Channel Error Suppression using Correlation of Sensing Infcmnation in Wrekss Sensor Networks

本研究的目的是利用单个传感器节点观测到的数据之间的相关性来提高无线传感器网络的性能。如果相关性为1，则可以将其视为分歧接收，因为情况与接收两个具有相同信息的stunk相同，因此，当观测数据存在相关性时，可以减小误差对无线通信信道的影响。在本研究中，我们考虑联合信道提取Turbo码的多个相关的数据，由传感器节点密集部署在传感器领域的观察。我们专注于观察数据的相关性，并试图通过迭代过程来减少煎煮臂。对于多于两个相关数据使用实际信道码的方法仍然没有提出。在扩展到多于两个传感器节点的情况中的问题是如何使用从观测数据获得的相关性信息。在本研究中，我们提出了一种迭代信道解码方案，使用它们与加权。我们发现，当传感器节点的数量增加，解码性能的改善不能实现简单的加权，因此需要一个更合适的权重。本文从无辅助BPSK的解析公式中找到了使误码率最小的最佳权值，并将其应用于Turbo码的情况，进一步给出了两个相关二进制信源的基于Turbo码的迭代联合译码方案的外部信息传递（EAT）。使用EXIT图研究了所提出的系统的收敛行为。

This research has aimed to achieve the performance improvement of the wireless sensor networks by using the correlation between the data observed by an individual sensor node.If the correlation is one, it can be thought as a diviersity reception as the case is the same as the reception of two stunk having same information, Thus, the influence of error on the wireless communication channel can be reduced when there is a correlation in the observational data.In this study, we consider joint channel decocting of Turbo code fur multiple correlated data that are observed by sensor nodes densely deployed in a sensor field. We focus on the correlation properties of observation data and try to reduce decocting arm by an iterative procedure. An approach to use practical channel codes for more than two correlated data is still not presented. A problem in the extension to cases of mom than two sensor nodes is how to use the information of correlation obtained from observation data. In this studs we propose an iterative channel decoding scheme that uses them with weighting. We show that when the number of sensor nodes is increased, decoding performance improvement cannot be achieved by simple weighting, and so a more appropriate weight is needed. We find the optimum weight that minimizes the bit error rate from the analytical formula for unaided BPSK and apply it to the case of Turbo code.Further, we show Extrinsic information transfer (EAT) of turbo based iterative joint decoding scheme f or two correlated binary sources. The convergence behavior of the proposed system is studied using EXIT charts.