Research On Capacity Analysis And User Residence For LTE-U Networks

In recent years,with the rapid development of mobile smart devices,the number of wireless users increases exponentially and the demand for the wireless communications increases continuously as well.The limited licensed spectrum resources have become unable to meet the needs of wireless users.In this situation,it is proposed as a solution that applying LTE technology to unlicensed spectrum,i.e.,long term evolution unlicensed(LTE-U).By utilizing the advantages of LTE technology,the spectrum utilization of unlicensed spectrum can be improved,so as to share the pressure of licensed spectrum.In this thesis,network capacity and user residence are studied for LTE-U networks.First,the capacity of delay-sensitive traffic for LTE-U networks is investigated.A discrete-time Markov chain(DTMC)model is proposed to model the behavior of the LTE-U network working with round-robin scheduling mechanism.To derive the transition probability of the Markov chain,we take advantage of Lattice-Poisson algorithm.Based on the QoS requirements of the delay-sensitive traffic(maximum tolerable delay and the delay bound violation probability),we find the capacity of the LTE-U networks.Simulation results not only demonstrate that the proposed analytical model can estimate the LTE-U network capacity with high accuracy,but also unearth the impacts of the two key packet-level QoS constraints,maximum tolerant packet delay and delay violation probability,on the network capacity,when coexisting with Wi-Fi networks of different network loads.Second,the performance of channel bonding in LTE-U networks is studied.We propose a throughput based performance analysis approach,which models the primary channel and the secondary channel independently.The channel is divided into four types of states according to its occupancy result.Based on the probability and the duration of different states,we find the occupation ratio of different states and then obtain the throughput of the LTE-U network and that of the Wi-Fi network.Simulation results not only show the effectiveness of the proposed analytical method but also unearth the relationship between the throughput of the LTE-U network and the load of the Wi-Fi network working on the primary and secondary channels.Finally,the user residence and cell selection criteria of LTE-U standalone system are studied.In order to alleviate the pressure of LTE cell,we put forward the idea of giving users priority to reside in an LTE-U cell,that is,to improve the priority of LTE-U cell by allowing users to search the LTE-U cell firstly.In the LTE-U cell search and synchronization process,to avoid prolonging user's residence time due to the non-continuous channel availability of the LTE-U cell,we use a synchronization timer based on the average time interval of the LTE-U base station(BS)'s two channel access chances.In addition,we propose a new cell selection criterion.Cell selection is decided based on the calculation of the remaining effective capacity and remaining throughput of an LTE-U cell,which takes into account both the cell reference signal's received power and the average time interval between the LTE-U BS's two channel access chances.Furthermore,the criterion exploits the useful information in user's historic traffic data,so that the user can choose a more suitable cell,thus improving the probability of successfully initiating traffic,reducing cell reselection frequency and the resource consumption associated with cell reselection.