Latest from 5G/6G & Fixed Wireless Access/Mobile Evolution
Considerations for Synchronization in NGN Packet-Based Mobile Backhaul Networks
Migration from legacy circuit-based to packet-based mobile networks (3G UMTS HSPA and later to 4G LTE) involves significant challenges for network operators. The compelling economics of Ethernet create pressure to move away from time-division multiplexing (TDM) to packet-based transport for wireless backhaul. However, this transition requires careful planning especially where it concerns the timing and synchronization of location-based services, mobility, and voice services.
The radio access network (RAN) backhaul of cell phone conversations and data from the cellular base station to the carrier’s switching point-of-presence is a critical part of the mobile network. (See Figure 1.)
Figure 1. Signal handoff from RAN to Ethernet-based backhaul network.
To achieve successful call signal hand-off over the IP network that connects base stations, and transport real-time data, mobile networks (2G GSM, 2.5G GPRS, 3G UMTS, and 4G LTE/WiMAX) must have extremely accurate clock synchronization.
Base stations must support an accuracy of +/- 50 parts per billion over a 10-year equipment life, as a drift will result in high dropped call rates and impaired voice and data services.
Currently, mobile backhaul is primarily done using a T1 (USA)/E1 (EMEA) synchronous TDM network that delivers the timing information needed for cell tower synchronization. However, today’s mobile backhaul networks face a major obstacle, as traditional TDM expansion is expensive to implement and cannot meet mobile traffic growth requirements. In the next-generation mobile network, Carrier Ethernet services will provide the necessary expanded bandwidth between base stations and controllers.
As mobile backhaul networks are upgraded to Ethernet, the base stations become isolated from the synchronization information that used to be carried over the TDM feeds. While a stand-alone solution such as GPS re-timers can be installed at each base station, this is an expensive solution and is typically not considered viable by most operators due to logistical and political reasons.
Syncing It Up
Today’s 2 leading methods for timing synchronization over Ethernet are ITU-T Synchronous Ethernet (Sync-E) G.8261 and Precision Time Protocol (PTP) IEEE 1588v2.
Sync-E Method
Sync-E is based on physical-layer Ethernet timing similar to SONET/SDH timing. A reference clock is injected by the physical Ethernet interface (PHY) into the Ethernet bit stream, and the PHY of the master port recovers the clock from the bit stream and sends it as a reference clock to all the slave ports. As a Layer 1 scheme, in order to maintain synchronization, all the stations between the master and slave nodes must support Sync-E. This means that even if the existing backhaul network is built as Carrier Ethernet, it must be upgraded to support Sync-E technology on all network elements across the connectivity path.
In Figure 2 we see how Sync-E technology is deployed in the radio access network, and how timing is distributed over the Ethernet physical layer. Here, 3 base stations are connected to a Carrier Ethernet backhaul network by Ethernet demarcation devices that support the Sync-E standard. These demarcation devices receive clock frequency via Carrier Ethernet switched RAN backhaul that is Sync-E enabled and transports the centralized primary reference clock.
Figure 2. Sync-E Network Architecture.
The network can be either Gigabit or 10G Ethernet, but each network attached device must have a Sync-E interface to properly read the physical layer clock signal.
One of the benefits of Sync-E is that its timing quality and consistency are independent of the traffic payload. Built into the Layer 1 transmission, the signals are unaffected by network impairments such as utilization, latency, or loss, thus achieving a purely deterministic and scalable Layer 1 timing solution. However, this technology requires end-to-end Sync-E infrastructure, as each Ethernet link is synchronous by definition. Additionally, Sync-E provides no phase and no time of day (TOD) and therefore requires a complementary protocol in order to support time division duplexing (TDD) networks such as LTE-TDD, Mobile WiMAX, CDMA2000, and WCDMA-TDD.
Precision Time Protocol IEEE1588v2 Method
IEEE1588v2 is similar in its concept to SONET’s network time protocol (NTP). It is a 2-way time transfer protocol with hardware time stamping. The master and slave end-points transport the timing information within MAC frames. In this method, the intermediate nodes do not have to be upgraded, and the advantage is that it delivers time and day synchronization that is required for TDD mobile networks. (See Figure 3.)
Figure 3. IEEE1588 Precision Time Protocol.
In this network, the demarcation devices act as slaves and communicate with short IEEE1588 messages to the centralized primary reference clock (PRC) via Carrier Ethernet asynchronous transport network. This does not require any changes to the physical layer signaling, but does make quality of service (QoS) a critical element of this network because the timing packets must have the highest priority to avoid congestion and impairments in the backhaul network.
The scope of such a backhaul network is limited since a packet-based protocol can suffer from network congestion and packet delay variation. This limitation can be addressed with correct network planning such as a dedicated network engineering path and limited number of hops in order to assure circuit-level quality for synchronization packet flow.
Today’s mobile networks have strict requirements for precise frequency synchronization as well as phase and time synchronization. While Sync-E can provide the synchronization frequency required by 2G and 3G networks, 4G LTE networks also require the phase and time synchronization provided by Precision Time Protocol.
Another Step in the Evolution
Unfortunately, these 2 technologies independently cannot meet all of the needs of today’s evolving mobile networks. However, new networking equipment is emerging that offers both timing technologies. One such option is MRV’s OptiSwitch 904-MBH.
With both timing methods in a demarcation device, carriers can use the same platform to enable timing gateway functionality between synchronous and asynchronous physical layer networks, or to facilitate packet-switched networks without the loss of traditional TDM-quality synchronization to cellular base stations. (See Figure 4.) Such flexibility enables end-to-end synchronization regardless of network type while maximizing investment protection of the existing network infrastructure, a key consideration as 3G networks are upgraded to 4G technologies.
About the Author
Zeev Draer is Vice President Marketing for MRV Communications. He has more than 15 years experience in the telecommunications industry, and is a member of the MEF. For more information, email [email protected] or visit www.mrv.com.