Research On Target Detection And Cross-Technology Communication In Smart Lighting Scene

Smart lighting utilizes a variety of Internet of Things(IOT)and wireless technologies(e.g.,Wi Fi,Zig Bee)to realize intelligent control of lighting equipment.It is widely used in hospitals and homes lighting due to its superior interaction capabilities,automated control methods,and effective energy-saving modes.The existing smart lighting control system is mainly built on a low-energy Zig Bee platform,based on the active control mode of the user’s handheld device,which brings discomfort and inconvenience to the user.In addition,the Zig Bee protocol is incompatible with wireless communication technologies such as Wi Fi.If expensive proprietary protocol conversion equipment is used,it not only increases the difficulty of system deployment,but also becomes a bottleneck in the system communication process.Therefore,the existing smart lighting system was faced with the problem of the control methods are not humanized,and the interaction of multiple types of devices lacks flexibility.In order to solve the above problems,the purpose of this thesis is to reduce deployment costs,equipment costs,and achieve humane control methods and improve the overall system flexibility.We study the target detection method based on Wi Fi signal and Cross-Technology Communication(CTC)method from Wi Fi to Zig Bee,dedicated to achieve the effect of "turn on the light when people come to,otherwise off" and interconnecting multiple types of equipment.The research results can make the smart lighting system separate from the dependency of the input control information of handheld devices and protocol conversion equipment.The main research contents of this thesis includes:1.We propose Wi Detect,a system which can leverages the changes in Wi Fi signal to accurate detect target.Compared to related work,Wi Detect is deployed in standard Wi Fi equipment,with device-free and no training for target detection.In the smart lighting indoor application scenario,wireless signals are dense and complex,noise rich and random.Both of them are serious challenges in analyzing Wi Fi signals.To solve the above problems,Wi Detect integrates Butterworth low-pass filtering and other algorithms to preprocess the original data,eliminate the influence of environmental noise,and extract target-sensitive data.Therefore,we design a dual-threshold target based on time domain and frequency domain.The detection model effectively avoids the high labor costs associated with training and surveys.In order to meet the needs of low-cost equipment,we implemented Wi Detect with commercial Wi Fi platform and deployed in three real experiment environments(open halls,labs and offices).The experimental results show that the detection accuracy of Wi Detect in different environments can reach at least 90.03%.2.We propose Sense Comm,a low-cost and interference resistant cross-technology communication which can sense the impact of data transmission on the channel RSSI value to realize the direct communication between heterogeneous wireless devices from Wi Fi to Zig Bee,without modifying the firmware or hardware.In the smart lighting indoor application scenario,multiple wireless network protocols coexist,where they compete for the channel and interfere with each other,and enable CTC but have limited decoding performance under interference scenario.Therefore,the communication between heterogeneous wireless network devices is hindered,which poses a challenge for the interaction of multiple types of devices in the smart lighting scene.To solve the above problems,Sense Comm extracts the RSSI of the disturbed channel signal source,using the feature difference rule to decode correctly and the communication between heterogeneous wireless network devices,whichbased on the difference in the disturbance amplitude and duration from the interference source and the signal source.It can effectively solve the problem of the lack of flexibility in the interaction of smart lighting multi-type devices.The prototype of Sense Comm is deployed on commercial Wi Fi and Zig Bee platform for real experimental verification.Our evaluation reveal that Sense Comm can achieves a throughput of 163.4 bps with less than 1.9% Symbol Error Rate in real environment.The results demonstrate that Sense Comm is feasibility and effectiveness.