Mobile backhaul: 4G impacts to mobile backhaul

Fri, 07/31/2009 - 8:40pm
Randy Eisenach, market development director, Fujitsu Network Communications

Mobile WiMAX service is now commercially available, and major cellular companies are on the cusp of upgrading to LTE. Service providers looking to perform cellular backhaul for these 4G technologies will find that there are significant differences from today’s 3G networks, and they will have to prepare their networks accordingly.

With 4G networks, simple text messaging and slow e-mail downloads are being replaced by high-speed connections that support true mobile office applications, real-time video, streaming music and other rich multimedia applications. 4G wireless networks will approach the broadband speeds and user experience now provided by traditional DSL and cable modem wireline services.

From the wireless operator’s perspective, 4G systems are vastly more efficient at using valuable wireless spectrum than 2G/3G systems. The spectral efficiency improvements support new high-speed services, as well as larger numbers of users.

The additional speeds and capacity provided by 4G wireless networks put additional strains on mobile backhaul networks and the carriers providing these backhaul services. Not only are the bandwidth requirements much higher, but there is also a fundamental shift from TDM transport in 2G and 3G networks to packet transport in 4G networks. Understanding the impact of 4G on mobile backhaul transport is critical to deploying efficient, cost-effective transport solutions that meet wireless carriers’ expectations for performance, reliability and cost.

wireless capacity requirements
Figure 1: Table of wireless capacity requirements.

Key objectives of 4G networks include supporting higher data rates, improving spectral efficiency, reducing network latency, supporting flexible channel bandwidths and simplifying the network by utilizing an all-packet (Ethernet/IP) architecture.

Although updated in recent years to include Ethernet, historically the 2G/3G wireless standards were based on T1 TDM physical interfaces for interconnection between network nodes. Given the wide availability of T1 copper and T1-fed microwave systems, this was a very logical choice for the physical layer. This traditional reliance on T1 physical interfaces has, up to this point, driven mobile backhaul transport requirements. LTE and WiMAX systems are based on an entirely new packet-based architecture, which includes the use of Ethernet physical interfaces for interconnection between functional elements.

From a mobile backhaul perspective, the major changes are the higher bandwidth required by 4G cell sites, as well as the use of native Ethernet as the physical interface for connection and transport of these services back to the MSC. Given that most cell sites will continue to support GSM 2G and UMTS 3G networks for many years, the addition of LTE and WiMAX means backhaul transport carriers need to implement systems that can support both native T1 TDM services and Ethernet services.

The amount of bandwidth on a wireless network is ultimately constrained by two factors: the spectral efficiency of the wireless interface and the amount of licensed spectrum a carrier owns. Spectral efficiency is measured as the amount of data (bits/s) that can be transmitted for every Hz of spectrum – the higher the number (bits/s/Hz) the better. Newer technologies, such as LTE and WiMAX, use advanced modulation schemes (OFDMA) that support higher spectral efficiencies and higher data rates than 2G and 3G wireless networks.

Figure 2: T1 performance metrics.

Wireless operators spend billions of dollars purchasing spectrum licenses from the FCC, making their licensed spectrum highly valuable assets. The higher spectral efficiencies of 4G allow them to use these assets more efficiently, supporting more users and delivering higher-speed services, resulting in higher revenues for a given amount of spectrum.

The amount of licensed spectrum a carrier owns and can operate over at a cell site is determined by how much spectrum was purchased through an FCC auction. In very simple terms, the maximum amount of bandwidth required at a cell site is simply the amount of licensed spectrum (i.e., channel size) owned by the wireless operator multiplied by the spectral efficiency of the air interface. As an example, the table in Figure 1 illustrates typical cell site bandwidths required for a number of scenarios.

Additional increases in LTE speeds can be achieved by utilizing multiple-input/multiple-output (MIMO) antennas. MIMO is a technique that uses multiple transmit and multiple receive antennas on the LTE radio, as well as on end-user devices (handsets, PC cards). MIMO is essentially able to transmit additional information over parallel air paths.

The higher capacities driven by larger channel sizes and MIMO antennas is somewhat balanced by the fact that data rates vary within a cell. Wireless networks adapt to users at various distances by automatically adjusting their modulation scheme for each user. Those close to the tower operate at the highest data rates, and those furthest from the tower utilize less complex modulation schemes that support lower data rates.

Ethernet overlay model
Figure 3: Ethernet overlay model.

While there is tremendous industry interest and excitement in 4G services and deployment, it’s important not to lose sight of the fact that there are approximately 190,000 cell sites in the U.S., and almost all of them require T1 TDM backhaul. This large installed base of 2G/3G wireless systems will remain in the network for many, many years, resulting in a continued need to provide T1 mobile backhaul transport.

As 4G systems are deployed alongside existing 2G/3G systems, there will be a need to transport both T1 TDM and Ethernet services for mobile backhaul. Several large wireless carriers have specified performance metrics that require their T1 TDM traffic to be carried in native TDM format (see Figure 2), as opposed to Circuit Emulation Services (CES).

Many wireless service providers are uncomfortable with the latency, jitter and efficiency issues related to T1 CES. For these carriers, insistence on carrying TDM services in native TDM format is very understandable based on these performance metrics. It’s critical for these carriers that the backhaul network and 4G LTE migration support both “native” T1 TDM services and Ethernet.

The GSM 2G and UMTS 3G networks that are currently deployed will remain an integral part of the wireless infrastructure for the next 10 to 15 years. At the same time, the introduction of WiMAX systems today, and LTE systems in 2010/2011, will impose new requirements for Ethernet services and Ethernet backhaul. Successfully supporting these legacy systems, as well as the new 4G Ethernet requirements, is critical to mobile backhaul business strategy.

For applications with native T1 TDM requirements, or that have a previously deployed SONET MSPP at the cell site, a TDM/Ethernet model provides a very cost-efficient solution for mobile backhaul.

In this scenario, MSPPs are deployed at cell sites to support native T1s, as well as Ethernet over SONET (EoS). Adding an Ethernet interface card to existing, deployed SONET MSPP nodes provides a simple, cost-effective method for providing connection-oriented Ethernet backhaul services. Newer hybrid Ethernet/SONET systems offer the flexibility of supporting Ethernet services as EoS, or as “native” Ethernet over a separate network connection. These devices provide a seamless transition from a TDM-centric network to a packet-centric network, morphing from a SONET MSPP into a native Ethernet edge platform.

A single node supporting both TDM and Ethernet services is very cost-effective and efficient for mobile backhaul transport. Unfortunately, due to regulatory, administrative or operational restrictions, not all carriers have the freedom to mix TDM and packet services on a common platform or infrastructure. For those carriers, an Ethernet overlay model is the best choice for 4G mobile backhaul.

In an Ethernet overlay model, legacy T1 TDM services for 2G/3G systems are provided by existing methods – either copper-fed T1s or optically fed MSPPs, as shown in Figure 3. As 4G base stations are deployed, a separate Ethernet edge platform is deployed to provide connection-oriented, carrier-grade Ethernet services (100BT, 100FX or GigE).

For true Ethernet greenfield models, where no TDM requirement exists, the model is the same as in the Ethernet overlay model, but without the legacy MSPP. This application might result from a new WiMAX or LTE-only tower location, or from the fact that existing T1 copper, microwave or alternative carrier connections are already meeting legacy T1 transport requirements.

Whatever the circumstances, the carrier requirements are for 4G Ethernet-only transport, and the recommended solution is a combination of an Ethernet edge platform at the cell sites and a Packet Optical Networking Platform (Packet ONP) to provide connection-oriented Ethernet aggregation at the MSC.

The introduction of 4G wireless systems involves fundamental differences in the architecture and backhaul requirements for 4G networks. LTE and WiMAX systems are based on a flatter, all-packet architecture with significantly higher bandwidths than existing 2G and 3G networks. Reliance on Ethernet to provide physical layer connectivity is driving the need for Ethernet services to cell sites for WiMAX today, and in anticipation of LTE deployments in late 2010 or early 2011.

Given the large deployments of existing GSM 2G and UMTS 3G systems, there is a real need to provide mobile backhaul using both T1 TDM circuits and Ethernet services. A single mobile backhaul solution will likely not fit all carriers or all deployment scenarios, resulting in a mix of TDM/Ethernet hybrid, Ethernet overlay and Ethernet greenfield models.



Share This Story

You may login with either your assigned username or your e-mail address.
The password field is case sensitive.