Ever since the industry’s early days when an army of switchboard operators manually cross-connected calls, many telecommunications advances made have been motivated by automation and productivity gains. Now, as wireless telecommunication migrates from 2G/3G networks to 4G, much of the progress derived from next-generation 4G technologies like Long Term Evolution (LTE) will come from automatic self-optimization techniques that will yield lower operating expenses (OPEX), improved call quality, and better service. Sometimes, the more things change, the more they stay the same.

The concept of self-optimizing networks (SONs) is integral to 4G LTE. In a SON, basestations will be able to automatically interact with each other and with the core network to seamlessly perform a broad range of self-organizing, productivity-enhancing activities. SON is needed to reconcile the industry’s need for higher capacity and faster networks with its desire for lower costs, which in turn brings wireless services to a broader segment of the global economy. Initiatives like the TR 36.902 of the Third Generation Partnership Program (3GPP) and other technical documents are integral to the SON movement. Indeed, it’s expected that LTE Advanced will make significant strides toward an environment made up of SONs.

A lot is at stake for service providers. Variable operating expenses have risen significantly. In recent years, energy consumption in particular has received intense scrutiny from service providers, as it has become a larger portion of overall operating expenses every year. As recently as 2006 at the Base Station Conference, European service provider Orange acknowledged that basestation power consumption had growth of approximately 75% of its network’s total power consumption.

Self-Simplifying

Not surprisingly, wireless networks have grown more and more complex with each new advance and each new generation of technology. Unfortunately, automation within the networks themselves has not kept pace with other improvements.

As the wireless infrastructure has moved away from a voice-only network to become a hybrid voice/data network, the transmission options such as rate of transmission, time of transmission, multi-antenna options, and re-transmissions have added to the complexity of network operations. As a result, the number of employees needed to maintain a network in efficient operation has increased significantly.

Wireless network operations such as the routine pre-operation and in-operation tests to satisfy certain grade-of-service/quality-of service (GoS/QoS) requirements for subscribers have become very labor intense. Network operators typically perform these tests by monitoring the radio frequency (RF) signals in an area or cell by driving vans with radio equipment through the area to collect the needed data. In addition, technicians sitting at consoles in a control center where the network provider’s operations administration and maintenance (OA&M) software is running monitor many operational tasks remotely.

As wireless networks and their equipment migrate toward 4G, the situation will only become more complex. Manual management of these complexities cannot be sustained for long. For instance, there will certainly be a transitional period when 2G, 3G, and 4G technologies will all be present and interacting with each other in the same networks. This sort of heterogeneous environment exacerbates the complexities.

The dynamic operating parameters of basestations will change even more rapidly and frequently in such an environment. For example, 2G, 3G, and 4G basestations must handle handoffs from one cell to the next, call camping, load balancing, and other network operations seamlessly.

Marketplace factors are also accelerating the complexity of wireless networks. Many operators are eyeing new form-factor basestations to better serve their subscribers, expand their services, and retain the loyalty of their users. One new small form-factor basestation technology, femtocells, targets interior installation in residences and small businesses.

Without some of the automation elements of a SON such as a new level of plug-and-play provisioning and self-configuration, operational costs relative to a significant deployment of femtocell basestations would likely be massive. Such a rollout of new and much needed technology could be delayed indefinitely without the productivity gains that SON promises. In fact, if 4G technology suppliers were to fail to provide SON solutions to replace manual operations with automation, the vast potential of 4G technology could be placed in jeopardy.

Component Parts

The elements of a SON fall into three fundamental strata: self-configuration, which occurs before a basestation fully joins the network, and self-optimization and self-healing, which occur dynamically and automatically in an operational SON. The aspects or operational characteristics of a SON that are automatically optimized cut across all three of the strata.

These operational aspects include, but are not limited to, energy savings, interference, random access channel (RACH) success, coverage and capacity maximization, and mobility optimizations. In all likelihood, additional optimization factors will be identified as conditions change and technologies advance.

SON basestations will be able to automatically configure themselves from the moment they are first powered up and before they join a wireless network. Once power is supplied, the basestation would configure its physical cell identity, including its Internet Protocol (IP) address, and it would authenticate its software and configuration data. Following the completion of these baseline tasks, the SON basestation would initialize the configuration of its radio by setting up its relationships with its neighboring cells and compiling its neighbor list.

All basestations would carry out all of these self-configuration processes. These include macrocell basestations, with their extensive range and soaring antenna towers, picocells, which are smaller and have a more limited range than macrocells, and the newer femtocell basestations that will eventually be installed in homes and small businesses. In fact, these auto-configuration process take on even greater importance as a greater variety of basestation types is installed to optimize the network’s coverage and capacity.

Until recently, the mainstay type of basestation for the wireless industry has always been the macrocell. But moving forward, operators will surely deploy the type of basestation that maximizes coverage and service quality indoors and out and, at the same time, minimizes both operational and procurement costs.

The smaller picocells will dot the wireless landscape, providing coverage in nooks and crannies that may not receive proper coverage from larger macrocells. The new femtocell also will provide better coverage inside homes and businesses, which should accelerate the already ongoing trend of wireless services replacing traditional wireline service.

In particular, auto-configuration will be an imperative feature of femtocell basestations if providers hope to control deployment and provisioning costs. A femtocell that can automatically configure its cell’s physical ID and construct its neighbor relation table will be a big step in the direction of plug-and-play capabilities that are needed in SON.

Based on a number of predetermined operational criteria such as energy savings, range requirements, and interference conditions, a SON basestation will begin the self-optimization process once its initial configuration has been completed and it has joined the network. One of the first optimization tasks it will undertake will be to dynamically prune and select the basestations that are on its neighbors list.

In addition, the basestation will begin measuring signals in its environment—its own signals and those emanating from handsets in its range. These measurements will form the basis for the station’s dynamic auto-tuning process, which is intended to optimize the operating state of the basestation at any moment in time. Auto-tuning is a perpetual process that automatically senses spatial and temporal changes in the network and within the basestation’s range, optimizing the station’s predetermined operational criteria.

The operating conditions present in a wireless network are constantly in a state of flux. Some changes, like increased call traffic caused by rush hour congestion along a highway, are predictable. But others, such as intermittent failures in network hardware or software, are not. In any case, a SON basestation must be able to launch self-healing procedures appropriate to the conditions currently being encountered. These self-healing measures would allow the service provider to guarantee a certain grade of service/quality of service to subscribers.

Because of the inter-relatedness of all aspects of a basestation’s operations, self-healing procedures cannot always follow a simple and straightforward decision tree. For instance, the notion of “breathing” or expanding and contracting cell borders juxtaposes a SON’s self-healing and self-optimizing processes. An example of breathing would be when a SON basestation automatically increases its power output and thereby extends its range and its cell’s borders to offload a neighboring cell that is overloaded with call traffic and failing to connect an unacceptably high number of calls.

This is called breathing because the basestation pushes outside its normal borders to assist with the self-healing processes in a neighboring cell. In contrast, the neighboring cell would essentially contract its borders to better serve the many callers who may have become concentrated in a certain limited area within the original cell borders.

This automatic breathing process can be quite effective at handling periods of peak call traffic. For example, rush hour traffic congestion is a routine daily occurrence in most large cities. When people are stuck in rush hour traffic, their first impulse is often to call home or wherever they are going to let someone know they are going to be late. This can quickly saturate the local macrocell basestation. In a SON, additional resources would be automatically deployed to absorb some of the peak call traffic during these periods.

The converse is also true. During the middle of the night when there is very little vehicular traffic on the highway, some of the basestation resources along the highway corridor might be automatically turned off to reduce power consumption for the service provider. Without SON technology, many wireless operators are currently using technicians to manually manipulate basestation energy consumption, channel utilization, and other operating parameters.

What’s To Optimize?

In a SON, operations can be optimized over several general operational metrics like energy consumption, signal interference conditions, coverage, and capacity. But maximizing one metric will often degrade the performance of other metrics. For example, coverage can be maximized at the expense of power consumption. In the real world, optimizing one operational metric often involves a tradeoff with others. As a result, an optimized network is one where the network as a whole is operating as efficiently as possible even though the utmost optimization of some metrics is not being achieved.

Generally speaking, several broad metrics can be applied to a network to determine how efficiently it is operating. As previously mentioned, energy consumption has increased in importance to operators in recent years. In a SON, power consumption will be reduced by automatically powering down cells when their resources are not needed or by reducing the transmit power output of certain basestations when extended range is not called for.

When the transmit power output is reduced for one basestation, the power output for a neighboring cell might be increased slightly to maintain coverage yet reduce the typical amount of power that the two basestations would consume under normal circumstances. Reducing basestation power consumption takes on even greater importance as more and more basestations, including the new form-factor femtocell and picocell basestations, are deployed.

Signal interference is another metric that a SON can automatically mitigate and thereby optimize the operation of the network. Signals coming from other cells in the neighborhood can cause interference. A SON can employ frequency management, beam forming, transmit-power reductions, and other techniques to reduce signal interference.

As 4G infrastructure equipment is deployed, interference will become even more critical. To increase capacity, coverage, and network bandwidth, more basestations will be installed. Some, like femtocells, will be nested inside of larger cells. A femtocell’s signaling must be closely controlled so it does not interfere with the operations of the macrocell within which it is located.

A basestation’s handling of RACH offers another optimization metric. Automatically setting up a SON basestation’s RACH configuration parameters such as the number of preambles on a packet and ramp-up power can reduce synchronization times, call setup times, and handover delays while improving other aspects of RACH performance.

The network’s coverage and capacity also can be optimized when SON basestations can dynamically alter parameters such as antenna tilt and reference power offsets to compensate for lapses in coverage and ensure adequate capacity where it’s needed. Mobility features like handoffs from one cell to the next can be optimized in a SON when basestations can balance load traffic among contiguous cells. SON basestations can monitor parameters like the elapsed time needed for handoffs, radio link failures, and access failures to maximize the number of handoffs that can be processed in any period.

Many other operational metrics can be optimized. Part of the beauty of 4G SON technology will be its adaptability to different service providers’ objectives as well as its ability to quickly respond to rapidly changing dynamic conditions in the network.

Delivering The 4G Promise

As wireless networks migrate toward 4G technology, the industry will be challenged to reduce operating expenses. The added complexity of 4G, if managed manually, will tend to drive operating expenses upward. Also, other variable costs like energy consumption have grown considerably in recent years. Automation, through 4G SON technology, is imperative if 4G networks are to reach their full potential and if the industry is to continue to provide service to a wide segment of the global economy.