How to migrate backhaul for 4G

By: Hazim Ahmadi

I wrote this article six months ago as when I was preparing my paper on Mobile Backhaul Asia 2nd in Bangkok, Thailand. So if you find any data that not match anymore with current condition, you may use the last one.  What I want to tell you here are about Indonesian mobile industry data and backhaul issues that should be taken by operators to date. Internet access cost, device price, Internet application, social networking are some issues which covered in this article.

Internet Access Cost Pressure

The method of service charge for data access in mobile data services can be based on minute of usage or on volume consumption. Some operators even offer contractual subscription bundling with notebook PC, modem data card or smart handset. To attract new customer mobile operators are competing to offer price as lowest as possible in either voice or data access services.

Figure 1. Data access tariff in GSM operators

However, high levels of competition are driving down tariff. I think a 5-10% per year fall in overall tariff in the coming years, with data tariff set to decrease still further. Eventually, flat rate data tariffs so familiar to fixed broadband users are adopted by the mobile operators. From this data, now you can imagine how much bandwidth you have to provide then how much dollar your customer will pay for it.

Figure 2. Data access tariff in CDMA operators

This places intense cost pressures on operators who need to ensure that the cost per bit of transporting data falls more quickly than the cost per bit that end users are paying. That means data traffic is going to continually increase.

(Device Price + Service Tariff) vs. Backhaul

Abundant of local and branded of smart mobile phone, modem data card, notebook, PC with low price available in market indirectly drives higher data network occupancy. Data tariff keep falling down as I mention at the previous slide. Today, the backhaul network is under an increasing amount of stress. The main contributing factor is the emergence of ever-more sophisticated, data-centric services that have a much greater impact on the network than the voice services.

These include:

• Web browsing, requiring 9.6 to 240 kb/s

• Media streaming, requiring 128 to 384 kb/s

• Real-time multimedia, requiring 1 to 3 Mb/s

Further, these high-bandwidth, lower data tariff and fast-response applications facilitated by new devices such as the smart phones are changing communication patterns, and the reliable delivery of these services impose backhaul network which operators have today. You must rethink your existing backhaul capability since the traffic will be different it was originally intended to handle.

Backhaul Is Often Neglected

When 3G technology come, and this may also happen with 4G technology, many of us focused on how radio network and handset can be adapted to the new technology, or on the kind of new services and data capacity 3G can support. These are important issues but connecting the cell sites through metro infrastructure to provide for 3G backhaul is often neglected.

Until now operators used various type of backhaul mix in one network for backhaul solution. Wireless backhaul can be an expensive proposition. TDM technology is not scalable especially for huge bandwidth of packet switched supplied to the 3G/4G cell sites. It will be costly to support 4G services.

So in my opinion for next the cost of backhaul shall be a great concern as operators migrate to advanced 4G data services, because the backhaul need per cell determines the effective maximum data load the call can support, whatever the technology might offer.

Driven Factors

Telkomsel says its data traffic has been growing by 178% this year than last year, in some cases because of blackberry, iPhone, smart handset and other type handset sales. You may took a look at my previous slide that Telkomsel spend 1.3 billion USD this year to expand its wireless and broadband networks with better coverage and deploy of HSPA+ service. Facing this situation operator should ensure data access services can be easily accessible to keep end-users happy otherwise they will leave us. The frustrating users will switch operators to find the better one.

Again, I underline operators shall optimize and upgrade their backhaul with right technology. But, we must consider the size of bandwidth with proper traffic demand. Optimization and upgrading backhaul shall be executed immediately. May be some backhaul technologies will co-exist within several years. Because at the meantime there is no one-size fit all backhaul technology. 3G often inherits the initial backhaul solution from an existing GSM network. Several options exist for scaling this up to meet growing data traffic demand and with minimum disruption to end-user services.

Current Situation

Current RAN deployments are typically based on T1/E1 leased lines, copper, microwave or fiber at the physical layer with TDM/PDH as the transmission technique. These solutions have easily managed traffic throughput requirements to date. The rapid growth of 3G high speed services is propelling lease requirement from one E1 per BTS for 2G basic voice services to two E1s per BTS for 2.5G services, up to four E1s and more per BTS/Node B for 3G services and it may not stop there – It is stressing these infrastructures. We need to deliver increasingly sophisticated services and applications that are always available and reasonably priced.

We need to juggle diverse traffic types, keep track of complicated service level agreements (SLAs) and speed time to market—all while reducing capital and operating expense (CAPEX/OPEX) and generating new revenues. Faced with these realities, mobile operators should rethink our approach to the backhaul network: Backhaul optimization, migrating from a mix of legacy transports and multiple leased lines to packet-based technologies and synchronizing the time, so as to support more capacity at much lower cost.

Implementing Bandwidth Optimization Technique

Optimizing the cost of backhaul is now a critical issue for most operators. But before we deploy new transport network, if it is possible we optimize bandwidth consumption by implementing optimization technique. This technique is to reduce the amount of bandwidth needed to support mobile 2G and 3G legacy services. Optimization and compression solutions can be deployed at cell site [BTS, NodeB], at hub/aggregation points in the RAN, at Base Station Controllers (BSC), Radio Network Controller (RNC) and so on.

The first technique is to convert the traffic to common packet. All traffic (2G, 2.5G, 3G, CDMA) is aggregated into traffic packet then statistically multiplexed. Secondly is by employing deep packet inspection technique. The overhead on mobile protocol layers is reduced and unused capacity (such as idle channels) is suppressed. Drop & insert mechanisms take the effective channels used on a BTS E1 interface, drop the unused channels, and then map the channels onto the E1 network link.

Then the last one is by using interface adaptation. In the access domain, interfaces are very often vendor dependent, requiring transparent processing. All optimization and compression equipment must be essentially transparent when introduced in the network. At the end of the day substantial OPEX savings can be gained and proportional revenue can be generated. In addition to those, this technique also off load to introduce new 3G/HSPA services with the same equipment.

Backhaul Network Evolution

From operator’s perspective protecting investment is important things while at the same time reducing cost of OPEX. We cannot replace our entire existing network with the newer infrastructure. We need network evolution. The evolving technology from today’s classic TDM networks to the all-packet network. In order to enable cost-effective capacity growth, traditional backhaul architecture should evolve to a full packet-based network step by step. The aim is a unified network that gives us network simplification, removing complexity from infrastructure and business processes to provide flexibility for changing market conditions.

Figure 3. Backhaul network evolution

However, it is critical that this evolution is performed smoothly to ensure uninterrupted service for end users. For many existing operators, the most cost effective solution will be to evolve to a hybrid backhaul network as an intermediate step to the full packet-based network. The target is not only to optimize the costs within the transport network and provide capacity to meet current demands, but also to provide a foundation for future developments cost-effectively.

Metro-E Urgency for 3G/4G Services

Mobile operators realize that someday data traffic is likely to surpass voice traffic on mobile network. As mobile technology trend evolve to 4G, metro backhaul topology will be more important. IP traffic grows; we must avoid concentrations of IP traffic. If this happens, performance will suffer and service credibility may be impacted. A few of operator has been deploying backbone network and metro Ethernet network. This strategy is to meet the demanding data traffic in either the mobile network or fixed broadband network. Metro infrastructure is far more important for 4G services than for 3G one. Because the expected growth in data traffic volumes and mixture of best-effort and premium content traffic will exist. Unless the metro network is planned as carefully as the radio network, 3G/4G services may not provide revenue and profit benefits expected.

Evolving to packet-based on access network.

Well, off course after we discus much about backhaul, we cannot let our legacy backhaul network without changes. Step by step we have to evolve our existing backhaul to meet emerging greater data demand in future. Fiber optic can substitute the old copper using Passive Optical Network (GPON) in some areas. May be this process will take several years. We are employing (Pseudo-wire Emulation Edge to Edge) PWE3 to provide tunnels to emulate services such as Frame Relay (FR), ATM, Ethernet, TDM SONET/SDH over packet switched networks (IP/MPLS). This protocol connects traditional networks and packet switched networks (PSNs) to share resources and extend networks.

Never Overspend

The current spending on backhaul isn’t so high, and it shouldn’t get too high. Why, because mobile data doesn’t bring in that much revenues. So operators must be careful not to overspend at least until 3G/4G services become mainstream.

It’s true that mobile backhaul spending will increase over the next five years, but it’s not necessarily a lot of money except in a relative sense. Operator shall get infrastructure spending to the extent that it remains a primary profit source, but too much spending will kill profit. I don’t expect to see any relevant increases in backhaul investment until 4G deploys because, until then, circuit-switched backhaul dominates because of 3G voice architecture.

If we add too much capacity, we are just wasting our money, with no monetary gain in return. If we don’t spend enough, our networks might be underpowered, causing to lose customers to other operators. We have to make priorities because there are so many places to spend and upgrade networks. The No. 1 reason to spend on a network is to gain or keep revenue, and the backhaul is a case where if we don’t have a quality backhaul, if we don’t have enough capacity, then our users are going to suffer, and many are not very patient.


It’s true that data traffic is climbing up and stressing infrastructure but voice services still important. Backhaul evolution should be performed smoothly to ensure uninterrupted service for end users. Backhaul deployment challenges for operators are timing & synchronization and hybrid strategy; leave voice traffic on existing TDM backhaul while have voice and data on a single IP/Ethernet backhaul. To maintain profitable operators can implement model voice services with guaranteed performance and broadband services delivered with best-effort performance. It will provide the right balance.

2 thoughts on “How to migrate backhaul for 4G

  1. hazimahmadi says:

    Ok thanks Rizkia, telah mengunjungi blog ini. Semoga isi tulisan membantu teman-teman pecinta telco.

  2. Benar-benar membantu tulisan Bapak dalam hal pengerjaan dan inspirasi Tugas Akhir saya…meskipun di tulis dalam bahasa Inggris…
    ayo pak…gencarkan terus ttg pencitraan teknologi 4G (LTE)..
    Terima Kasih pak…

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s