| With the rapid popularity of smart devices as well as the Internet of Things in the past decade,the volume of mobile data traffics supported by the communication networks have undergone an exponential growth.In addition,some new application scenarios such as driverless,telemedicine and smart home have emerged quickly,thus demanding for more advanced broadband services that supported by a higher spectral efficiency.Unfortunately,the traditional base-station(BS)-centric architecture for the wireless access networks often leads to a high deployment cost as well as a low flexibility in terms of mobile users'loca access to the network.By employing the device-to-device(D2D)technique,the underlaying cellular networks are capable of not only improving the spectrum utilization,increasing the system capacity and expanding the radio coverage area,but also reducding the traffic load of the cellular networks,decreasing the terminal power consumption,and improving the sum data rate,etc.At present,a critical challenge in terms of resource allocation in D2D communications has been arisen.To address this problem,a variety of studies have been carried out.However,most of the existing research works adopted the single-interference-suppression or mode-selection method,i.e.,without calculating the optimal/maximum number of D2D users(DUs)that can be accommodated by a single link.Therefore,adopting reasonable resource allocation methods is critical to improving the system performance of the D2D-aided underlaying cellular networks.To improve the capacity of the proposed heterogeneous networks,this thesis aims to suppress the mutual interference between D2D and/or cellular links by considering the effects of a variety of aspects,including the number of D2D links,power control,mode selection,etc.The innovations of this thesis are reflected in the following:(1)The mutual interference between multiple links in D2D-aided heterogeneous networks has been analyzed by employing an optimization model based on the number of D2D users.Relying on the proposed model,the trade-off between the increasing interference and the improved performance(i.e.due to the contributions of the D2D links)can be effectively treated.By keeping the sum number of users(including both the conventional cellular users(CUs)and DUs)to be fixed,it was shown that the capacity of the proposed system can be improved by increasing the density of DUs.Furthermore,it was observed that there always exists an optimal/maximum number of DUs for optimizing the sum data rate.For example,when the DUs' density is set to 0.6,the optimal number corresponding to the maximum data rate and the maximum(tolerable)number of users is shown to be 12 and 34,respectively.(2)The effect on system performance in terms of sum data rate was evaluated by using the method of interference limited(ILA)area,which can be implemented by relying on the location of DUs.The closed-form expressions for both the coverage probability and the sum data rate of both cellular and D2D links under the proposed ILA-based mechanism are derived,followed by analyzing the influence of key parameters such as signal-to-interference-plus-noise ratio(SINR)threshold and inter-D2D-peer-distance.By measuring the quality of service(QoS)of the communication links(i.e.by adopting geographical location),the system can determine whether to satisfy the requirement of a D2D-link-creation or not.ICompared with the existing classical algorithms that based on interference to signal ratio(ISR)and interference over thermal noise(IoTN)by using Monte Carlo simulations,the closed-form expressions for the sum data rate of the the proposed ILA-based scheme have been given out,greatly reducing the computational complexity.Furthermore,the sum data rate is shown to be increased by about 83.3%and 86.4%respectively,for dBR=50m,or be increased by about 10.5%and 8.8%,respectively,for dBR=500m.(3)The problems concerning D2D communication mode selection and link multiplexing in the multi-channel multi-user scenarios have been resolved.In particular,two regional division methods are proposed for the purpose of identifying both the optimal relay node in the D2D-relay mode and the optimal multiplexing link.The cost function can be established,aiming for optimizing the resource allocation mechanism in the proposed heterogeneous networks.Compared with the area division(AD)mechanism,the candidate number of nodes as well as cellular links can be decreased by 41%and 76%,respectively.Furthermore,compared with random selection from all users(RS),random selection in range division(RAS)and AD mechanisms,the sum data rate of the proposed system can be increased by about 71%,41%and 10%,respectively. |
PDF Full Text Request