An Empirical Study On Link Quality And Spatial Correlation In Wireless Sensor Networks

Integrating data collection,transmission,processing and wireless communication,Wireless Sensor Networks(WSNs)have the merits of large coverage area,simple implementation and low cost,and has been drawing increasingly attention,and implemented in industry,agriculture,military,medication and environment surveillance and etc..Being influenced by path loss,fading,noise,multipath effect,external interference and other factors,the transmission quality between two nodes in WSN is unstable,and the wireless links perform unstably with packet error/loss and vulnerability to the environment.Hence,the characteristics of wireless link have become research hotspots.On the basis of measurements run on a wireless links measurements testbed designed by this paper,this paper mainly focused on the link quality estimation and spatial correlation in WSN and the main work are as follows:(1)This paper designed a testbed for link measurements in WSN called STWM(Simple Testbed for Wireless Links Measurements).This testbed consists of a WSN module,a data collection module,a data process module and a node control module,and provides users for measurements of PRR,LQI,SNR,etc.By module reconfiguration,users can run other measurements on STWM according to specific requirement.(2)Traditional mean LQI value only takes the LQI value of successfully received packets into the calculation of mean LQI value on links,resulting in imprecise estimation.To compensate the drawback of traditional method,this paper presented an improved LQI-based link quality estimation mechanism called imLQI(improved LQI).By analyzing 3 different causes of packet loss:1.Sync bits error;2.Payload error;3.Signal strength lower than the sensitivity threshold,imLQI adapted different methods to estimate the LQI values of lost packets caused by different reasons,and used these values in the mean LQI calculation.After measurements on real testbed,we found that the mean LQI value computed by imLQI had better correlation with PRR(Packet Receive Ratio),which mean imLQI reveals link quality more precisely.Meanwhile,we built a link quality estimation model based on LQI according to our experiment results by curve fitting,and testify the correctness of this model by analyzing the standard variance of its estimation error.(3)Considering the impact that link special correlation have on the upper protocols and schemes design and optimization,this paper analyzed the influence that distance and power have on link special correlation by running large amount of link measurements,and utilizes the κ factor to quantize spatial correlation among links.To start,this paper studied the relationship between the κ factor and the distance between sender and receivers by alter the distance between sender and receivers.Then,by changing the distance between receivers,this paper analyzed the relationship between the κ factor and the distance between receivers under different distance between sender and receivers.At last,by adjusting the transmission power of sender,this paper attained the relationship between the κ factor and the transmission power.