With the exponential data traffic growth associated with unprecedented emerging bandwidth-hungry network applications and services, the fifth generation (5G) of mobile networks is currently being adopted worldwide, which is targeted to provide significant increased signal transmission capacities, massive machine-type communications (MTC), and ultra-reliable low-latency (URLL) real-time services. The 5G network architecture that aims to support these targets adopts the cloud radio access network (C-RAN) where mobile fronthaul connects remote units (RUs) and virtual baseband units (vBBUs), whilst mobile backhaul connects a pool of vBBUs and data centre. To further increase the signal transmission bandwidth of mobile fronthaul/backhaul links in a cost-effective manner, passive optical networks (PONs) are considered as one of the most important candidates. Moreover, intensity-modulation and direct-detection (IMDD) is preferred in these networks to improve its cost-effectiveness and lower the transceiver architecture complexity. From the signal transmission technique point-of-view, the initial stage of 5G should have sufficient transparency to 4G. Since orthogonal frequency division multiplexing (OFDM) is widely used in 4G, thus OFDM is still a promising signal modulation technique for 5G. As such, this dissertation research aims to explore the feasibility of utilising digital signal processing (DSP)-enabled multiple information-carrying dimensions to improve the performance of optical OFDM (OOFDM) IMDD PON systems in terms of signal transmission capacity, system power budget, transceiver design flexibility and system performance adaptability.
A subcarrier index-power (SIP) information-bearing dimension is introduced into conventional OOFDM by setting the subcarrier power level at either low or high according to an incoming data sequence in order to convey an extra information bit per subcarrier. As a result, a novel signal transmission technique termed subcarrier index-power modulated optical OFDM (SIPM-OOFDM) is proposed for the first time. Compared with conventional OOFDM adopting similar signal modulation formats, this technique offers an increase of 17% in signal bit rate without compromising the minimum required optical signal-to-noise ratio (OSNR) for achieving a specific bit error rate (BER). Moreover, such improvement does not degrade the performance tolerance to both chromatic dispersion and fiber nonlinearity. As the usage efficiency of high power level subcarriers is not fully maximised in SIPM-OOFDM, a technique termed SIPM-OOFDM with superposition multiplexing (SIPM-OOFDM-SPM) is proposed by applying the superposition multiplexing (SPM) operation for high power subcarriers. SPM passively adds different signal modulation format-encoded complex numbers and assigns the sum to a high power subcarrier. As a direct result, compared to SIPM-OOFDM, SIPM-OOFDM-SPM increases the signal bit rate by 28.6% without increasing the signal modulation formats. To further enhance the power usage efficiency of both high and low power subcarriers, an improved version of SIPM-OOFDM-SPM, termed SIPM-OOFDM with dual superposition multiplexing (SIPM-OOFDM-DSPM) is proposed. Compared to SIPM-OOFDM-SPM, SIPM-OOFDM-DSPM increases the signal bit rate by approximately 11% while using lower signal modulation formats. It should be noted that both SIPM-OOFDM-SPM and SIPM-OOFDM-DSPM are capable of improving the system power budget and performance tolerance to both chromatic dispersion and fiber nonlinearity compared to the SIPM-OOFDM technique operating at the same signal bit rate. To further increase the number of information bits conveyed per subcarrier in the above-mentioned techniques, multi-level SIPM-OOFDM (ML-SIPM-OOFDM) is proposed and investigated, in which the number of subcarrier power levels can be increased to a predefined multilevel (ML). As a direct result, compared to SIPM-OOFDM, ML-SIPM-OOFDM improves the signal bit rate by approximately 30%. Moreover, in terms of transceiver design, ML could be applied easily in SIPM-OOFDM, SIPM-OOFDM-SPM and SIPM-OOFDM-DSPM as the ML-associated operating principles, their DSP implementation procedures and corresponding performance advantages are very similar for these transmission techniques.
In all the above outlined signal transmission techniques, each individual subcarrier is regarded as a separate unit to carry extra information bits. To enable a group of subcarriers of various power levels to carry extra information bits, SIPM-OOFDM with subcarrier grouping (SIPM-SG-OOFDM) is proposed, where each symbol is divided into multiple subcarrier groups to bear extra user information bits in the subcarrier group (SG) information-bearing dimension. In addition, SIPM-SG-OOFDM is equipped with an additional capability of automatically detecting and subsequently correcting errors at the receiver without consuming any valuable transmission bandwidth. As a direct result, compared to SIPM-OOFDM, SIPM-SG-OOFDM not only increases the signal bit rate by 11%, but also improves the system power budget by 1.0dB. This implies SIPM-SG-OOFDM can improve performance capacity, adaptability and flexibility. Moreover, the performances can also be further enhanced by combining the ML and SG operating principles in each of the above-mentioned signal transmission techniques which considerably increases the information-carrying dimension.
The above descriptions indicates that, compared to conventional OOFDM employing similar signal modulation formats, the proposed techniques are capable of providing cost-sensitive IMDD PON systems with improved signal transmission capacities and system power budgets. In terms of the transceiver architecture, the proposed techniques still maintain the exact same transceiver design as conventional OOFDM, except that slight modifications in the encoding/decoding DSP elements occur in each of these techniques. These different DSP elements can be implemented in the digital domain in parallel in the transceivers, thus depending upon the traffic requirements and network status, a suitable technique and/or their combination can be selected to improve both the transceiver performance flexibility and adaptability.