YOU ARE AT:Network InfrastructureReader Forum: LTE and the need for nanosecond-accurate network timing

Reader Forum: LTE and the need for nanosecond-accurate network timing

Editor’s Note: Welcome to our weekly Reader Forum section. In an attempt to broaden our interaction with our readers we have created this forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].

Because LTE has been introduced in a somewhat staggered fashion and the true “4G” as defined by the International Telecommunications Union is yet to come in the form of LTE-Advanced, many wireless operators have failed to notice the limitations inherent in last-generation timing based on TDM lines and GPS satellites. TDM lines can only deliver frequency synchronization while spectrum-efficient Time-Domain (TD)-LTE and LTE-A need frequency and time-of-day synchronization. The difficulty of acquiring GPS signals in urban environments has been obvious for many years. GPS is also susceptible to jamming, with some highly publicized incidents of GPS jamming having occurred recently. Since LTE and LTE-A often are deployed in urban small-cell environments where multiple high-rise buildings prevent line-of-sight fixes on a GPS satellite, it is imperative to find alternative, low cost synchronization solutions for both frequency and time-of-day.

Thanks to the general migration from legacy TDM to IP/Ethernet networks for mobile backhaul and the work by the IEEE, the 1588v2 standard for packet timestamping has arrived in time to solve the problem of delivering nanosecond-accurate frequency and time-of-day synchronization to cost-effective next-generation IP Edge packet networks. The 1588v2 standard builds on Synchronous Ethernet, which brought frequency synchronization to the packet world, but did not provide the necessary time-of-day synchronization required for LTE and LTE-A. Market analysts predict that by 2015, Synchronous Ethernet will be used in 30% of timing solutions, greater than T1/E1 or GPS deployments, but lagging well behind 1588v2, which will be used in 50% of all deployments by 2015. The reality is that both Synchronous Ethernet and 1588v2 combined will be used together to provide the best synchronization results.

(Figure 1: IEEE 1588v2 Precision Timing Protocol estimates the time at the slave by calculating the propagation delay between master/slave via a series of time-stamped messages. Packet networks can have large packet delay variations, introducing timing inaccuracies.)

What makes time stamping tick

In 1588v2, timing information is carried directly in data packets through the process of time stamping. The precision timing protocol, or PTP, uses these timestamps to recover the clock in the slave nodes, such as a base station. However, the packet delay variations that lead to jitter in packet networks also introduce timing inaccuracies in the recovered clock. The PTP protocol further relies on a fixed and symmetric latency between network nodes, which is often not true in real networks. The synchronization error caused by PDV and asymmetry is cumulative across every node in the network in the path between the network node generating the master clock and the base station recovering the clock. This becomes even more challenging when microwave links are used as interconnects, which is the case today in over 60% of all mobile networks according to Infonetics Research. The latency over these links changes dynamically with modulation format as the weather changes, and often the bandwidth and latency can be different in the upstream and downstream directions.

PDV can be addressed in 1588v2 through a defined hierarchy of clocks. The standard defines boundary clocks, as well as transparent clocks. A BC node requires an expensive oscillator, a digital phase-locked loop, and microprocessor, while a TC node can simply reuse the PHY or switch in the datapath as long as the switch and/or PHY can support accurate time stamping and time stamp correction mechanisms. In fact, such a port-based PHY solution is completely sufficient to implement a highly accurate TC node. BC nodes are generally more expensive and complex to implement, but greatly benefit from port-level accurate time stamping to meet LTE and LTE-A synchronization requirements. PHY and switch silicon solutions available today incorporate time stamping correction that compensates not only for network PDV and asymmetry, but also solve the very difficult 1588v2 timing-over-microwave link challenges.

1588v2 timing and small cells – simplicity in diversity

Wireless operators often roll out improvements to infrastructure in a piecemeal fashion. While the conversion from TDM or GPS timing to packet-based timing is required now to satisfy LTE requirements, the migration from macro to metro, micro, and femtocells will take place in an incremental manner, with possible early reliance on BCs, followed by the migration to the lower-cost TCs as LTE services proliferate. TCs can allow for picocell synchronization in an outdoor environment and synchronization down to the femtocell in large indoor multi-floor installations. In the former case, TCs are carried over microwave and millimeter-wave links, meeting TD-LTE and LTE-A specifications while eliminating the requirement for GPS signals or fiber links.

The simpler hardware requirements of TCs fit the model of the small cell base station requiring small footprints, low cost and minimal power consumption. The latter is particularly important for wireless operators with LTE licenses in higher frequency bands (1900 MHz and the 2.x GHz bands) with limited indoor coverage. In such an indoor environment, the access network itself can generate 1588v2 timing, or a GPS antenna on the roof of the building can generate time packets for synchronization services inside the building.

Thus, the capex-constrained operator can move a few key network elements to BC support, followed by efficient network elements reliant on TC clocks to lower cost during volume deployments. This makes it easier to add smaller cells to improve coverage and capacity in a cost effective manner.

ABOUT AUTHOR