsets among different nodes, plays an important role in WSNs. Clock skew, the instantaneous clock drift rate between two or more clocks, i.e., the difference of ticks from different clocks, is the inherent and dominant reason causing clock de-synchronization. The clock skew is mainly caused by imperfect oscillator and the peripheral circuit. How to accurately estimate clock skew and compensate it is non-trivial as the status of clock is highly related to the working environment, e.g., temperature. ..The WSN applications usually rely on wireless communication links. Compare with the wired communication, wireless channels are usually open and suffer from the fluctuations, such as shadowing, fading, etc. It is well known that the relative moving between transceivers will introduce the channel fading. Actually, the changing of environment will also cause the change of channel conditions, e.g., humidity, moving objective around. The channel fluctuation caused by the change of environment is usually in slower pace compared with the fading effect. How to adaptively adjust the communication schemes according to the dynamic environment is a promising research direction. .

无线传感器网络（WSN）由大量的低成本、低功耗的小型无线传感器节点组成，可以实现多种传统手段较难完成的任务，如环境监测，健康监护等。通信管理是协调、调度及控制网络资源的重要手段，对于WSN网络的工作有着举足轻重的作用。然而相较于传统的网络系统，WSN节点的工作环境通常都是高度变化的，例如温湿度、光照、电池电压等。这些动态变化的环境因素会影响WSN节点的工作状态，例如节点的时钟电路及通信链路质量，给WSN的通信管理带来挑战。因此在本课题中，我们拟在现有研究基础上，首先通过实验手段进行测量，通过测量数据确定若干关键的环境因素，研究其具体的作用方式和机理，并对其建模。在此模型的基础上，我们可以利用本地物理环境信息，在WSN的工作过程中对可控因素进行补偿。同时，针对不受控的物理环境，如无线链路，结合数据流特征，设计一套自适应的通信管理策略，以期提高WSN的通信效率，降低能耗，延长网络的生存时间。

在项目的支持下，我们系统的研究了环境温度因素对低成本、大范围部署的WSN节点时钟同步的影响，实验结果 表明，只要偏离了室温，晶振的频偏就会逐渐增大/减小，使得传统的同步方案效果变差。基于这样的观察， 我们提出一种基于局部信息的时钟同步方法，在不依赖时钟同步数据包交换的情况下，能够将时钟同步周期提 高1~2个数量级;在此基础上，我们又进一步研究了电池供电电压对时钟同步的影响，电池电压会发生变化， 提出一种自治的时钟偏移补偿方案，能够进一步提高时钟同步精度。我们还研究了非二进制通信的可行性及其 性能。对于二维平面而言，六边形的分割效率更高，这就意味能够使得通信系统更好的应对由于环境变化导致 的动态、时变信道。我们提出H-QAM(Hexagonal Quadratic Amplitude Modulation)。同时，由于六边形调制 往往不能提供2的整次幂个星座点，因此我们提出利用非二进制编码直接对信息进行差错控制编码，以更好的利 用六边形调制。链路吞吐量和网络吞吐量均有显著提高。同时，追踪国际研究热点，开展了可见光通信的研究 。不仅提出一种新的可见光通信调制方法，相比于传统的OOK(On-Off Keying)调制系统，提出MILC(Multi-I llumination Level Communication)，即自适应的依据环境光强变化、利用多个不同的光信号强度对数据进行 调制，显著的提高链路吞吐量。此外，结合频率调制，提出一种联合频域与幅度变化的可见光调制方案FAM(Fr equency Amplitude Modulation)，能够在两个信号自由度上自适应的改变调制速率，在此基础上我们还提出 结合移动计算，利用可见光进行伪AP(Access Point)的识别。研究目标及项目各约束性指标均已完成。