Software-Defined Networking-based Adaptive Resource Allocation in Optical Networks

Abstract

Programmable optical transmission based on software-defined networking (SDN) is being considered as a promising solution for the fifth-generation (5G) cellular networks, data center networks, and passive optical networks (PONs). In particular, the stringent capacity and latency requirements of the fronthaul segment call for innovations in the design of new architectures and transceivers that can adapt themselves to highly variable application traffic patterns and physical-layer conditions. We propose a software-defined pseudorandom sequence-based synchronization scheme, which enables us to in-band adapt the modulation type and code rate per connection. Based on this scheme, we further propose a cross-layer resource allocation framework for hybrid mobile fronthaul (MFH) networks that can support dynamic processing resource sharing and comprises both coherent detection (CD) and direct detection (DD) transceivers. We develop new modulation schemes to support simultaneous data delivery to both CD and DD terminals. Using theoretical analysis and two separate SDN-enabled downlink and uplink transmission testbeds, we demonstrate the advantages of adaptive transmission in MFH networks as well as the benefits of probabilistically shaped three-level pulse amplitude modulation (PAM-3) and five-point quadrature amplitude modulation (QAM-5) for multicast transmission in hybrid networking scenarios. As a step-forward towards the real-time implementation of proposed concepts, based on a high-speed FPGA, we design and implement a generic adaptive shortened and punctured irregular low-density parity check (LDPC) code emulation system using layered-decoding with iterative scaled min-sum algorithm. We evaluate the hard-decision, soft-decision, and burst-error performance of the irregular LDPC code that is under consideration for ITU’s 50G-PON standard and demonstrate high potential of properly designed irregular LDPC coding for PON, data-center, and 5G+ applications.