Mobile service providers are under tremendous pressure to transform, and address unprecedented competitive pressures and service demands.
Operational automation plays a crucial role in this transformation by addressing the growing cost and complexity of mobile networks and by enabling
agile service creation. The demand for operational automation has resulted in the standardization of SON (Self Organizing Networks). SON was initially
applied to LTE and small-cells with Distributed SON (D-SON) architectures. Subsequently the scope of SON has broadened tremendously to support
multi-vendor and multi-technology environments, with sophisticated algorithms operating in Centralized (C-SON) and Hybrid (H-SON) environments.
While C-SON orchestrates D-SON functionality, richer interaction between D-SON and C-SON is contemplated with H-SON architectures. SON is also ripe for
further innovation as operational automation gains market traction. Notable examples include proactive network-centric customer experience management
(CEM), and orchestration for network function virtualization (NFV) and cloud architectures.
A variety of players offer SON solutions. Radio infrastructure vendors including Ericsson, Huawei and Nokia have evolved their LTE based D-SON solutions to include C-SON and H-SON capabilities, and operate in multi-radio and multi-vendor environments. Cisco provides an advanced C-SON solution through its acquisition of Intucell, as does Amdocs, since acquiring the capabilities from Actix and Celcite. SON solutions are also provided by a number of start-ups including Cellwize, P.I. Works, Eden Rock, Reverb, Airhop and Cellmining. These players are capitalizing on opportunities to capture market traction with innovative SON solutions that address specific service provider needs, and deliver sophisticated automation and optimization capabilities.
While service providers generally recognize the importance of SON and operational automation, they are challenged by the disruptive impact it has existing operational processes and procedures. Overcoming this challenge is complex and requires service providers to carefully phase the introduction of automated operations. While this phasing varies depending on circumstances, there are a variety of factors that are important, which include the following:
Although operational automation is inevitable and necessary for the success of the mobile industry, it is human rather than technology factors that will determine its rate and manner of adoption. Service providers and technology vendors must pay careful attention to these human factors as they forge a path towards an automated future.
It is widely documented that mobile service providers are coming under tremendous pressure to transform. It is necessary for service providers to respond proactively to emerging and disruptive market dynamics. Notable examples of this include the following:
Exhibit 1: OTT players differentiate with agility and operational efficiency
Source: Tolaga Research, 2015
Although the transformation drivers for service providers are well documented, the transformation journey they need to follow is scarcely scripted. This is particularly the
case for mobile network operations which tends to be organized in specialist silos with processes and procedures that are focused towards managing
complexity. In contrast, the disruptive operational models being pursued by companies like Amazon and Google emphasize the abstraction of complexity,
iterating deployments quickly and a focus on the customer.
This report investigates the technical, organizational and human factors that impact the transformation and automation of mobile network operations.
Section 2 summarizes the research methodology used in this report, including the primary and secondary research and subsequent analysis that was conducted.
Section 3 provides an overview of current network operations with a focus towards automated radio network optimization, planning, deployment and fault management.
Section 4 reviews SON and the various SON solutions that are available in the market.
Section 5 investigates service provider strategies for fuelling the adoption of operational automation.
Section 6 outlines conclusions and recommendations.
In the coming months, Tolaga will be publishing a web based interactive report that builds on the analysis in this report to investigate and quantify the business case and adoption cycles for operational automation.
This report investigates the following thesis:
“Network operational automation is a core requirement for mobile industry transformation. Although automation technologies and solutions are evolving rapidly, they are hindered primarily by human factors. To overcome these human factors, service providers must carefully phase their automation initiatives.”
The following research was conducted to investigate this thesis:
Pivotal to this report is the notion of operational automation. In this context, the operational functions not only incorporate current activities,
such as network planning and optimization. These functions also span to broader considerations such as agile operations which by their very nature
require customer centricity. Although automation is initially targeted towards existing processes and procedures, it also transforms operations
with an emphasis towards abstracting complexity and eliminating traditional functional silos.
The mobile industry has established operational processes and procedures, which have been developed and refined over a period several decades
to address the core business objectives of a service provider’s operations, see Exhibit 2. The business objectives are determined by a variety
of external factors, including investor expectations, the availability of capital resources, operational legacy and the state of market
competition. The key performance indicators (KPI) which drive network development and optimization strategies are fuelled by a service
provider’s business objectives and normally form the basis for incentivizing operations staff. More recently some service providers have
started to incorporate customer experience management (CEM) considerations in their network development and optimization strategies, as was
discussed in a Tolaga Research report from December 2014 entitled, Operationalizing Network Centric Customer Experience Management. These
CEM initiatives are still relatively nascent and typically inform rather than drive network optimization strategies.
From a radio infrastructure perspective, network development and optimization activities are commonly divided into three categories, namely:
Exhibit 2: Traditional processes and procedures for network operations
Source: Tolaga Research, 2015
While the fundamentals of network operations have essentially remained unchanged since the early days of mobile communications, they have
benefited from tremendous innovation. This is particularly the case when new and more advanced radio technologies have been introduced.
In the early 1990’s, when the author began his telecommunications career, most mobile networks used AMPs or NMT radio technologies, GSM, (which had been developed in the late 1980’s) was seeing its first deployment, Digital AMPs (TDMA) was also just being deployed, CDPD had recently been invented and while CDMA technology showed great promise, it had some still had some technical challenges to iron out. Frequency plans for AMPS networks were sketched out manually with pen and paper and in many respects resembled a game of Sudoku. The frequency plans required that neighbouring cells did not use co-channels or adjacent channels and that the neighbours of neighbours did not use co-channels. The challenge for AMPs was to allocate a little more than 600 channels amongst tens and in some cases hundreds of thousands of mobile users, each producing between ten and twenty mErlangs of busy hour traffic. Excel based exception reports collated information from network element managers to identify poor performing sites and to predict emerging network expansion requirements. Network planning tools operated on Unix workstations and used relatively rudimentary radio propagation models to estimate coverage and interference conditions. Periodically, radio engineers would re-plan the frequencies across each network region, a process that typically took weeks to complete for a network of several hundred sites.
When GSM and CDMA technologies were deployed, radio engineers could no-longer rely on the Sudoku like process that had been used for AMPs. Instead they required advanced network management systems (NMS) and optimization algorithms to support the creation of exception reports and network planning activities. Sophisticated network functionality, such as synthesized frequency hopping, device assisted handover, soft and softer handover, payload scheduling, hierarchical networking and fractional spectrum re-use schemes provided tremendous optimization support directly in the network. In many respects these network functions were the initial steps towards operational automation and were further enhanced with 3G-CDMA, UMTS and HSPA.
When LTE was standardized, the 3GPP recognized that there was a need for automation techniques to deal with growing network complexity. This resulted in the standardization of Self Optimizing Networks (SON) with 3GPP-Release 8, to introduce automated configuration management, optimization and self-healing functionality.
Since its standardization, SON has been widely used for LTE and small-cell (femto-cell) deployments, and more recently for 2G/3G and multi-radio
technology (multi-RAT) optimization. SON solutions are provided by infrastructure vendors like Alcatel-Lucent, Cisco, Ericsson, Huawei and Nokia,
and by a variety of other software and service companies, including Amdocs, Cellmining, Cellwize, Eden Rock, PI Works and Reverb Networks.
The SON standards are widely documented and consist of self-configuration, self-optimization and self-healing functions and can be implemented with distributed, centralized; or a coordinated hybrid of distributed and centralized functionality.
Distributed SON (D-SON) is implemented directly in LTE edge network elements, such as radio base stations. D-SON is provided by network infrastructure vendors and most commonly used for provisioning and configuration management of new sites (i.e. self-establishment). In the case of LTE, D-SON capitalizes on the flat IP architecture used, and leverages X-2 interfaces to enable interaction between e-NodeB radio base stations. To ensure D-SON does not create unwieldy X2 signalling, it only has a localized view encompassing only nearby base stations in the optimization process. As a consequence, it is used primarily for LTE and small-cell base station deployments, and some basic optimization functions such as Automatic Neighbor Relationships (ANR), Plug and Play (PnP), and Physical Cell Identity and Code Allocation. It is also used for functions that demand low latencies, such as inter-cell interference control (ICIC) and short time constant energy saving (ES).
Centralized SON (C-SON) is seeing growing support because it orchestrates automated management, optimization and self-healing functions across network clusters, has access to all network data and historical patterns, and is capable of enriching the network data set with other data such as CEM. When implemented, C-SON servers operate northbound and aggregate the SON interfaces from the network. Many of these interfaces have been standardized by the 3GPP. Solutions based on C-SON have the advantage of a broad network view and the ability to optimize multi-radio and multi-vendor environments and across network clusters. However C-SON has the disadvantage of creating round trip delays that restrict its ability to respond to low latency network optimization demands, such as ICIC and certain use-cases of ES. To overcome these challenges C-SON vendors have been advancing their orchestration capabilities with D-SON. Technology infrastructure vendors, like Huawei have been using Hybrid-SON (H-SON) to distinguish the orchestration processes between C-SON and D-SON. However we believe that to distinguish H-SON from C-SON, H-SON must introduce sophisticated orchestration capabilities between D-SON and C-SON algorithms, so that automated functionality can be implemented with optimal management hierarchies. This orchestration calls for innovation, which we anticipate will create opportunities for SON solution providers to differentiate for many years to come.
SON focuses primarily on those operational functions that are well defined and impactful. Examples of this functionality include:
Since SON is disruptive and has far-reaching implications for traditional network operations, it has the capability to implement processes with
open-loop and supervised or unsupervised closed-loop configurations and incorporate advanced data visualization and roll back capabilities. This
enables service providers to review the impact of SON initiated changes and to roll back the changes if they have a negative effect on network
SON is uniquely positioned at the intersection of data analytics and operational automation at a time when the mobile industry is transforming. This provides SON with significant scope for innovation in terms of:
For most service providers, the transformative potential of SON is daunting and calls for a phased implementation approach. In particular, D-SON
is already widely used for LTE and small-cell deployments, and C-SON has been applied by some service providers for targeted 3G optimization initiatives. However
service providers are more measured when applying SON solutions to broader operational functions; and for good reason. The architectures and algorithms that
underpin SON are complicated and can have a far reaching impact on a service provider’s KPIs. Choosing a SON solution is not merely an exercise of checking that
it has the necessary features. Instead it normally requires extensive field trials to compare the relative performance of different solutions and careful evaluations
to ensure they align sufficiently with the service providers’ specific objectives. In many cases, service providers have implemented solutions from multiple suppliers
to address different automation priorities, and to enable the monitoring of the relative performance of different SON solutions.
A variety of players including network infrastructure and software vendors, and innovative start-up companies are offering SON solutions. Some of the start-up players have been acquired, while others continue to gain market momentum with innovative SON solutions. We expect further acquisitions over the next 24 to 36 months as Tier 1 vendors look to fortify their market positions.
Although SON has its origins with the automation of radio network operations, it provides a basis for automating broader network functions and service capabilities, such as virtualization, cloud and network centric CEM. These service capabilities are the target of many innovations being developed by both start-up and established vendors and will see tremendous advancements in the future.
Radio network infrastructure vendors like Alcatel-Lucent, Ericsson, Huawei and Nokia developed SON as part of their LTE solutions, and have subsequently retro-fitted SON
into their legacy radio network platforms. These vendors have widespread D-SON implementations as a consequence of their LTE
deployment activity and have been expanding their presence with C-SON and H-SON solutions, which generally incorporate both
multi-vendor and multi-radio operating capabilities.
Ericsson, Huawei and Nokia have established the OSSii alliance with the aim of easing inter-operability challenges through bilateral agreements amongst themselves. In March 2015, Reverb Networks also joined the OSSii and established a bilateral relationship with Nokia. While the OSSii provides a means for Ericsson, Huawei, Nokia and Reverb to accelerate their interoperability efforts, the interface specifications are ultimately made available to all SON solution providers. For this reason, we do not believe that OSSii is critical for the success of service providers’ SON initiatives. However OSSii does reduce the need for vendors to reverse engineer competitive solutions and the associated delays that this creates.
Alcatel-Lucent (ALU) has extensive operations and network optimization experience that it draws upon for its SON solution. It has also participated extensively in the development
of the SON standards. However as a consequence of its restructuring efforts, ALU has curbed its investments in SON and radio network centric automation. We believe that this a
necessary step for ALU as it pursues its restructuring process, and that in the future it is likely to fulfil SON solution requirements through partnership arrangements, if and
when required. It is for this reason, we believe that ALU is conspicuously absent from the OSSii initiative.
Ericsson benefits from its dominant position in the mobile infrastructure and service industry. Relative to its peers, Ericsson is taking a measured approach towards SON and operational automation, which is in many respects in lock-step with its service provider customer demands. A SWOT analysis for Ericsson’s SON and operational automation capabilities is summarized in Exhibit 3 below.
Exhibit 3: Ericsson SON and operational automation SWOT Analysis
Source: Tolaga Research, 2015
Huawei benefits from an extensive 2G/3G and LTE network footprint across the globe and a relatively robust Balance Sheet from which it drives competitive strength. As of February 2015 Huawei reported that it had over 130 networks running its D-SON and over 20 networks with C-SON deployed. A SWOT analysis for Huawei’s SON solution and strategies for operational automation is summarized in Exhibit 4.
Exhibit 4: Huawei SON and operational automation SWOT Analysis
Source: Tolaga Research, 2015
Nokia benefits from a strong market position in mobile network infrastructure and is relatively innovative in the field of operational automation with its iSON solution, which is built on its NetAct OSS platform. Nokia has complemented its SON offering with several other innovations, including its Predictive Operations and CEM offerings. A SWOT analysis for Nokia’s SON solution and strategies for operational automation is summarized in Exhibit 5.
Exhibit 5: Nokia SON and operational automation SWOT Analysis
Source: Tolaga Research, 2015
Although Cisco is an infrastructure vendor, it has limited presence in the radio network infrastructure market, other than
with small-cells through its acquisition of Ubiquisys in 2013. Also in 2013, Cisco acquired Intucell and entered the C-SON
market, with a strategy to bolster its radio network competencies. At the time of its purchase, Intucell already had large
contracts with Pelephone in Israel and AT&T in the United States. Cisco integrated Intucell into its Quantum product suite
and has continued to expand its C-SON market footprint, with other service providers including America Movil.
Cisco adopts a unique approach to its SON implementations. Instead of the typical approach of having SON implemented initially in an open-loop configuration, Cisco is insists on closed-loop deployments from the outset, initially in areas) that are less impactful on overall network performance (such as amongst rural cell sites. Once service providers become comfortable with the implementation, closed-loop SON is applied to areas that are more impactful on network performance. We believe that this approach is compelling since it is likely to accelerate a service provider’s use of closed-loop operations and the rate at which service providers evolve their operational models accordingly. A SWOT analysis for Cisco’s SON solution and strategies for operational automation is summarized in Exhibit 6.
Exhibit 6: Cisco SON and operational automation SWOT Analysis
Source: Tolaga Research, 2015
IP Network Competencies:Although Cisco does not provide radio network technology, it is a formidable infrastructure vendor because of its dominant market position with IP network technology. This gives Cisco extensive experience with large and complicated networks and the opportunity to drive SON automation beyond radio network environments. From a business perspective, it already has key relationships as a respected partner with mobile service providers worldwide.
In 2013 Amdocs acquired Actix and in 2014 it acquired Celcite to enter the RAN optimization market. It launched its C-SON solution in 2014, with the combined the capabilities of both Actix and Celcite. Amdocs recognizes the important role that automated RAN optimization plays in CEM, which is an area where it has strength with its business support system (BSS) solutions. It also believes that it has an advantage over traditional infrastructure vendors by being radio technology agnostic, without a commercial interest in selling hardware infrastructure and from having a strong software legacy. A SWOT analysis for Amdocs’ SON solution and strategies for operational automation is summarized in Exhibit 7.
Exhibit 7: Amdocs SON and operational automation SWOT Analysis
Source: Tolaga Research, 2015
The relative performance of the SON solutions offered by Ericsson, Huawei, Nokia, Cisco and Amdocs are summarized in Exhibit 8 using Tolaga’s Market Momentum Index. This index grades and maps the SWOT (strengths, weaknesses, opportunities, threats) analyses shown above for each company according to “Market Position” (i.e. Strengths plus Weaknesses”) and “Market Potential” (i.e. Opportunities plus Threats).
Exhibit 8: Market Momentum Index for SON and operational automation
Source: Tolaga Research, 2015
New Entrant Players Drive SON Innovation
With SON still being in its infancy, it creates opportunities for new entrant players to offer innovative solutions that are targeted towards specific market demands. Notable examples of these innovations include solutions that:
There are variety new entrant companies with C-SON solutions. The companies that are briefly profiled in this report include Cellwize, P I Works, Eden Rock, Reverb Networks, Cell Mining and Airhop.
Cellwize was founded in 2012 and has its headquarters in Singapore. In 2013, its “elastic-SON™” platform was adopted by Cellcom (Israel) and Beeline (Vimpelcom). In 2014 established a multi-national relationship with Telefonica. This relationship was initially to support the operations of the network that Telefonica shares with Vodafone in the UK in a multi-radio and multi-vendor environment. More recently, Cellwize has been supporting Telefonica for other networks in Europe and Latin America. Also in 2014, Cellwize established a partnership with ALU to provide SON for the CloudBand NFV platform, and introduced its “Value-Driven SON™” solution. Cellwize touts its Value-Driven SON™ as having advanced analytics that shifts the focus of the value of the network to the value of the subscriber, while at the same time achieving measurable network performance improvements.
P.I.Works was founded in 2004 and is headquartered in Istanbul, Turkey. Since 2004, P.I.Works has provided solutions to mobile service providers globally. Its P.I.SON platform provides closed-loop SON functions and performance management, reporting and root cause analysis capabilities. P.I.Works has targeted LTE, LTE-A and VoLTE C-SON market opportunities, and has publicized relationships with Tier 1 service providers in the US, Softbank in Japan, “BOLT!” in Indonesia, and Avea and Turkcell in Turkey.
Based on customer endorsements and the variety of market solutions it has deployed, P. I. Works is differentiated by its strong network team and the products that it develops using this experience and expertise.
Eden Rock was founded in 2007 and is headquartered in Bothell, WA, USA. It positions its C-SON solution as a platform that can be customized by its service provider and OEM customers using Python scripting language. In June 2014, Eden Rock announced a relationship with T-Mobile USA as the SON provider for its nationwide network and was selected by the US Government to support research into radio spectrum sharing.
Reverb Networks was founded in 2007 and has been a SON vendor since 2011. It is headquartered in Sterling VA, USA. Reverb has established partnerships with a variety of professional service and systems integration companies to promote its Intellison C-SON solution. Notable partners include LCC/Tech Mahindra, Ascom Network Testing, GET Wireless (North Africa), NVision (Eastern Europe and Russia) and lacroccasolutions (Italy). In 2014 it signed a deal with Mobilicity in Canada and it has publicized a relationship with a Tier 1 service provider in North America. In March 2015, Reverb joined the OSSii initiative to establish a bilateral relationship with Nokia.
Reverb positions its SON solution as being differentiated in terms of its Predictive SON technology for pre-optimization, SON Director for multi-radio management and process coordination and its orchestration between C-SON and D-SON (a.k.a. H-SON). As with other solution providers, Reverb is also driving differentiation through its IP portfolio which covers key SON methods and algorithms and capabilities for cloud SON and self-learning.
Airhop was founded in 2007 and is headquartered in San Diego, CA, USA. Airhop’s “eSON™” product is primarily focused towards real-time interference management and optimization functionality for HetNet and small-cell architectures. Its customers and partners are primarily technology vendors who embed the eSON™ solution in their products. Airhop’s solution is aided by its eSONify™ partnership program to support interoperability and its vSON™ platform to simulate the impact of proposed SON optimization initiatives.
CellMining was founded in 2013 in Caesarea, Israel. Its SONATA SON incorporates advanced subscriber analytics, which it refers to as “Behavior-Based SON”. This allows service providers to analyze and refine their SON optimization initiatives in the context of end-user experiences. Cellmining has publicized contracts with service providers in Israel and Europe
“It has become appallingly obvious that our technology has exceeded our humanity.”- Albert Einstein
Service providers have large teams of specialized employees managing their networks. The roles and responsibilities of these employees have been refined over decades to align with traditional processes and procedures and to ensure that activities are sufficiently coordinated across operational silos. However, service providers cannot sustain these legacy organizational structures, because they are costly and complex, and incapable of delivering the agility needed for future service demands. As a consequence, transformation is inevitable even though it disrupts status quo by eliminating organizational silos and embracing automation.
To succeed, service providers must phase their transformation initiatives, with each phase being carefully crafted and communicated so that it is relevant and actionable for the staff involved. While operational automation is necessary, it is often met with a thorny response from operations staff, who are threatened by its implications. This thorny response can be exacerbated by company executives, when they drive automation initiatives to achieve their own objectives without adequate buy-in from operations.
Exhibit 9: Assessing the transformation drivers for operational automation from the perspective of company executives and operations staff
Source: Tolaga Research, 2015
Exhibit 9 identifies several of the key drivers for operational automation. It identifies existing operational functions, such as mundane and repetitive scripting activities, that can be easily automated. It also identifies changes in network architectures with virtualization, and new business models that require automation. The key drivers described in Exhibit 9 create varied responses from customer executives and operations staff, as is described below.
Service provider employees, and particularly company executives, are well aware of the increasing requirements for business transformation. The notion of transformation, while abstract, is threatening. When it is related directly to an employee’s day-to-day activities, such as automation in the case of network operations, it is generally viewed negatively. For this reason, we do not believe that business transformation should be used as a primary driver to justify operational automation initiatives.
Improvements to network KPIs are important for both company executives and operations staff. However, since the mobile industry is mired with offerings that tout KPIs improvements, SON and other operational automation solutions are subject to significant scrutiny (i.e. the burden of proof). For this reason, service providers generally require extensive field trials before deploying SON and other operational automation solutions.
When automated solutions are deployed, service providers have the ability to introduce new operational priorities, such as those that aim to target network resources to high value subscribers. These operational priorities have the ability to drive automation initiatives, so long as operations staff are incentivized by relevant and actionable KPIs.
The business case for automation is often justified by operational cost savings and that are represented in terms of the number of personnel hours saved. These savings are of tremendous value to company executives, but are naturally interpreted negatively by network operations staff, who fear for their job security. Some countries have laws to restrict the rate at which employment arrangements and organizational structures can be changed, which reduces the rate of cost savings that can be achieved from operational automation.
A more palatable alternative for operations staff is to use automation eliminate mundane operational activities (such as base station neighbor relationship management), so that the staff can concentrate their attention towards higher value activities.
Market competition amongst service providers is a significant driver for innovation in the mobile industry. This competition is fuelling the adoption of SON and will drive further demand for automation as solutions mature and service providers jockey for competitive advantage.
New technologies normally create new operational challenges. For example, LTE management and optimization is complex and different from 2G and 3G technologies, Network Function Virtualization (NFV) requires the orchestration of virtual network resources and small-cells introduce unprecedented scalability challenges. Increasingly, service providers are introducing automated solutions to overcome these challenges. These solutions enable service providers to establish new operational models that incorporate automation and can subsequently be applied to legacy technologies. For example, the automated provisioning, and optimization of LTE and small-cells is now being applied to 2G and 3G technologies by some service providers.
The ability to address resource short-falls with automation is generally compelling for company executives. For network operations staff, the impact generally depends on their perceived entitlement. In particular, if a resource short-fall is a consequence of an unfilled head-count, it is likely that the operations staff would view automation as an unfavorable solution. Alternatively if a resource short-fall is a consequence of the introduction of new technologies such as NFV, LTE or small-cells, operations staff are more likely to embrace automation to address the short-fall.
With changing market dynamics service providers are starting to pay greater attention to the role of network operations in the overall customer experience that they deliver. This is investigated in a research report that Tolaga published in December 2014, entitled “Operationalizing Network-Centric Customer Experience Management”. However as long as service providers continue to maintain historical operational KPIs without introducing targeted key quality indicators (KQI), network operations teams have no incentive to embrace customer centricity in their day-to-day activities. This contrasts OTT players, whose agile service development regimes require that the customer be central to their operations and service innovation.
As service providers map their transformation strategies for operational automation, it is crucial that they introduce a phased approach that corrals support from their network operations staff without creating protracted implementation timelines. This phasing can vary greatly amongst service providers, depending on their organization structures, market priorities and competitive positioning.
Exhibit 10 illustrates a how a service provider’s phased approach towards operational automation might be implemented. Initially automation is targeted towards routine processes for new technologies, such as using D-SON “plug-and-play” for base station deployments. When a service provider implements automated processes for new technologies, the efficacy of automation can be ascertained without directly impacting legacy processes and procedures.
As the benefits of automation become better understood, service providers are increasingly compelled to apply similar automation to legacy network technologies. This is particularly the case if efforts are made to incentivize operations staff to embrace automation.
In the early stages of adopting automation, service providers often use open-loop and supervised closed-loop operations, particularly when applied network optimization initiatives. As confidence increases and the reliability of the algorithms used for automation is verified, more processes and procedures can be converted from open to supervised and unsupervised closed-loop operations.
The KPIs that are used to incentivize operations staff must evolve as service providers advance their automation efforts beyond routine processes and procedures. Ideally these evolved KPIs must be transparent, measurable and repeatable, with the ultimate goal of maximizing a service provider’s return on invested capital with customer-centric network design and optimization philosophies. Examples include KQIs that capture and optimize network resources according to key customer insights and other indicators that measure the impact of operational complexity, such as mean time to deliver optimization initiatives.
Exhibit 10: The evolving role of automation in mobile network operations
Source: Tolaga Research, 2015
Operational automation is inevitable for mobile service providers as they transform in response to competitive market conditions and evolving service demands. While a variety of solutions have been developed under the guise of SON, the path towards automation is far from certain and impacted by a variety of factors. In particular:
When service providers and technology vendors implement SON and automate networks operations, there is no blue-print that they can follow. Instead, they must remain nimble and adapt to changing market dynamics as the mobile industry transforms.;