Self Organizing Networks, LTE and OPEX
2G to 4G
Mobile users has moved from second generation satisfying their” voice” communication needs to the third generation networks for their voice and data communication needs. Now, with the next generation networks(NGN) like LTE promising to offer a plethora of services at significant speeds to satisfy the users completely, right from voice to streaming multimedia to social networking to online gaming to web surfing on the go. Though these NGNs, promises to open new revenue vistas but the costs associated in deploying and managing these networks will dictate the profitability equation and the confidence from the operator’s end to go ahead with the deployment.
LTE and OPEX
Almost 64 network operators in nearly 31 countries evolving either from HSPA roadmap or CDMA-EVDO roadmap, have committed to deploy LTE in years to come. LTE is expected to provide increased wireless bandwidth , enabling number of new broadband applications and opportunities in mobile space. It aims to provide these services with a higher spectral and network efficiency using a flatter all-IP architecture. But , the key challenge here will be to provide these high quality services in a cost-effective manner meeting CAPEX costs constraints and controlling the associated OPEX costs for the new LTE infrastructure. OPEX is estimated to contribute to almost 30% of the total costs associated with the network over an average network lifecycle. OPEX primarily includes the costs associated with operation and maintenance of the networks in order to maximize the total performance of network which is the key factor in delivering enhanced user experience. This demands for the network operator to monitor, repair, optimize LTE networks and expand this ability to the legacy (2G & 3G) networks preserving seamless backward compatibility. For this level of management, operator will require additional resources which means increased OPEX.
SON – Self Organizing Networks
Thus, in an effort to decrease these infrastructure configuration and O&M associated OPEX costs, Next Generation Mobile Networks (NGMN) Alliance and 3rd Generation Partnership Project (3GPP) have standardized a set of capabilities known as Self-Organizing Networks (SON) in Releases 8 and 9. SON brings in the level of automation in operations and maintenance to minimize the lifecycle cost of running a network by eliminating manual configuration of equipment at the time of deployment, right through to dynamically optimizing radio network performance during operation. SON simplifies network operation and extends the management to the rapidly increasing number of base stations’ configuration and management with least possible human interaction.
The exhibit below depicts the use-cases of SON across the entire network life-cycle from planning to deployment to optimization and maintenance. SON use cases can be grouped into functional domains such as Self Planning, Self-Configuration, Self- Optimizing and Self-Healing as shown below. These SON algorithms are capable of self-configuring. self-optimizing and self-healing automatically and dynamically as per the network and traffic parameters measurements over the entire network lifecycle.
SON Architectural Implementation
The real-time SON functions(algorithms) may be implemented using either centralized or distributed or even hybrid architecture.
Centralized: SON algorithm may reside within the so called Element Management System/Network Management System or a separate SON server controlling the eNBs(base stations)
A centralized architecture approach can be used for deploying real-time SON functions, such as Automatic Neighbor Relations and Automatic Physical Cell ID. In this approach, the EMS/NMS is the key decision maker in the real-time SON functionality. This centralized approach has significant disadvantages -If the NMS is down, no deployments, SON neighbor updates, or SON RF optimization can take place. As the network expands to thousands of eNBs this task becomes more burdensome on the NMS and is also exposed to the operator. Since decision-making is in the NMS, localized convergence is not possible; all data must be forwarded to the NMS, creating burst issues as SON algorithms at the network elements converge to a steady state. NMS-based SON functions also discourage multi-vendor networks and open network element interfaces. Finally, due to the volume of data and computation required at the NMS, the NMS can become a bottleneck for SON changes, introducing significant latency in the commitment of automatic SON changes
Distributed: SON algorithm may reside within one or more network elements (eNBs) SON in a distributed approach allows to push its functionality to the far edges of the network even if the number of eNBs increases. With this approach the eNBs are now equipped with smart and intelligent SON functionality in an autonomous fashion leveraging decision making process as well as reducing complexity, load and footprint issues with NMS. The distributed architecture also enables open interfaces to the network elements by isolating the real-time SON capabilities within the network element. This simplifies support of multi-vendor SON in a single geographic area.
Hybrid: SON algorithm may reside partly in network elements and partly in the NMS. In Hybrid SON, part of the optimization algorithms are executed in the NMS while others are executed in eNB.
All implementations have their advantages and disadvantages and it will be up to the operators to choose the most suitable approach assessing the synergy with the current infrastructure deployment.
Current State:
This concept of SON introduced by 3GPP in Rel-8 specification and advanced features in Rel-9 have been adopted by major network equipment vendor gearing towards LTE deployment for their client network operators. Recently Motorola has announced advanced SON solution which is a part of its LTE offering. Other vendors such as NSN, Ericsson and Huawei are also working intensively on their SON solutions tested.
Summary:
Thus, SON will bring in high level of plug-n-play functionality that will allow the network operators to quickly integrate new network elements and implement configuration changes to the network easily which was not entirely possible in legacy networks. It also means reduced expertise requirement and high-level oversight for the operator . SON in LTE networks will reduce the manual work associated with network operation and maintenance with an ability to self-configure and continuously self-optimize in response to network and traffic changes helping operators to reduce the OPEX , increase operating efficiency delivering higher level of performance through such a robust, energy efficient and intelligent network.
Some Quotes from Industry on SON -
Fabricio Martinez, Consulting Director at Aircom
“Understanding that traffic is growing in a particular area is a given; we can do that. Processing that information and making the right decisions and implementing those decisions back into the network is the problem, Speed heals”, he said. “You have to react in a timely manner,” he said. “There is a degree of intelligence [built into networks] at the moment but it needs to go to a point where this intelligent decision-making happens immediately. If you’re able to gather traffic data in real time and take action I real time to change the network configuration you will optimize the resources of the network.” Fabricio said self-optimized and self-organized networks meet the 4G challenge from two sides: intelligent firmware and external software that is embedded into the networks’ operational systems. “Firmware gives you faster response. Software [which is more flexible and upgradable] will give you more possibilities [to resolve problems] but will be a bit slower than firmware,” he said. Tom Anderson, Senior Director of Wireless Network Architecture and Evolution at Alcatel-Lucent “Certainly there will be a basic set of SON [self-organizing in this instance] things that all LTE deployments will have and others that will fold in over time. It will keep on expanding as we go through the various 3GPP releases to further optimize and fine tune,” he said.Go to the main page
