Cross-layer Design And Optimization In Multi-relay Cooperative Communications

There has been explosive growth of smart phones, tablet, intelligent wearable devices and many other wireless mobile devices in recent years, together with the increase of users’requirements on different kinds of services, contents and experience, which have brought many challenges to the next generation mobile communication system. Considering the fact that spectrum is finite, it is a key issue needed to be sovled by future mobile communication networks that how to support high-speed and high-reliability transmissions for cellular networks, wireless sensor networks, Internet of vehicles and so on. From the aspect of network-layer, the heterogeneous cellular network consisting of macro base station and many small cells can improve the frequency reuse factor. However, dense deployment of small cells results in more complex and severe inter-cell interference. From the aspect of physical-layer, the massive multiple input multiple output technique can achieve spatial diversity or multiplexing gain by deploying antenna array at base stations. Considering the fact that it is difficult to pack multiple antennas per terminal due to the limited physical space of wireless devices, relay-based cooperative communication has been proposed as one of the key technologies to provide high-reliability and high-speed uplink transmissions.The core idea of cooperative communications is that by sharing the same time and spectrum resources, the distributed users form a virtual antenna array, achieving spatial diversity. Cooperative communication has many advantages, such as expanding the coverage, providing high-reliability transmissions and decrease inter-cell interference. However, it also has some shortcomings. On the one hand, due to the half-duplex nature of antennas, the original one-hop link has been replaced by multi-hop transmissions, decreasing the data rate and spectral efficiency. On the other hand, multi-hop transmissions means more energy consumptions. The existing cooperative strategies cannot obtain high energy efficiency, especially under the circumstances with energy constraints. Therefore, this dissertation is concentrated on the designs and analysis of cooperative communications, in order to improve spectral and energy efficiency. This dissertation establishes the interference model in the link-layer for the single user two-path relay (TPR) model, and proposes a complex field network coding (CFNC) based cooperative strategy for multi-user TPR model. Furthermore, in the energy harvesting (EH) scenario, a novel link-layer based cooperative strategy is proposed for multi-relay cooperation. All these works present effective solutions to improving the spectral and energy efficiency of cooperative communications with half-duplex antennas.The main works and contributions of this dissertation are as follows.1) Based on the single user TPR system, the link-layer based IRI model is proposed. On the condition that system stability is satisfied, the end-to-end throughput is maximized through power allocation between user and relay nodes.The TPR model utilizes multiple relays to take turns to receive and forward the user’s signal, which successfully improve spectral efficiency, but at the same time leads to inter-relay interference (IRJ). Based on queueing theory, this dissertation proposes a link-layer based IRI model, considering the arrival and service process of data packets at each transmitter. The system stable throughput region is deduced. Besides, an optimization problem is established to maximize the end-to-end throughput via power allocation among the user and relay. Through the inner property of the proposed model, the dimension of optimization variables is reduced, due to which the solution can be obtained. Moreover, when the signal of direct link cannot be ignored at the destination, the dissertation introduces a method to decouple the odd slots and even slots transmissions, each of which can be equally treated as the same scenario where two links share the same time and spectral resources. The stable throughput region of such scenario is also deduced. The simulations results verify the improvement of the link-layer model against the conventional full buffer assumptions. The system performance with different deployments of relays is given to provide a reference for applications.2) The multi-user TPR model is established, and the CFNC-based IRI and inter-user interference suppression strategy is proposed. Besides, symbol error probability (SEP) is further decreased via optimizing precoding factors.The dissertation extends the single user TPR model to a multi-user TPR scenario. Through CFCN-based precoding at users and relays, it can be treated as a distributed jointly modulating system. The receiver adopts maximum likelihood detection to decode the message and suppress the IRI, inter-user interference and the interference of direct link. Subsequently, Union bound and Chernoff bound are used to deduce the tight upper bound of the end-to-end SEP. Based on SEP or mutual information, a selection strategy of decoding methods is proposed at the destination to reduce the decoding complexity. Furthermore, considering the fact that the precoding factors affect the SEP, the dissertation proposes the equiphase precoder (EP) scheme to decrease the SEP, via designing the constellation of distributed jointly modulating. The simulation results verify the coding gain of EP design and the SEP performance gain of the proposed CFNC-based IRI cancellation strategy.3) With energy constraints, a link-layer based multi-relay cooperation strategy is proposed in an EH scenario. In addition, on the condition that system stability is satisfied, the end-to-end throughput is maximized by configuring the cross-layer cooperation parameters.The dissertation proposes a link-layer based three-node decode-and-forward cooperation strategy in an EH scenario, overcoming the shortcomings of conventional physical-layer and link-layer cooperation. The stable throughput region of the proposed model is deduced. Besides, an optimization problem is established to maximize the throughput with energy constraints, via searching the optimal configurations of related parameters. Through the inner property of the proposed model, the original problem can be divided into several simple one-dimensional sub-problems. Subsequently, the single relay model is generalized to multi-relay scenario. The indexing strategy among all the relays is established. Under such circumstance, the stable throughput region and the optimization problem are also presented. A heuristic algorithm is introduced to provide a sub-optimal solution. The simulation results verify the improvement of energy efficiency of the proposed scheme. Meanwhile, the relationship between the number of relay nodes and energy efficiency is given to provide a reference for the deployment and optimization of EH cooperative communications.