DSP-based Real-Time Enabling Technologies for Future Cloud Access Networks

Electronic versions


  • Ehab Al-Rawachy

    Research areas

  • Cloud Access Network(CAN), Optical Communication, Intensity Modulation and Direct Detection(IMDD), digital filter multiple access(DFMA), Digital Signal Processing (DSP), PhD, School of Computer Science and Electronic Engineering


Optical network technologies employed in access networks, metro networks and mobile fronthaul and backhaul must support the unyielding exponential growth in user bandwidth requirements whilst also providing adaptive connectivity solutions to meet the rapid evolution in the diversity of data traffic patterns and characteristics. Furthermore, it is essential that future optical networks support converged fixed and mobile data to leverage the associated cost advantages.. The three distinct 5G services to be supported are, enhanced mobile broadband (eMBB), massive machine type communication (mMTC) and ultra-reliable, low latency communications (uRLLC). As a consequence of the heterogeneous requirements, network slicing is a critical feature for 5G, this allows various services to be provisioned in independent logical channels on the same physical network infrastructure, with different quality of service (QoS) levels. To meet the abovementioned challenges in a cost effective way, the next generation of optical networks have to evolve to be extremely agile, offer highly elastic bandwidth provision and possess a highly reconfigurable network architecture supporting network sliceabilty and software defined networking (SDN), this will allow dynamic adaptation to the changing traffic patterns and enable use of network resources in a highly efficient manner, thus realising commercially viable solutions for network operators. To meet the aforementioned challenges cloud access networks (CANs) have been proposed as a cost effective solution for future optical networks which support converged access/metro networks and mobile fronthaul/backhaul networks. The CANs employ digital signal processing (DSP) to enable various dynamically reconfigurable network devices, network architectures and embedded algorithms for signal impairment mitigation. The fundamental concept of the CAN is built upon digital orthogonal filter-based channel multiplexing, implemented in DSP to achieve highly flexible physical layer connectivity at multi-sub-wavelength levels.
The main objective of this dissertation research is to comprehensively investigate fundamental technical concepts that enable the realisation of the CANs. Using both off-line processing-based experiments and fully real-time DSP-based experiments the technical feasibility of the following CAN technologies and techniques are investigated, i) a cross-channel interference cancellation (CCIC) technique to mitigate physical channel frequency response-induced interference between orthogonal channels, the technique is demonstrated in 2 channel IMDD point-to-point (PTP) SMF-links and implemented with an offline receiver, ii) a real-time, intensity-modulation and direct-detection (IMDD) multipoint-to-point (MPTP)- digital filter multiple access (DFMA) passive optical networks (PONs) with four independent channels, incorporating a real-time ultra-low complexity CCIC technique embedded in a real-time optical line terminal (OLT) in a 26km SMF, IMDD DFMA PON consisting of two real-time ONUs and iii) a fully real-time soft reconfigurable optical add drop multiplexer (soft-ROADM) providing channel add/drop functions for channel switching within the CAN. The soft-ROADM is constructed with commercially-available low-cost electrical/optical components. In summary, this PhD thesis experimentally validates, performs detailed performance analyses and demonstrates the practical feasibility of various fundamental technologies required for the realisation of CANs, which are proposed as a cost effective solution for future optical networks.


Original languageEnglish
Awarding Institution
Thesis sponsors
  • Iraqi Ministry of Higher Education and scientific research
Award date16 Dec 2019