Research On Efficient Data Collection Algorithms For Wireless Sensor Networks

As one of the important facilities of the Internet of Things framework,Wireless Sensor Networks(WSNs)have extensive development and application prospects in the areas of early warning,monitoring,measurement,etc.The transmission and collec-tion of sensor data as one of the basic functions and tasks in WSNs has always been the focus of researchers' concerns.In the data collection sensor network,the sensor node generates sensor data periodically or intermittently uploads it to the gateway node called SINK by means of multi-hop transmission.Large-scale,dense distribution of sensor networks and limited node energy,communication and computing power pose challenges for routing and transmission optimization design.How to deal with these challenges and design a lightweight and distributed multi-hop transmission optimization algorithm is an urgent problem to be solved in the research of data collection wireless sensor networks.In this thesis,we use the problem of energy limitation as the scenario,by adopting low-power sleep/wake scheduling,renewable energy and SINK mobility we mitigate the energy bottleneck problem and improve network performance.This thesis compre-hensively uses convex optimization and dynamic programming theory to design and an-alyze metrics for different data collection scenarios,optimizing the sensor nodes,SINK nodes and other network facilities,respectively,as well as focusing on lightweight and distributed efficient optimization algorithms.The main work of the study is as follows:1)Aiming at the design defects of energy consumption measurement in oppor-tunistic routing under low duty-cycle sleep/wake scheduling,an energy-saving energy metric is designed to accurately estimate the energy consumption for the transmission for a data packet to the SINK,and an efficient optimal relay candidate set optimiza-tion algorithm is proposed;aiming at the imbalance of existing metric design,a utility metric that combines end-to-end transmission delay,transmission energy consumption and delivery successfully ratio is designed,and an optimal relay candidate set selection algorithm and an approximation algorithm that can achieve the approximate ratio of(1-1/e)are proposed under the model of acknowledgement and none-acknowledgement scenarios,resepectively,with objective of maximizing the value of the proposed utility metric.2)The data collection optimization problem of a mobile SINK aided sensor net-work under energy harvesting scene is studied,and an optimization framework is pro-posed.Firstly,aiming at the uncertainty and instability of harvested energy from natural environment,an energy budget allocation management strategy is proposed,which can dynamically allocate the energy budget for the sensors according to the time-varying en-ergy harvesting rates,avoiding energy shortage and energy waste from occuring.Notic-ing that the traditional mobile data collection path planning methods ignores the data uploading opportunities from the sensors to the SINK while the mobile SINK is trav-eling,a novel data uploading and data collection path planning strategy based on vir-tual aggregation nodes is proposed.Finally,due to the centralized characristics of the proposed optimization problem,a distributed optimization algorithm is proposed.The experimental results show that the proposed optimization framework and algorithm can effectively improve the network performance while maintaining energy balance.3)For the problem of excessive energy loss of omnidirectional antennas in wireless energy transfer,the directional antenna energy transmission is used and the energy and data transmission optimization problems in this scenario are studied.With the goal of optimizing energy efficiency,on one hand,we need to adaptively adjust the directional antenna orientation,transmit power and antenna gain of the energy base station to mini-mize the energy consumption cost.On the other hand,we need to optimize the transmit power of each sensor node,and data generation rate and transmission rates of sensor,in order to maximize throughput while meeting a given quality of service(Qos)require-ment,and construct optimization problems.In order to solve the optimization problem,a heuristic approximation algorithm based on sector selection a rotation-based virtual hybrid antenna optimization strategy is proposed,which can achieve the approximate ratio of(M+1),where M is the number of sectors.The numerical simulation proves that the proposed strategy and the optimization algorithm can significantly improve the network energy efficiency compared with the omnidirectional antenna strategy.