Performance Analysis And Optimization For Device-to-Device Communication System

With the rapid development of mobile communication technology, information sharing between users becomes more common and the data traffics generated between adjacent users are growing. Consequently, the local characteristics of the traffics become more and more obviously. However, there is a lack of relevant mechanisms and techniques to take advantage of the local characteristics in the conventional cellular systems. Different from the information processing method in the conventional cellular network, Device-to-Device(D2D) communication enable two proximity users to transmit signal directly without going through the base stations. D2 D communication has been proposed as a key feature to be supported by the 5th generation mobile communication systems(5G) due to the potential of increaseing spectrum efficiency, saving user equipment power, offloading cellular traffic, and enhancing system reliability. Futhermore, with the adoption of D2 D communication underlaying cellular networks, there are large numbers of mobile devices which can be used as relays to help other users communicate with each other. Therefore, the relay-assisted D2 D communication underlaying cellular networks has attracted considerable attentions for extending cellular coverage, enhancing the quality of service and improving the system throughput. In this dissertation, the interference management of D2 D communication, and the performance performance analysis and optimization for the relay-assisted D2 D communication system are investigated. The main contributions are summarized as follows:1) The interference problem for D2 D communication underlaying cellular networks is investigated. Firstly, considering the case in which one D2 D pair and cellular user share the downlink resources, an interference limited area of cellular user is built and the radius of the area is deduced under the cellular user interference constraint. A D2 D pair access strategy is proposed based on the interference limited area. Furthermore, the outage probability of cellular user is analyzed and simulated. The results demonstrate that the outage probability performance can be improved greatly when the proposed D2 D pair access strategy is performed. Secondly, considering the case in which cellular user and multi-pair D2 D exploit the same uplink resources, the closed-form expression of outage probability of the cellular system is deduced under the constraint of multi-pair D2 D link aggregate interference which is modeled as a weighted chi-squared random variable. Based on numerical simulation, the effect of the number of D2 D pair, the position of D2 D user and related network parameters upon outage probability performance are investigated. Simulation results indicate that the interferences received by BS increase with the increase of numbers of D2 D link. The results also show that the overall system capacity can be improved with the increase of D2 D link when the distance constraint between D2 D link and the BS is met.2) The system performance and optimal power allocation scheme for Amplify-and-Forward(AF) Two-Way relaying D2 D communication systems are investigated. Firstly, for the D2 D communication with three-time-slot fixed-gain AF two-way relay protocol, the approximate expression of outage probability is deduced considering a scenario with co-channel interferences and noises at all terminals. Furthermore, a power allocation optimum method to minimize the outage probability is developed, and the closed form expression for the optimal power allocation coefficient at the relay is derived. Secondly, for three-time-slot variable-gain AF two-way relaying D2 D communication system where interferers and noises affect all nodes, the lower bound of the outage probability and upper bound of the ergodic capacity are deduced based on the tight upper bounds of the equivalent signal-to-interference-plus-noise ratio(SINR). A power allocation optimum method to maximize the ergodic capacity is developed, and the closed form expression for the optimal power allocation coefficient at the relay is derived. Simulation results demonstrate accuracy of the derived outage probability and ergodic capacity expressions. They also denote that the system performance can be improved using the proposed optimal power allocation method.3) The relay selection schemes for Decode-and-Forward(DF) two-way multi-node relay D2 D communication system are investigaed. Considering a general scenario with co-channel interferences and noises at the two source nodes and all relay nodes, the probability for the number of relays that can decode the two sources' message successfully given L relaying candidates is derived utilizing order statistics. Then, according to the derived result, the closed-form expressions of the outage probability for max-min relay selection, random relay selection and direct transmission scheme are deduced considering that all the channels are independent and not identically distributed Rayleigh faded. Numerical results are provided, sustained by simulations which corroborate the exactness of the theoretical analysis. Based on the simulation, the effects of the number of relays, the position of relay nodes, power allocation coefficient and related network parameters on outage probability performance are investigated. Simulation results also indicate that the performance of D2 D communication system can be improved under the relay selection scheme and the max–min relay selection scheme achieve better performance than other scheme.4) The relay selection schemes for AF two-way multi-node relay D2 D communication system, considering transmissions over independent and not identically distributed Rayleigh fading channels, are investigated. The closed-form expressions for the cumulative distribution function of the tight upper bound of the SINRs are presented. Then the lower bound of the outage probability for max-min relay selection and random relay selection scheme are deduced in the presence of co-channel interferences and noises at all nodes. Simulation results demonstrate that the derived analytical results fit well with Monte-Carlo simulations. Based on the simulation, the effects of the number of relays, the position of relay nodes, power allocation coefficient and related network parameters on outage probability performance are investigated. Simulation results show that the system performance can be improved under the relay selection scheme and the max–min relay selection scheme always achieve lower outage probability when compared with random relay selection scheme.