Tag Identification And Network Planning In Low Power IoT

The rapid advancement of Internet of Things(IoT)technologies in recent years has made it become essential infrastructure to support automatic sensing,prompt response and scientific decision-making in a broad range of applications in smart city.As a promising future trend,low power IoT,being desirable for low-speed and massive-connection scenarios,consists of sense layer,network layer,service layer and application layer.In the sense layer,tag identification plays a crucial role,since it provides reliable data source for networks;while in the network layer,network planning is an important technology to optimize the coverage and transmission of the nodes.Due to their importance to the overall performance of the whole system,tag identification and network planning in low power IoT have attracted considerable attention from both academia and industry.There is a large volume of research that devotes to tag identification and network planning of low power IoT;nevertheless,further efforts are required to address the following core issues in prior literature:Existing tags in the context of low power IoT cost more than barcodes,and the commonly used tag identification under 13.56MHz is subject to limited communication range,which makes it challenging for large-scale deployment and long-distance detection;Previous ap-proaches are incapable of accommodating highly diverse requirements due to the heterogeneity of the nodes in low power IoT;Existing schemes for resolving the contention where there are mas-sive nodes cannot be directly applied to low power IoT systems.To tackle the abovementioned concerns for tag identification and network planning tasks in low power IoT,this dissertation de-vises low-power,low-cost and efficient strategies which achieve the ultimate goal of boosting IoT system performance.The primary contributions can be summarized as follows:1.A brief review on low power IoT,including the features,the applications and typical tech-nologies are provided.Related work and the challenges are also introduced.2.A chipless tag design paradigm for low power IoT is proposed to reduce the cost of tags.Firstly,the typical features in the antenna mode of the reflected signals are analyzed.Then,inspired by the analysis,circular slotted tags are designed to increase the attenuation of the reflected coefficient and reduce the impact of polarization.Furthermore,this dissertation encodes the tags by reading the peak and trough in the spectrum,and the encoding algorithm is developed based on the different resonate frequencies of the tags.The resultant low-cost system,with conventional RFID tag chip removed though,is able to accommodate over 1024 tags,and meanwhile achieves a remarkable detection accuracy of 95%.Passive relays are designed to push the range limit of tag detection in low power IoT.Firstly,the model of near-field propagation and mutual coupling are built.Then,inspired by the model analysis,a greedy algorithm is put forward to obtain the optimal parameters of the relay coils.In order to mitigate the interference from human body,this dissertation designs Magnetically Coupled Resonant Wireless Power Transfer(MCR-WPT)model.With tunable capacitors integrated,the resonate frequencies of the coils are optimized and bidirectional propagation is established.Experiments show that the communication range is increased by ten times.To the best of our knowledge,the relay device is the first portable passive one to boost the communication range,which is still robust in multipath environment and is compatible with smartphones4.A hybrid gateway planning of unlicensed low power IoT by taking account of the coverage requirements of different nodes is proposed.Firstly,based on the path loss and the data rate at edges,this dissertation analyzes the coverage range of the gateways.Then,gateways which utilize different frequencies for uplink and downlink are deployed to cover the nodes.The considered problem is formulated as a point coverage problem,for which a greedy algorithm is devised to attain the optimal number and the optimal locations of the gateways.In addi-tion,the channel is treated as a Rayleigh channel,and the system throughput and the energy efficiency can be obtained by analyzing the channel model.To the best of our knowledge,this is the first strategy for gateway planning in hybrid networks of unlicensed low pow-er IoT,which mitigates self-interference from full-duplex gateways.Extensive simulations demonstrate that increasing the coverage redundancy effects substantial system performance improvement over one-gateway-coverage network.5.Contention resolution in low power IoT is modeled and analyzed when there are massive nodes competing channels.Queue theory is adopted to model the buffer length of the nodes.After analyzing the random access procedure of low power IoT,Markov Chain is used to describe the state transition of the retransmission times and the queue length of the buffer concurrently.Then,the probability of transmitting a packet successfully and the model of the system throughput can be derived by analyzing the steady distribution of this Markov Chain.Finally,empirical results validate the theoretical analysis,which illuminates the ef-fects of packet generation rate,buffer length,retransmission times and number of nodes on the performance.6.The dissertation is concluded with a research summary and future research outlook.