Delav-aware Cross-laver Optimization For Wireless Sensor Network

Wireless sensor networks get more and more worldwide attention in recent years.Especially with the proliferation of micro-electro-mechanical system (MEMS) technology, smartsensors have been developed rapidly. These sensor nodes can sense, measure and collecteffective information from the surroundings, then transmit the data to the user or other relaysensors after some local decisions. In practical applications, the communication delay is a veryimportant performance metric of QoS evaluation standard in wireless sensor networks. There aremany theoretical studies on the delay aware of wireless sensor networks. Recently, more andmore research efforts to investigate the queueing delay in multi-hop wireless networks. however,still only a few research focus on the worst case delay guarantee. In addition, the sensor nodes inwireless sensor networks are always powered by battery with limited energy, and sometimescharge in the battery can’t meet the need of sensor nodes in some operating. Therefore, it isnecessary to consider both external energy supply and energy harvesting by the sensors.Based on the full understanding of the research on delay aware and energy resourceallocation in the wireless sensor networks, we in-depth study the network throughput and energyresource allocation optimization with the constraints of the worst case delay and stability. Themain work is shown as below:1) Based on the traditional throughput optimization problem in multi-hop wireless sensornetworks, we further consider the worst case delay constraint. We construct a novel virtual queuewhich is used to analyze the worst case delay and the upper bound of the delay. At the same time,we introduce a control parameter to adjust the proportion between the backlog of the virtualdelay queue and that of realistic data queue. Considered in only two kinds of extreme conditions,we give the expression of worst case delay which can be regarded as a function of the controlparameter, and then we attempt to minimize the lower bound of the worst case delay by setting aappropriate value for the control parameter. 2) Besides the worst case delay constraint, we consider a wireless network throughputoptimization problem with the constraint of the end-to-end delay. At the same time, we add thenetwork coding in this model. At last, we transmit the end-to-end delay constraint to the queuestability constraint through Lyapunov optimization technique and solve the problem successfully.3) Based on the study of1), we consider the energy consumption and power allocationproblem. The article considers the heterogeneous energy supplies including renewable energy,electricity grid and mixed energy. At the same time, the article also considers the problem ofpower allocation with interference. The article obtains the convergence of the objective byapplying the Lyapunov optimization technique while the network is stability and achieves a goodperformance in terms of the time average packet dropping.