Research Of Key Technologies In Mode-Division Multiplexing Transmission Systems

As the continuous improvement of the degree of the social informatization, the rapid popularity of Internet applications and the increasing multimedia services, the transmission bandwidth demand for communication networks has witnessed an explosive growth, moreover, network traffic is currently approaching the limits of existing transmission technologies. Development of new transmission technology to meet future bandwidth needs of network development has become an urgent task. As the main content of multi-dimensional communication transmission mechanism and networking method, space-division multiplexing has become the remarkable research and hot topics of the forefront of optical fiber communications. Mode-division multiplexing belongs to space-division multiplexing. Using different modes to be the carries of information and using few-mode fiber to be transmission link, mode-division multiplexing transmission systems utilizes multi-input multi-output structure and still has more critical scientific problems to study and solve. For example, the impact of random mode coupling in the few-mode fiber transmission link on transmitted signals and the performance of the system, and the internal mechanism and design of a variety of key equipments such as mode coupler, mode multiplexer/demultiplexer, few-mode optical amplifier and so on. Based on "The National Key Basic Research Development Plan (973Plan)" project "Multi-dimension multiplexing optical fiber communication basic research", this dissertation studies mode-division multiplexing transmission system and its key technologies, and the main content of this research is the MIMO DSP algorithm which is used for compensating mode coupling in the receiver of mode-division multiplexing transmission system.The main research work and innovations of this dissertation are as follows.1. Establish the transmission link model of few-mode fiber with random mode coupling.In the mode division multiplexing transmission system, the few-mode fiber transmission link is crucial. In establishing the transmission link model of few-mode fiber, this dissertation mainly considers the impact of random mode coupling. According to describing changes to field patterns, group delays or other modal properties or not, the author can establish field coupling model or power coupling model respectively. Depending on the different coupling strength, the mode coupling can be divided into strong coupling regime and weak coupling regime. The dissertation utilizes the power coupling model, simulates and processes the strong coupling regime and weak coupling regime separately with algorithm in the receiver.2. Propose and simulate the general model of mode multiplexing transmission system.By collecting, reading and studying a large number of literatures, comparing and summarizing, the author proposes the general model of mode-division multiplexing transmission systems, which consists of transmitter, mode converter, mode multiplexer, few-mode fiber, mode demultiplexer, coherent receiver, digital signal processing module and so on. The author has deep studied the principles to achieve the above devices in the current. On this basis, the author sets up a4×4mode-division multiplexing transmission system which can transmit four modes simultaneously, and each mode carries a56GBaud quadrature phase shift keying(QPSK) signal.3. Propose multi-input multi-output constant modulus algorithm and use it to compensate mode coupling.The dissertation proposes to use multi-input multi-output constant modulus algorithm to compensate mode coupling at the coherent receiver of mode-division multiplexing transmission system for the first time. Extend constant modulus algorithm which was used to compensate polarization mode dispersion in the single-mode fiber to multi-input multi-output constant modulus algorithm, use it to compensate mode coupling in few-mode fiber. With extended multi-input multi-output constant modulus algorithm, the author realized to improve the performance of bit error ratio of the four modes received signals at different levels in different coupling strength. And deep study the impact of the length of filters and step size in the algorithm on algorithm convergence.