TDD pattern estimation and auto-recovery from estimation error for accommodations of fronthaul and secondary services in TDM-PON

Time division multiplexing passive optical networks (TDM-PONs) are attracting attention as a cost-effective way of realizing mobile fronthaul (MFH), which is an optical link between a central unit (CU) and a distributed unit (DU) in a centralized radio access network (C-RAN). To realize an even more cost-effective network, we have studied the accommodation of MFH and secondary services in a single TDM-PON. We focus on the fact that a time division duplex (TDD) scheme will be applied to future radio access (FRA). Since a PON link is a wavelength division multiplexing (WDM) scheme, whereas a radio propagation link is a TDD scheme, unallocated intervals occur periodically in the PON link. For the above reason, we have already proposed a technique for estimating unallocated intervals in a TDD-based MFH link that involves prelearning based on a test signal transmission, and the signal of the secondary service is accommodated in the unallocated intervals. However, a scheme for recovering from an error when the estimation is wrong and the test signal is not transmitted has yet to be studied. This paper proposes an estimation technique without a test signal transmission and an estimation-error recovery scheme. The proposed estimator calculates the correlation coefficient between captured uplink traffic and the seven types of uplink and downlink ratio patterns standardized by the third generation partnership project (3GPP: http://www.3gpp.org/). We evaluate the effectiveness of the proposed estimation technique by using a numerical simulation. The recovery scheme detects the estimation error and re-allocates the bandwidth. We implement the estimator and the recovery scheme in a 10 Gigabit Ethernet PON (10G-EPON) prototype. We report the feasibility of the proposed scheme along with experimental results.