Optical transport is widely viewed as a necessary component of fifth-generation (5G) mobile access, chiefly due to the large bandwidth requirements. Edge optical transport in support of 5G presents numerous challenges, including absolute latency, latency variation (i.e., jitter), operation and cost; fiber scarcity in some metro areas is another limiting reality. This paper clearly enumerates these challenges and then addresses the problem on numerous fronts, taking advantage of standardized technologies to present practical solutions. The solutions span both the physical and networking layers. In the first approach, a wavelength-division multiplexed (WDM) solution is advocated between the distributed radio units (RUs) and the more centralized shared equipment, rather than an individual fiber per RU. The scheme is tunable, which avoids the operational limitations of a fixed-wavelength solution, but is implemented with a standardized mechanism that avoids the need for costly wavelength locking and calibration componentry at the RUs. The details of the mechanism and the associated remotely-tunable transceiver technology are presented. Using a low-cost polymer waveguide chip for the transceiver, experimental results indicate that tuning accuracy can be maintained over a wide temperature range. The second approach, based on gray optics rather than WDM, utilizes high-speed Ethernet (e.g., 100 Gb/s) to aggregate the signals of multiple RUs. This achieves capacity benefits through statistical multiplexing while taking advantage of low-cost Ethernet components. The use of the O-band is advocated, to avoid limitations due to chromatic dispersion. Experimental results indicate the feasibility of using either discrete multi-tone (DMT) or four-level pulse amplitude modulation (PAM-4). At the networking layer, the paper addresses the requirement for time-sensitive networking with stable low-latency packet delivery by taking advantage of intelligent frame pre-emption strategies. Excellent latency and jitter results are demonstrated for a two-level traffic priority scheme. In addition to these proposals, the paper speculates on the role passive optical networks (PONs) may play in future 5G access.
You must log in
to add comments.