Mobile Netwk Opt for 3G and 4G Netwks

ABIresearch

®

Research Report

Mobile Network Optimization Performance Monitoring and Optimization for 3G and 4G Networks

Joe Madden Analyst Aditya Kaul Practice Director, Mobile Networks

NEW YORK

LONDON

SINGAPORE

SCOTTSDALE

ABIresearch

®

Mobile Network Optimization

Section 1. EXECUTIVE SUMMARY 1.1

Introduction In mobile communications, the act of “monitoring” and “optimizing” the network has historically referred to technicians visiting cell sites with test equipment, checking RF power levels and antennas, and performing drive tests. At the network operations center, wireless carriers have performed this act by watching parameters such as the dropped call rate and bit error rate/frame error rate to keep track of statistical performance. Throughout the 1990s and even into the early 2000s, the most important parameters were related to coverage and voice. The arrival of 3G created a longer list of key process indicators (KPIs) to track. Yet, operators generally focused their attention on voice-related metrics because most of their revenue and profit comes from voice services in 3G. Recently, operators have been surprised by the huge increase in mobile data, especially from Internet-friendly smartphones such as the Apple iPhone. As a result, they are beginning to emphasize data efficiency and signaling metrics in 3G and are focusing much more heavily on data-related metrics for LTE and WiMAX. Monitoring and optimization solutions are following this change in direction. Internet solutions such as deep packet inspection (DPI) are suddenly in high demand from mobile operators. Transport efficiency and radio signaling are becoming important metrics, surpassing the previous simple focus on dropped call rates and voice call handovers.

1.2

Market Drivers The rise of mobile data is the most compelling market driver today, and multiple aspects of data demand are driving a variety of “bottlenecks” in the mobile network, including: •

Streaming video content is flooding the radio channel and transport equipment.

•

Client applications on the smartphone are constantly enabling radios, checking some data, and then logging off again. When millions of these smartphones are authenticating radio channels every five minutes, the signaling and TCP (Transmission Control Protocol) transport setup is overwhelmed in a 3G network.

•

Gaming applications demand ultra-low latency, which is not always met with 3G networks.

•

Other bottlenecks are appearing – in almost every 3G network element.

End users are beginning to notice the quality of service issues resulting from the bottlenecks. Voice calls are increasingly dropped on some networks, and latency issues on others are creating noticeable delay for gaming and even VoIP applications. Both uplink and downlink throughput are degraded by the constant signaling traffic on the network, thus reducing user-perceived throughput dramatically. In short, end users are now beginning to drive network improvements through their complaints and churn to new networks. What’s different about this? Nowadays, subscribers are moving to a new operator in order to get better data coverage – not necessarily better voice coverage.

© 20q0ABI Research • abiresearch.com The material contained herein is for the individual use of the purchasing Licensee and may not be distributed to any other person or entity by such Licensee including, without limitation, to persons within the same corporate or other entity as such Licensee, without the express written permission of Licensor.

2

ABIresearch

®


Table 1.1

Mobile Network Monitoring & Optimization Equipment Revenue, World Market, Forecast: 2009 to 2015 ($ Millions) Segment UE Clients Radio Test Radio Probes Routing/Transport OSS DPI Network Offloading

2009 3.0 200.0 99.0 193.0 378.0 65.0 2.0

2010 3.6 208.0 108.9 231.6 453.6 162.5 4.0

2011 7.2 216.3 119.8 277.9 544.3 325.0 14.0

2012 14.4 225.0 125.8 347.4 653.2 552.5 36.4

2013 21.6 225.0 132.1 416.9 783.8 828.8 69.2

2014 32.4 225.0 138.7 500.3 940.6 1,077.4 103.7

Total

940.0

1,172.2

1,504.6

1,954.6

2,477.3

3,018.0

2015 48.6 225.0 145.6 600.3 1,128.7 1,292.9 155.6

CAGR (09-15) 59.1% 2.0% 6.6% 20.8% 20.0% 64.6% 106.6%

3,596.6

25.1%

(Source: ABI Research)

To address bottlenecks throughout the mobile network chain, operators are investing in multiple monitoring and optimization solutions, ranging from client applications on the terminal to DPI solutions in the core network. The operators are investing in two directions:

1.3

•

The network equipment manufacturers are providing more OSS software and self-organizing network (SON) features in both 3G and LTE networks.

•

Independent suppliers of monitoring and optimization solutions are finding strong growth in direct sales to operators. Overall, the independent market for various monitoring and optimization solutions will grow at CAGRs of between 2% and 106% over the next five years.

Technology Network operators are trying to balance the SON networks promised by their network OEMs with several independent monitoring and optimization ideas scattered through the network. In general, the SON use cases currently focus on radio parameters and self-configuration. SON functionality is expected to expand into true optimization of additional network layers in the future. Independent monitoring/optimization solutions include: •

Client-based optimization solutions with software on the mobile terminal, typically to improve efficiency in the transport layer

•

Portable RAN test equipment, which is used to verify compliance to standards and optimize the coverage/capacity of the radio layer at each cell site

•

Fixed probes and related software for the radio layer to track call parameters and handovers to isolate the root cause of radio access and mobility issues

•

Protocol analyzers and transport optimization hardware, installed in the core network to streamline packet sessions and improve on transport costs

•

Operations support system (OSS) software, which aggregates performance data, inventory information, and fault information to speed up root cause analysis

•

DPI infrastructure, which examines each packet of data to categorize it for application content

•

Offloading solutions, which can divert Internet traffic to save bandwidth in media gateways and Gateway GPRS Support Node (GGSN) infrastructure

© 2010 ABI Research • abiresearch.com The material contained herein is for the individual use of the purchasing Licensee and may not be distributed to any other person or entity by such Licensee including, without limitation, to persons within the same corporate or other entity as such Licensee, without the express written permission of Licensor.

3

ABIresearch

®


No one single solution is a “silver bullet” that will boost the capability of a network. Each solution can provide between 15% and 50% improvement in efficiency for one network element, but multiple bottlenecks exist simultaneously. Several of these solutions must be used together in order to optimize all of the links in the chain. Thus, it is critical for network operators to have a comprehensive, interoperable strategy.

1.4

Outlook The growth of monitoring/optimization solutions is accelerating. These solutions will grow at 24% overall during the next five years. The most advanced data networks in North America, Europe, and Japan are the most important growth markets in the near term, as these large networks currently have insufficient data capacity to handle anticipated demand over the next few years.

Chart 1.1

Mobile Network Monitoring & Optimization Equipment Revenue, World Market, Forecast: 2009 to 2015 4,000 3,500

($ Millions)

3,000 2,500 2,000 1,500 1,000 500 0 2009

2010

2011

2012

2013

2014

2015 (Source: ABI Research)

1.5

Recommendations The unanticipated rush to the mobile Internet is happening now, and time is critical for operators, network OEMs, and independent monitoring/optimization suppliers. The companies that can work in a coordinated way in teams to attack multiple bottlenecks will achieve the greatest network improvements, and thus will capture market share. Network monitoring and optimization suppliers should ensure that they have focused solutions for 3G networks that are future-proof for use with LTE in the near future. Large suppliers capable of handling multiple monitoring solutions should strive to cover end-to-end requirements and integrate probes, analyzers, and OSS software. Smaller, specialist suppliers should partner with other vendors in order to bring simple, highly targeted solutions to market together with a comprehensive portfolio.


4


1.6

ABIresearch

®

Conclusions A shift in focus has long been expected for mobile networks, and it has been predicted that LTE/WiMAX networks will address data issues. In reality, the data tsunami has hit 3G networks before the deployment of LTE, forcing operators to invest early in solutions to better diagnose and optimize for data applications. As LTE networks are rolled out, the market for network optimization will grow even stronger.


5

ABIresearch

®


Section 2. MARKET OVERVIEW Today’s complex mobile networks simply cannot be managed by an army of technicians in blue jeans that drive to site locations and climb up towers to tweak the system. The techniques and processes used to optimize an analog or 2G network would require a huge amount of manpower in order to support 3G and 4G systems. Instead, the mobile network must be managed statistically and automatically, with the technicians reserved for key interventions that are triggered by statistical analyses of the system. It is possible to operate a voice network without sophisticated monitoring and optimization tools. In fact, many 3G networks are currently in operation with what is essentially an extension of the old 2G dependence on manual intervention. This approach can work reasonably well for voice traffic, but data traffic impacts the network quite differently, and ignoring the differences can result in poor OPEX efficiency. AT&T Wireless’ famous problems with backhaul and signaling in its 3G network are a prime example. AT&T has been surprised by the sheer weight of iPhone data traffic, as over 55% of its data traffic comes from less than 5% of its subscriber base. This issue has lingered for over a year because the solution involves more than simple increases in backhaul capacity. The iPhones, as well as Android and many other smartphones, continuously shut down their radios to save power, then re-establish the data link for a quick data update. The result: a heavy load of signaling traffic. This unprecedented rise of signaling traffic from iPhones has created a bottleneck in the routing and radio signaling channels, which had basically been designed to handle normal voice and SMS traffic throughout the AT&T network. Chart 2.1

Mobile Traffic by Application, World Market, 2009 (TB per Month)

Web Brow sing, 30,242 Video, 35,897

VOIP, 4,579

Gaming, 4,615

Peer-to-Peer, 15,496



6

ABIresearch

®


Chart 2.2

Mobile Traffic by Application, World Market, Forecast: 2014 (TB per Month)

Web Brow sing, 621,610

VOIP, 156,829

Gaming, 173,177

Video, 2,336,732 Peer-to-Peer, 276,952


This problem is not unique to AT&T. Every carrier faces a similar loss of network quality if smartphone, tablet, and mobile PC web traffic is allowed to continue growing without a corresponding increase in network sophistication. The bottlenecks are manifesting themselves in multiple locations throughout the network, from the user equipment (UE) to the radio access network (RAN) to the core network.

2.1

Networks Evolve into Optimization Networks tend to mature along a predictable timeline, beginning with simple RF planning and provisioning of network elements and moving toward an automated, optimized mobile network where business processes are tuned in to the network, optimizing both cost and revenue. Most 2G networks followed the process to monitoring/optimization with simple and manual interventions for optimization. 2G business processes are extremely simple, covering some pricing differences for time of day, total number of minutes, or SMS messages. 3G networks have also evolved along the timeline to the point of implementing optimization solutions. In this case, simple manual interventions are not always possible due to the high level of complexity. Interventions such as neighbor list updates or power level adjustments can be achieved remotely. In 3G, tiered billing structures to take advantage of end-user data preferences are uncommon, often due to the lack of information available to operators at a detailed level regarding the content downloaded by each user.


7


ABIresearch

®

LTE networks will evolve even more quickly along the timeline, with self-configuration features taking a major role in the deployment of the network and optimization algorithms built in from the beginning. Business processes still lag behind the optimization features available for network performance, but the rise in consumer data consumption will strongly drive the development of more innovative pricing policies. Figure 2.1 Typical Timeline for LTE Network Optimization


2.2

Key Performance Indicators KPIs for 2G systems were normally tracked on a short list of five to six items, including simple voice-related metrics such as: •

Call drop rate

•

Handover success rate

•

Bit error rate (or frame error rate)

For 3G networks, the list of KPIs grew to include a few data metrics and the complexity of handovers between multiple air interfaces, but the primary focus for most operators remained on voice performance: •

Call setup success rate

•

Packet Data Protocol (PDP) context activation success rate

•

SMS/MMS success rate

•

Call drop rate

•

PDP context drop rate

•

Soft handover success rate

•

Hard handover success rate

•

Inter-radio access technology handover success rate

•

Call setup delay


8


ABIresearch


®

9


ABIresearch

®

In LTE, 3GPP defines KPIs in five categories with a clear focus on data quality of service. The 3GPP organization has been much more careful in the case of LTE to define KPIs with standard definitions, including precise mathematical definitions of the KPI calculation. In doing so, 3GPP hopes to address the confusion that has resulted from 3G KPIs defined differently by each vendor, thus improving interoperability at an equipment level and communication at a human level. The five LTE KPI definitions are: •

Accessibility: Measured by the E-UTRAN radio access bearer setup rate

•

Retainability: Measured by the radio access bearer drops per unit of time

•

Integrity: Measured by downlink throughput and latency

•

Mobility: Measured by handover success rate

•

Availability: Measured as the fraction of total time that a cell is unavailable

Most operators use additional KPIs to address cost optimization targets as well, and a wide variety of preferences are used by operators around the world. Both CAPEX and OPEX are tracked by operators in calculations of parameters such as spectral efficiency, throughput per backhaul fiber or microwave link, and even energy efficiency.

2.3

R&D/Testing/Deploying the Network The ecosystem for test equipment in mobile networks has matured over the past twenty-five years, with established vendors such as Agilent, Anritsu, Rohde & Schwartz, and Tektronix filling most niche areas. Test equipment is distinct from the market for monitoring and optimization gear. The test equipment is generally designed to be portable and temporary so it can be used for setting performance on a given network element and then taken away to another site. The market’s evolution to 4G does not change the equation for radio test gear very much. Radio test equipment will continue to be required for drive testing and adjustments to the radio layer, which cannot be achieved through software and automated controls alone. The difference between 3G and 4G will be seen in the manpower required to deploy, configure, and maintain the network. While technicians will always need a spectrum analyzer within reach, SONs will hopefully reduce the number of times that they need to start up their truck and drive to the cell site. SON technology will result in automatic programming of cell site parameters such as neighbor lists and default frequencies, and power levels will increasingly be set through coordinated data from neighboring eNodeB sites.

2.4

Monitoring the Network So far, SON committees related to the 3GPP standards have focused primarily on use cases that impact Layers 1 and 2, which are most closely tied to the radio access network, as well as maintenance functions such as alarm management, inventory, and QoS optimization within LTE only. This leaves an open field for optimization tools to augment Layers 3 and 4 with routing and transport setup. In particular, suppliers of optimization tools are able to provide signaling analyzers, probes, and OSS software that span multiple network vendors and can deal with the multi-technology complexity of a modern mobile network. Real-time monitoring is achieved through the use of probes at multiple interfaces throughout the network, including both 3G and 4G probes at the interfaces to NodeB and eNodeB sites, at SGSN and GGSN locations, and at the LTE Mobility Management Entity (MME) and serving gateway. Through the collection of statistical data on retransmissions, error rates, and signaling traffic, the network monitor can synthesize a summary of the performance of a complex, multivendor network in which the network equipment manufacturers (NEMs) do not coordinate data with each other.


10


ABIresearch

®

In the 100-year history of telecom networks, an ongoing market for OSS has developed slowly and gradually. Almost any OSS on the mobile market can track basic KPIs such as dropped call statistics and handovers between different vendors. However, operators complain that the level of detailed information available in the OSS for troubleshooting is inadequate. A typical 3G network today includes an OSS provided along with the mobile network, as well as a parallel system of probes throughout the network, in order to isolate the signaling issues and identify the root cause of traffic issues. One of the primary functions of protocol analysis tools and related probes throughout the network lies in the need for operators to quickly diagnose and solve problems. Today’s probes and monitoring systems provide more specific data than a typical OSS, isolating problems through analysis of the signaling between network elements on standard interfaces.

2.5

Optimizing Performance Self-optimizing networks (as distinct from self-organizing networks) are still in their infancy. They will remain immature for years because the focus of major NEMs centers on self-configuration of network radio elements, not on the fine-tuning of optimization. Furthermore, while most major OEMs support the collection of data from multivendor network elements, none of the major OEMs can offer true optimization of a multivendor network. In short, interfaces are well standardized, but intervention algorithms are not. When the operator has visibility into the data patterns on the network, questions quickly arise: •

What can we do to improve data throughput efficiency?

•

Can we reduce CAPEX by avoiding future equipment upgrades?

•

How can we reduce OPEX by eliminating site visits and reducing time spent by technicians?

•

What adjustments will most directly improve quality of service?

In the area of optimization, network operators are focused on three basic concepts: cost reduction, quality improvement, and revenue increase. Cost reductions and quality improvements are the two areas that can be impacted through changes to network settings, since revenue is generally independent of network performance.

2.5.1

Cost Reductions One of the most direct and natural outcomes of improved network monitoring is that the existing equipment can be tuned to be optimally efficient, resulting in a deferral of upgrades and capacity overhauls. In addition, operating costs can be reduced significantly if issues can be resolved quickly with information available in a central database – without a truck.

2.5.1.1

CAPEX Reduction Inefficiency caused by retransmissions, interference, or poorly set radio parameters can be eliminated with effective monitoring and timely changes and upgrades, which allow the infrastructure to run closer to its rated capacity. By deferring upgrades and new line cards, multiple vendors believe that they can create between 15% and 30% capital savings for a particular network element. Assuming the low end of 15% applies to key core network elements such as SGSN and the RAN in 3G, the annual CAPEX budget for network capacity is significant. As an example, for a European operator with 30 million subscribers, the €800 million budget for CAPEX might be reduced by €120 million.


11


ABIresearch

®

Similar figures apply to LTE deployment, since the system will be initially installed to handle limited capacity and will be stretched to accommodate users as LTE services are adopted. The savings are difficult to distinguish in hypothetical comparisons to an “unoptimized” LTE network, but vendors and network operators agree that significant savings are built into their plans for LTE optimization. 2.5.1.2

OPEX Reduction Operating costs constitute another significant area of cost savings. The self-configuration of network elements saves technician time and the number of drive tests required, and means fewer trips to the NodeB or eNodeB site. Monitoring tools with probes installed throughout the network can save a great deal of technician time. The tools provide the ability to troubleshoot problems instantly from the network operations center, so the technicians do not need to drive to multiple locations. In addition, streamlining data packets can save a surprising amount of bandwidth in the transport layer, improving throughput efficiency by roughly 15% to 25%. Harnessing transport optimization tools to take advantage of 15% efficiency improvement can translate directly into reduced transport cost, which averages roughly 30% of total operating cost for mobile operators worldwide. In the United States, backhaul cost reduction is particularly important, since as much as 50% of operating costs is spent on fiber and leased lines. The OPEX savings from a suite of optimization solutions can be significant. As an example, for a regional American mobile network with 6 million subscribers:

2.5.2

•

The savings in technician labor and associated trucks/support can be in the range of $30 million per year, or roughly 5% of the total operating cost budget.

•

Transport efficiency improvement of 15% can result in cost savings of $24 million to $28 million per year in transmission costs.

Quality Improvement While some KPIs are useful in cost efficiency optimization, others are more useful in tracking quality of service and quality of experience for the end user. QoS metrics have been changing quickly during the recent explosion of mobile data because the experience of a VoIP user is very different than a streaming video user or web browsing user. As a result, the tools for monitoring and optimizing QoS are evolving to examine a greater number of KPIs and the definition of “success” has moved well beyond a simple view of dropped calls. As an example, for VoIP and game users, latency is key but throughput may not be critical. Conversely, for streaming media, the throughput efficiency is critical but latency can be absorbed through buffering of the media. Therefore, these two users require different KPI profiles for optimized QoE. True QoS metrics are increasingly focused on parameters that refer to the efficiency of the network itself, such as packet retransmissions and overall throughput. Vodafone has proposed the establishment of different grades of service (GoS) in order to separate different users into virtual networks. This process would allow the operator to track different KPIs for different grades of service and even optimize each group’s traffic for their particular usage pattern.


12


2.6

ABIresearch

®

Optimizing Revenue The final – and arguably the most important – step in optimizing the network will come in the form of pricing plans tailored to extract more money from the end user. The mobile operator finds itself in a strange dilemma today, spending 70% or more of capital investment on the data network and deriving 75% of revenue from voice traffic. Further investment in data networking capability must eventually be supported by a commensurate increase in revenue. As with cable TV in the 1980s, mobile networks will be moving away from flat-price data plans and toward a variety of pricing tiers. Technology is now coming into place to make new business schemes feasible. For example, DPI tools can identify the content in each IP packet, triggering the OSS/BSS (Operations Support System/Business Support System) to enforce policies on whether individual users may access each content type. Coming out of the Internet market, DPI tools are capable of high capacity, ranging up to 80 Gbps. However, network operators have yet to deploy DPI widely enough to implement new pricing plans. A complete suite of DPI boxes (to accompany every GGSN on the network) is necessary before an operator can implement a content-based pricing structure. Simpler protocol analyzers can identify the source of content as a captive server, an Internet server, or another mobile user to establish some coarse data categories. Many network operators feel that these solutions are more suitable for troubleshooting and diagnostics; they are not detailed enough to utilize as a basis for complex pricing plans. The potential for increased revenue remains untapped in the market. ABI Research expects operators to continue driving their networks toward better monitoring and diagnostic tools so that they can implement policies based on content. During the past year, Verizon Wireless has hinted that its strategy calls for migration to multi-tiered billing structures. Over the next five years, we expect several major operators to gradually move pricing plans toward tiered, content-driven plans tailored to end-user preferences.

2.7

Test Equipment Market Mobile networks have antennas and cables mounted in exposed, outdoor locations. Thus, they will always need RF test equipment and trained technicians to check antenna position, power output, spectral emissions, and other RF performance parameters. The equipment is growing in sophistication as LTE signals are added and analyzers have additional signals to demodulate and new parameters such as error vector magnitude (EVM) to monitor in the RF chain. In addition, with the increase in the number of frequency bands used worldwide, passive intermodulation (PIM) has become an important test parameter. In the past, intermodulation from passive components was considered negligible, but dual-band antenna and radio front-end systems make it possible for loose connectors or dirty interfaces to create interference for mobile signals. Special PIM testers have emerged as a new item on the market to address the loss of capacity and performance that accompanies PIM. Overall, vendors supporting the test equipment market, such as Agilent, Anritsu, Rohde & Schwarz, and Tektronix, occupy a relatively stable capital equipment market that is currently entering a phase of supplying deployment/monitoring gear for LTE. The rise of LTE is resulting in a surprisingly smooth transition from legacy spectrum analyzers and power meters to newer models with additional functionality.


13


ABIresearch

®

Network operators procure portable RF test equipment as a part of their comprehensive monitoring and maintenance plan. In addition to protocol analyzers, OSS systems, and the other diagnostic tools discussed in this report, the operator must periodically send a technician into the field to measure radio parameters, perform a drive test, adjust the position of an antenna, or replace failed components. Figure 2.2 An Example of a Portable Spectrum Analyzer

(Source: Anritsu Co)

2.8

Probes While portable RF test equipment is sometimes necessary, most problems can be diagnosed immediately through information available from the network elements themselves. A monitoring system that uses probes installed at interfaces to RNC, SGSN, GGSN, MME, and LTE gateways can provide information for almost instant diagnosis of simple problems. Probes are generally passive data-collection devices that can be inserted at the interfaces of a network without affecting communications between the network elements. They can range from 19-inch rack-mounted boxes to standardized (ATCA) racks with multiple probe cards. The most common configurations involve probes in the 3G mobile core network, with a recent trend to add more probes at 3G and LTE radio network interfaces, as well. Probes report data up to a centralized aggregation point, and are sold along with software to view a summary of network data. They also typically include Ethernet ports so that a technician can plug in a laptop to view a probe’s data output locally.


14


ABIresearch

®

Figure 2.3 An Example of an ATCA Rack-Mounted Network Probe

(Source: Tektronix Communications)

The market for network monitoring probes and related software/transport optimization includes more than $300 million in routing and transport layers, typically located at interfaces to SGSN and GGSN equipment, and $160 million in probes for the radio and Layer 2 signaling, typically located near the RNC equipment in 3G networks. When LTE systems roll out, an interesting level of competition will arise between probe vendors and network equipment manufacturers such as Ericsson, Huawei, NSN, and Alcatel-Lucent. As the network OEMs introduce more self-organizing and self-optimizing networks, the network diagnostic capability will expand and the network’s ability to automate a response will also improve. Operators have mixed responses to this trend: •

Operators that want to maintain flexible and manual control over the network are reluctant to turn over control to an automated system. Thus, in many cases, the OEMs will limit initial SON features to self-configuration aspects. In these cases, the probe vendors will continue to see growth in the number of units sold and complexity/capacity of the probes, driving an ongoing growth market.

•

Operators that want integrated data collection (especially operators that have experienced problems with multivendor interoperability) see the network OEM as the best way to summarize diagnostic data. These operators are more likely to rely on diagnostic elements built into the network and summarized in the OSS provided by their primary supplier.


15


ABIresearch

®

Overall, probe vendors such as Astellia, Agilent, Radcom, and Tektronix are filling the holes left in modern networks by providing valuable missing information in a simplified and aggregated central console. Due to the reluctance of operators to give up control, ABI Research does not expect to see these vendors step too deeply into automation of network responses. Nevertheless, these vendors will continue to increase the scope of their probe offerings to accommodate LTE networks. They will also gradually offer higher sophistication in the analysis of problems and recommended response by operators. Figure 2.4 Network Monitoring Solutions Applied to Different OSI Layers


2.9

Protocol Analyzers To fully understand the root causes of limitations in the network, technicians must be able to duplicate the loading conditions that stress the network, and then visualize the impact on Layers 1 through 4. While network probes generally focus on the operation of Layers 1 and 2, protocol analyzers are typically brought in to investigate issues that span multiple communication layers, tracing a call or data session to determine what bottlenecks are affecting performance. In many cases, the protocol analyzer and related equipment can simulate heavy loading conditions to emulate the conditions during a recent failure. As such, protocol analyzers are a key link between performance monitoring systems and optimization processes for the network operator. Analyzers can be portable test equipment or stationary installations in rack-mounted enclosures. Each subsystem includes the data collection/capture electronics, as well as a hard disk drive for storage of data session signaling information. Rack-mounted protocol analyzers normally allow multiple users to log on simultaneously. Emulators are sometimes used to simulate heavy loads of network traffic (late at night) so that testing a more optimal solution can take place without impacting too many users. Companies such as Tektronix Communications, EXFO, Astellia, and Agilent Technologies provide protocol analyzers with diagnostic capability supplied through a graphic user interface at the network operations center. Some companies, such as Tellabs, EXFO, and Spirent, step further into optimization of the transport layer through active intervention in session setup.


16


ABIresearch

®

Figure 2.5 Use of a Protocol Analyzer to Monitor Network Messaging

(Source: Radcom)

2.10

Impact of Managed Services on Optimization Roughly half of Nokia Siemens Networks’ revenue comes from services, while 38% of Ericsson’s revenue is derived from mobile operator services. Ericsson has made some high-profile announcements of its management of major networks on behalf of major wireless service providers such as Sprint Nextel and Bharti. These cases illustrate the undeniable trend toward the outsourcing of network management toward the major OEMs. How will this affect the market for network monitoring and optimization gear? It may be too early to draw any final conclusions to this question, as two-thirds of network operators still manage their own equipment. However, the dynamics of selling independent solutions into the changing network management community will shift as OEMs increasingly call the shots. Network equipment manufacturers are increasingly gathering revenue through services contracts, with companies like In cases where the OEM manages and optimizes the network, in theory the optimization is, well, optimal. However, large network equipment suppliers are often more reluctant to engage with independent suppliers of probes, DPI equipment, and other handy tools from competing outside suppliers. In the end, too much reliance on in-house solutions can leave the network without the best tools for optimizing each communications layer. Over time, we can expect the simple, radio-centric SON algorithms offered by major OEMs to become more sophisticated for services customers. DPI, offloading, and other high-layer capability will be added to the simpler radio algorithms. If the trend toward services outsourcing continues as planned, the leverage applied by the major OEMs will put limits on market growth for independent suppliers.


17

ABIresearch

®


2.11

Operations Support Systems The multibillion-dollar market for OSS software spans wired telephony, cable networks, mobile networks, satellite, and Internet systems. Service providers and even large enterprises employ OSS to monitor and manage their network elements over a growing breadth of network complexity. In the mobile world, the OSS is responsible for organizing information on the status of network elements, including inventory and relationships between network elements, and aggregating performance information reported by each network component. Many network deployments include sales of both network elements and an OSS. Thus, the “captive” or bundled market for OSS is difficult to separate from the sales of network equipment. Bundled OSS from the network equipment manufacturers comprise roughly 65% of total 3G NodeB sites globally.

Chart 2.3

Captive OSS Market vs. Merchant OSS Market, World Market, 2010

Merchant Market, 35%

Captive Market, 65%


The served market for mobile OSS software is at $400 million per year and growing steadily, due to the difficulty for any one company to cover all of the possibilities in complex multivendor, multi-generation systems. Both network OEMs and independent suppliers can access the standardized data from network interfaces, so differentiation takes place in the speed and flexibility of customization and in the ease-of-use for the applications using OSS information. Independent OSS vendors tend to move more quickly than NEMs to adapt to changes desired by the network operators. Independent vendors such as Agilent, Anritsu, and Tektronix also offer more flexibility in OSS integration with business systems. Smaller players such as Mycom offer visibility in simple software that aggregates information from multivendor networks so that operators can visualize the entire network in one place.


18


ABIresearch

®

OSS typically consist of the software itself, which aggregates and displays information for the operator. In most cases, the OSS provided by a major network OEM is connected to its own network elements, resulting in islands of information. Many of the independent OSS solutions sold today are packaged together with network probes or protocol analyzers in order to create a more complete picture of the entire multivendor network. The OSS does not typically interfere with any systems because the process is oriented around information gathering, rather than controlling network elements. As LTE and WiMAX networks are overlaid on top of existing 2G/3G networks, ABI Research expects the OSS software to need even more flexibility and capability in order to gather multivendor data. LTE/WiMAX metrics are a bit different than the 2G and 3G KPIs gathered in some of today’s systems, so a significant trend in the LTE/WiMAX OSS market will involve software upgrades to integrate new parameters. Figure 2.6 OSS Architecture and Applications

(Source: Nokia Siemens Networks)


19

ABIresearch

®


Section 3. TECHNOLOGY In the 1990s, cell technicians would endlessly drive around and watch the received signal strength indicator (RSSI) on a special handset. They would then, in many cases, climb up on a rooftop or even onto a tower to change the tilt of an antenna, hoping to solve many different issues with this manual, mechanical intervention. Those days are over. The complexity of 3G and 4G networks is making this simplistic type of radio optimization much less effective, and thus manual intervention is a smaller part of the overall picture. Instead of simple antenna tilting and other tweaks to the radio layer, network optimization today involves changes to Layers 2 through 7 using a wide variety of tools and techniques. Tools available to network operators include the following:

Table 3.1

•

Portable test equipment for initial network deployment is used periodically to monitor and maintain the radio layer.

•

Stationary probes can be installed at almost every interface in the mobile network, collecting data from base station and NodeB sites, SGSNs, GGSNs, RNCs, and media gateways.

•

Protocol analyzers are often used to monitor the traffic and statistical data such as retransmission rates, error rates, throughput, and other KPIs in Layers 2 through 4.

•

The traffic itself can be monitored through deep packet inspection (DPI) tools, which examine the data packets for signatures that indicate what type of application is used.

•

Operations support systems (OSS) are implemented to centralize the collection and control of multiple elements within the radio access and core networks.

•

In all cases, optimization algorithms using the monitoring data collected can be implemented to change radio, routing, or application parameters.

•

Monitoring data can be reported to the network operator to initiate a business action, such as suggesting a new data/pricing plan to the customer.

Applicability and Impact of Various Monitoring Solutions

Solution Type UE Clients Radio Test Radio Probes Protocol/Signaling Analyzers OSS DPI Network Offloading

Equipment Location UE NodeB and eNodeB sites NodeB, eNodeB, RNC sites SGSN, GGSN, MME, S-GW Software at a NOC server GGSN, S-GW RNC/SGSN interface

Primary OSI Layer Addressed Layers 5-7 Layer 1 Layers 1-2

Example of Impact End-to-end optimization possible Improved spectral efficiency, QoS Improved handover, QoS

Layers 3-4

Improved transport efficiency

Layers 1-7 Layers 5-7 Layers 3-7

Improved troubleshooting time Improved traffic prioritization Clears capacity for other traffic (Source: ABI Research)

Almost every major mobile network today involves a mixture of 2G and 3G, as well as a multivendor environment in which elements of the mobile network, core network, and backhaul come from competing vendors. As operators begin to deploy LTE, making an overall monitoring and optimization solution work requires an active and organized strategy.


20


ABIresearch

®

Overall, many network operators plan to continue monitoring and optimizing 2G and 3G networks with their existing, fairly manual processes and tools while relying on SONs to save manpower and effort in the LTE overlay network. The SON approach does not solve many problems, as SON use cases are primarily aimed at radio-related issues in early deployment. In addition to SON, many network operators will deploy DPI solutions or other end-to-end monitoring solutions to oversee network performance.

3.1

Self-Organizing Networks Self-organizing networks (also often known as self-optimizing networks) encompass a vast range of use cases in which the network itself takes care of tasks that previously had been handled manually by the RF planning and maintenance technicians. In practice, SONs can save time in multiple different areas – from RF planning to inventory management.

Figure 3.1 Self-Organizing Network Features

(Source: Mobile Experts)

3.1.1

Self-Configuring Networks The first use cases considered and implemented in 3GPP clearly rank among the most important in terms of operational cost savings. In LTE networks, setup and configuration of a new eNodeB relies on the new network element to configure itself for frequency selection, populate the list of neighboring cells, and set optimal parameters for handoff to/from the neighboring cells. As it recognizes the new eNodeB, the network will also automatically update the neighbor lists for the other eNodeB elements and trigger the re-optimization of their respective handoff parameters.


21


ABIresearch

®

Femtocell implementation makes this functionality crucial, since femtocells are intended to be installed in the field by end users instead of trained network technicians. In a femtocell, the typical functionality of a NodeB and a radio network controller are rolled into a single low-cost consumer box. As a result, the configuration algorithms must take into account both the radio parameters typically set for the NodeB and the handoff and transport setup that is typically achieved in the RNC. Typical LTE or WiMAX use cases for self-configuring networks include:

3.1.2

•

Installation and radio provisioning of each eNodeB

•

Automatic neighbor relations (ANR) used to populate a list of handoff candidates

•

Authentication of the new eNodeB to the element management system (EMS)

•

Setting radio parameters (C/I and RSSI) for handoffs to neighboring cells

Self-Optimizing Networks Optimization of a 4G network normally focuses on the collection of KPIs, which can then be used to fine-tune the network settings. The goal for a network operator is to optimize for operating cost efficiency, spectral efficiency, or transport cost efficiency and balance this optimization with its targets for quality of service. 3GPP has defined several use cases for self-optimizing networks, including:

3.1.3

•

RF power or antenna tilt optimization to reduce pilot pollution

•

Load balancing between cells

•

Interference avoidance through changes in channel selection and power levels

•

Transport optimization through adjustment of packet sizes and signaling

•

QoS adjustments based on DPI to change packet scheduling priorities

Self-Operating Networks With a network that sets itself up and optimizes its operation, the intelligence can easily be applied to more mundane tasks to assist in tracking network elements and software updates. Key use cases include:

3.1.4

•

Automated inventory tracking for each network element

•

Automated software updates for eNodeB hardware, combined with reconfiguration or re-optimization as necessary

•

Reporting of KPIs to key personnel in the network operator organization

•

Tracking of fault data within the network – with 2G/3G and even multivendor complex networks

Self-Healing Networks Automated network responses to problems constitute an important part of maintaining high-quality services available for the end user. Use cases include: •

Rapid network reconfiguration to compensate for eNodeB outages

•

Re-routing of traffic for loss of service in the transport layer

•

Isolation of faults to identify the root cause of a failure automatically


22

Mobile Network Optimization 3.1.5

ABIresearch

®

Trends in Self-Organizing Networks OEMs are introducing features of SONs ahead of the launch of 4G/LTE systems. Many 3G HSPA and HSPA+ networks, as well as some EV-DO networks, include features for self-configuration and self-healing, as well as inventory control features. The development of femtocells has accelerated this trend, since the self-configuration and self-optimization features of femtocells are implemented in extremely inexpensive equipment and can be easily applied to the RF layer of any NodeB. The end game for SONs is to achieve a network with truly self-optimizing performance, whereby KPIs are tracked and adjustments are made to parameters in Layers 1 through 7 to maximize efficiency and minimize cost. Yet, this Holy Grail is still several years away due to the technical barriers to true optimization in multivendor network environments where the system is essentially untestable until deployment in the field. Operators know that their networks are not prepackaged solutions that are fully tested in a sterile lab environment. Therefore, many operators are very reluctant to implement any optimization algorithms that take over control of the network. Overall, ABI Research expects SONs to become normal practice for the configuration of eNodeBs, as well as 2G and 3G base stations. Inventory and quick-response actions to recover from failures will also be implemented, with steady growth in the sophistication of solutions over the next five years. However, we anticipate that SONs will not result in fully automated networks. Above the radio access layers, independent tools will provide more granular, user-friendly, and actionable information than the 4G SON.

3.2

Monitoring and Optimization Tools A typical mobile operator has 2G voice and SMS services, 3G voice and data services, and will add 4G data services soon. The operator’s data applications demand a wide variety of jitter, latency, and throughput requirements from the network. Trading off radio, transport, and network parameters to help all of these services coexist optimally is challenging for any one network equipment manufacturer. Independent suppliers of monitoring and optimization tools therefore have fertile soil in which to grow businesses based on the aggregation of network performance data, presenting a unified picture to the operator, and suggesting changes to the network. In 3G networks today, the OSS provided with the original network deployment is often seen as inadequate, leading to a separate parallel monitoring system with probes at several interfaces in the network chain. RAN probes and protocol analyzers provide data to a central graphical user interface that displays more granular data than the OSS, allowing for more specific troubleshooting. Here’s an example to illustrate a typical troubleshooting scenario. A common problem arises when peering delays in the IP core produce latency, causing the radio access network to assume that the packets have not been successfully transmitted. The RAN re-transmits the packets (e.g., from the RNC to the SGSN) two or three times, but in fact the packet transmission was successful each time. The processor of the RNC is eventually overloaded, resulting in a failure that appears to be related to call traffic and RAN performance. A coordinated set of protocol analyzer probes between the RNC and SGSN, as well as between the SGSN and GGSN, can isolate the problem and identify the IP latency as the true root cause. As such, network operators plan to use end-to-end monitoring equipment and specific analysis gear from tool vendors to gain the overall visibility and multivendor analysis that their OEM solutions lack.


23


ABIresearch

®

Figure 3.2 An Example of a Network Probe Analysis Report

(Source: Tektronix Communications)

3.2.1

Smartphone Clients Companies such as Keynote, Mobidia, and Venturi Wireless take the challenge of end-to-end performance monitoring all the way to the user device itself. These companies have clients that reside on the smartphone, monitoring the throughput for data applications and, in some cases, working with its sister software in the core network to optimize the transport layer. By positioning an optimization client in the smartphone, Mobidia or Venturi can optimize the TCP link through either compression or a proxy for the TCP sessions. Up to 30% efficiency improvement in the transport layer is possible through these techniques, though some of the compression techniques will cause problems with any DPI tools that are deployed by the operator since the normal data signature will be altered. The client-based monitoring and optimization approach is quite new; it is currently emerging with pre-revenue startup companies. Roughly twenty trials are underway worldwide to verify compliance and interoperability with OSS tools, DPI, SONs, and other network-based optimization features.


24

ABIresearch

®

Mobile Network Optimization 3.2.2

Network Probes Equipment inserted into the network to monitor performance can monitor traffic on almost any interface. This is an established market in 3G networks, with companies such as Agilent, Anritsu, Astellia, and Tektronix providing probes both to network equipment manufacturers and directly to the network operators.

Figure 3.3 Probe Locations in a Typical LTE Network

(Source: Agilent Technologies)

A probe can be used temporarily for initial setup of a network, but increasingly the probes are being implemented as a permanent, stationary feature of the network monitoring strategy. In a 3G network, probes are often installed to monitor multiple locations throughout the system as follows: •

A passive air interface probe can monitor the Uu interface (between the handset and the NodeB) for mobile throughput/error rate issues

•

The Iub interface (between nodeB and RNC) can be monitored for call drop, handover, signaling, and other Layers 2 through 4 KPIs

•

The Iur interface (between RNCs) can be monitored for handover and neighbor list updates

•

The Iu-CS and Iu-PS interfaces (between RNC and MSC/SGSN) can be monitored for transport efficiency

•

The Gn interface (between the SGSN and GGSN) can be monitored for transport efficiency

•

The Gi interface (“north” of the GGSN, at the interface to the IP core network)

As the network migrates to include LTE, a few new interfaces are added, including: •

The X2 interface between eNodeB sites duplicates the Iur functionality

•

Multiple Sx interfaces (S1-U, S1-C, S3, S4, S6a, S11, etc.) connect to the Evolved Packet Core elements such as the Mobility Management Entity (MME) and serving gateway, offering an opportunity to monitor IP traffic for content and efficiency, as well as handover between 3G and LTE


25


ABIresearch

®

A typical monitoring probe solution is usually employed to keep track of performance data. Currently, probes are utilized as a troubleshooting tool rather than an automated optimization tool, working together with a signaling or protocol analyzer to summarize network performance data. Note that these probes are typically sold as a package, together with a graphical user interface or integrated with the vendor’s OSS software.

3.2.3

Signaling and Protocol Analyzers The wired Internet world has created a robust market for IP protocol analyzers, which can provide analytical data on the signaling at each network element, highlighting bottlenecks and issues in the data flow through the network. Portable analyzers are a common historical method for this type of troubleshooting. I complexity and speed of today’s mobile data environment demands that the signaling and/or protocol analyzer must be permanently attached to the network and multiple probes, providing data to multiple users through web interfaces. Signaling analyzers can store analytical data representing a user’s data usage, trace calls, configure a UMTS NodeB, or even record and store a voice call. This type of analyzer can be implemented as a rack-mounted box with integrated storage, or it can reside on the packet core network elements as a software application. In summary, signaling and protocol analyzers offer operators a convenient user interface with which they can see graphical representations of traffic loading patterns at Layers 2 through 4 and a limited view into specific users and applications.

3.2.4

Deep Packet Inspection To analyze the content in the data stream, instead of simply monitoring the type of signaling used to establish data connections, several suppliers have introduced DPI tools. DPI tools typically include equipment permanently installed in the IP layers of the mobile network, and the current trend is for high-capacity solutions (60 Gbps and trending up to 100 Gbps) to monitor every packet transmitted through the network. DPI tools can isolate the type of traffic transmitted, allowing the operator to precisely understand which users are downloading video content, web pages, FTP files, VoIP traffic, or other applications. Current state-of-the-art DPI tools provide the ability to perform real-time analysis on high-capacity data networks. At CTIA 2010, companies such as Allot Communications, CCPU, Procera, and Tellabs were showing DPI capabilities up to 60 Gbps to 80 Gbps, which enables every packet to be inspected in real time. DPI solutions have advanced to the extent that they can pick off signaling traffic to identify the individual user device and/or NodeB associated with each data packet. So far, DPI vendors appear to be focused on Layer 5-7 identification of applications and session traffic, while protocol and signaling analyzers focus on Layer 3-4 efficiency. DPI is presently moving toward the prioritization of traffic by application while other solutions are focused on network/transport layer efficiency, independent of the data application.


26


ABIresearch

®

Although DPI tools are becoming highly sophisticated and capable, the use of the data will require time to mature. Mobile operators today typically use the analysis to understand data traffic patterns (VoIP vs. web browsing vs. video, etc.), but the future evolution of these tools will become more integrated with different levels of optimization. Consider the following: •

Certain types of network traffic will automatically be given higher priority (i.e., VoIP traffic may require higher priority to avoid latency, while web browsing may be more tolerant).

•

New pricing plans for end users will evolve based on the capability inherent in DPI engines in order to separate the pricing for different kinds of data. Facebook fanatics may be willing to pay a premium for each Facebook megabyte, but not for VoIP traffic. Conversely, other users will have different priorities. The rich diversity of end-user preferences creates fertile ground for DPI-based pricing plans to grow.

Note, though, that privacy is an issue that may inhibit the growth of DPI tools. So far, the general public has not reacted to an operator’s capability to monitor traffic usage, but there is a risk that privacy concerns in the wider consumer audience will impact an operator’s ability to prioritize traffic. In addition, any ongoing discussion of “net neutrality” in regulatory circles could inhibit the growth of DPI tools in the mobile data network.

3.3

Centralized vs. Distributed Optimization The algorithms used for decision making in network optimization are often carried out in a central operations support system (OSS), taking advantage of a central point for data collection. Actions that require data from multiple network elements and affect parameter settings in multiple modules are usually centralized. On the other hand, actions that can take place independently of other network resources can be localized in a single eNodeB, and software optimizing these areas can reside in the eNodeB itself.


27


ABIresearch

®

The distributed SON architecture is best suited to use cases involving network element self-configuration, which typically take place upon deployment of the system. Optimization of efficiency, QoS, and adaptation to failure conditions are use cases that are expected in future LTE systems but have less relevance in the early deployment phase. As a result, ABI Research anticipates SONs will move from distributed architectures toward more centralization over time, with different use cases handled in different ways. Figure 3.4 Centralized SON vs. Distributed SON Architectures

(Source: Comarch)

3.4

Interoperability for Multiple Optimizing Algorithms The danger of creating a system with dozens of algorithms lies in the tendency for separate processes to work at cross purposes to each other. While one algorithm may be adjusting parameters such as frequency settings to avoid interference, a second algorithm may be changing the frequency settings back to accommodate changes in another cell. An overall strategy is necessary to resolve any conflict between systems. This issue becomes especially acute in the case of multi-generational, multivendor systems. 3G optimization on Vendor A’s NodeB may conflict with 4G optimization on Vendor B’s femtocell or Vendor C’s LTE infrastructure. Concern over interoperability leads to a major question for mobile operators: Should they combine SONs supplied by their major OEM suppliers with monitoring/optimization solutions from third-party vendors?


28


ABIresearch

®

The scope of endless possible issues between different vendors and algorithms is too large to address in this report, but we can highlight a few of the known issues as examples: •

Any optimization in the Layer 2 or 3 packet protocols that compress data will render DPI useless for monitoring the content at higher layers.

•

RF frequencies and power levels can be adjusted by multiple algorithms as new cells are configured or as interference avoidance is adjusted. Multiple algorithms that address Layer 1 and 2 radio settings may compete with each other, resulting in either constant oscillation in the optimization algorithms or convergence (a non-optimal compromise).

•

A firewall crash (or virtually any failure in the core network) can drop thousands of users at one time. When the users re-register on the network, hundreds or thousands of data connections will initiate at once, overloading the NodeB, RNC, HLR, and SGSN simultaneously.

Overall, the concern about interference between algorithms is not well articulated in the market today, but it remains a nagging doubt for operators. Vendors with partnerships and pre-tested solutions will emerge over the next few years in order to satisfy customer anxiety. To some degree, standardization of SONs and DPI will help to define the boundaries for each solution. Nevertheless, the overall complexity of possible interactions between solutions is so large that standardization and OSS software alone will not be enough. Growth of the market for independent optimization solutions depends on each vendor’s ability to pre-test its solutions in a wide variety of network environments.

3.5

Mobile Offloading The notion of pulling mobile data traffic away from the mobile core network has risen very quickly to the forefront of the industry recently. Femtocells, Wi-Fi, and backhaul offloading have all become very hot topics lately, as the mobile tsunami threatens to swamp many mobile data operators. Femtocells and Wi-Fi lie outside the scope of this report since they are consumer premises solutions. In brief, though, the network can also offload its own data traffic. New solutions from companies such as Stoke can recognize the originating application for network traffic and can redirect the traffic directly to an Internet node instead of routing the traffic through a 3G SGSN/GGSN or through an LTE media gateway.


29

ABIresearch

®


Section 4. KEY INDUSTRY PLAYERS 4.1

Actix Ltd Based in London, Actix is an independent private company that supplies OSS software. Actix offered a self-optimizing network solution during early 2009 for LTE networks. The company supports both network operators and OEM customers with software development related to network management and optimization.

4.2

Agilent Technologies Agilent has participated in the mobile ecosystem from its early days as a supplier of test equipment in the radio layer and protocol and signal analyzers to develop network infrastructure. As networks have developed, Agilent has migrated toward more support of “live” networks, with probes and software to monitor and aggregate information for the operator. The company also supplies OSS software for mobile networks.

4.3

Alcatel-Lucent Alcatel-Lucent supplies network infrastructure to the mobile market, and today’s organization combines business groups from Alcatel, Lucent, and Nortel to cover GSM, CDMA, and UMTS/HSPA technologies. The company offers its “Wireless Network Guardian” as a “dynamic control” OSS solution, whereby Alcatel-Lucent can monitor and troubleshoot multivendor networks. The company is focused on the multivendor aspect of this monitoring/optimization product line, and has placed over twenty systems with network operators globally over the past eighteen months. Overall, Alcatel-Lucent has been able to move quickly into the monitoring/optimization space and will directly compete with nimble smaller companies for a sizable share of the independent monitoring market.

4.4

Allot Communications Allot is an independent supplier of DPI solutions with a focus on the mobile market. The company devotes significant R&D effort toward finding efficient “signatures” for mobile data traffic so that it can handle high-capacity DPI cases in which 45 Mbps to 60 Gbps must be handled in real time. Allot’s products examine every packet and can isolate the user and the cell site involved in order to integrate with policy control functions. The company reached $41 million in both fixed and mobile network revenue during 2009. Allot is in a good position, with a reputation for having efficient solutions for high-capacity DPI applications.

4.5

Amdocs As a large supplier of OSS for a wide variety of networks, Amdocs collects roughly $2.8 billion in revenue annually. Only a small proportion of its revenue comes from the mobile infrastructure market, due to the integrated bundles offered by mobile specialists such as Ericsson, Huawei, and Nokia Siemens Networks. However, Amdocs has excellent know-how in defining the customer experience and should be able to offer some useful tools for the monitoring/management of higher layers (Layer 4 and above). Still, the company is likely to continue to have difficulty capturing a major share of the general mobile OSS market.


30


4.6

ABIresearch

®

Anritsu Company Anritsu Company and its parent Anritsu Corp in Japan are focused on the test and measurement market. They have established leading market share for portable radio test gear for mobile networks. Anritsu collected roughly $490 million in test and measurement-related revenue in 2009. In addition, the company supplies OSS-related software monitoring solutions to 3, T-Mobile, Telcel, Vodafone, and other operators.

4.7

Astellia Astellia has supplied radio probes and related monitoring software to as many as 150 operators worldwide, focusing on helping operators optimize their radio layer. The company collected roughly $35 million in revenue during 2009, growing slightly despite the global recession. Astellia currently supports 2G and 3G technologies and can aggregate KPIs from its own probes, as well as information from network elements. Astellia is publicly traded with headquarters in France.

4.8

Azimuth Systems Inc Azimuth Systems focuses on pre-deployment laboratory testing for mobile networks, with new development stepping out into field testing for trials and deployment. The company works closely with either network operators or OEM customers to bring its staff’s experience in radio frequency propagation and channel conditions into drive test scenarios. While Azimuth does not participate heavily in the monitoring of live networks, it does get involved with troubleshooting and optimization of the radio performance layer.

4.9

Comarch Based in Warsaw, Comarch is an established supplier of OSS mediation software, taking information from network elements or network management systems to present a unified view of network performance. The company also focuses on inventory management and the use of network configuration information to trace the root cause of network issues. Comarch is more active in OSS outside of the mobile/wireless market, but has focused recent efforts on the stronger growth in the mobile OSS market. Overall, Comarch will be challenged to compete with large network equipment manufacturers and independent mobile specialist vendors.

4.10

Continuous Computing Corp Continuous Computing provides enabling technologies directly to OEMs, rather than to network operators. The company supplies ATCA-based racks with software for network monitoring, protocol stacks and software for 3G and 4G networks, as well as professional services to customize monitoring solutions. Continuous Computing addresses both macro layer and femtocell solutions. During the past few years, it has migrated into the DPI market, providing at least twelve different mobile infrastructure manufacturers with a basic DPI engine to analyze their traffic.

4.11

Empirix Privately held and headquartered near Boston, Empirix has roughly 280 employees. The company provides quality assurance software for IP networks in multiple markets, ranging from wireline telephony to cable networks and mobile networks. Empirix has had a slow start in capturing major mobile customers and may be challenged to find a place in the mobile market as LTE vendor relationships solidify.

4.12

Ericsson AB Ericsson holds leading market share in the mobile network infrastructure market. Based in Stockholm with over 80,000 employees worldwide, Ericsson maintains its leadership through strong relationships with hundreds of network operators. The company supplies mobile radio


31


ABIresearch

®

access, transport, and packet core infrastructure, including SON systems for LTE. In addition, Ericsson provides OSS software solutions and services to assist operators in monitoring and optimizing their networks. Roughly 38% of Ericsson revenue comes from the services organization. It is interesting to note that in 2003, only 24% of revenue came from services, showing a trend toward more service support – even though there is no official company strategy to move in that direction. Ericsson acquired Redback Networks in 2006, giving the company greater capability and market share in IP routing. Recently, this organization has been able to provide Ericsson with DPI tools for mobile networks as well, thus providing Ericsson with better diagnostic capability for mobile networks. Note that Ericsson systems present an opportunity for independent vendors to enter with OSS software, probes, and DPI solutions because the company’s infrastructure is perceived as relatively inflexible for monitoring and optimization customization.

4.13

EXFO Inc EXFO acquired NetHawk Oyj in March 2010, creating a combined company to offer a strong competitor in IP packet capture, DPI, and transport monitoring tools, as well as protocol analysis and troubleshooting tools. The company’s solutions address fixed and mobile IP networks, with a clear focus on the mobile market despite revenue behind some competitors in protocol analysis and DPI. It reported roughly $60 million in revenue related to EXFO’s protocol analysis/troubleshooting tools and wireline focus and roughly $40 million related to NetHawk’s protocol analysis, network simulation, DPI, and service assurance solutions for mobile networks.

4.14

Huawei Technologies China-based Huawei has quickly become a strong competitor to Ericsson and Nokia Siemens Networks in the mobile infrastructure market. Huawei has been extremely effective in capturing new network opportunities for GSM and WCDMA networks, overlaying its infrastructure on legacy systems provided by other vendors in most cases. The company has won contracts with very inexpensive hardware, and the software provided with the network has been useful for managing the Huawei elements in the network. Huawei’s OSS suite is less popular with regard to aggregating performance monitoring from multivendor networks. A growing opportunity to support Huawei networks with software solutions and monitoring systems will be a key part of the independent market for optimization solutions.

4.15

Keynote Systems, Inc Keynote Systems provides a unique mobile monitoring solution based on a smartphone client. The Keynote “Mobile Device Perspective” measures download speeds for web sites using a typical smartphone, creating a useful service for network operators and enterprise customers to check the performance of mobile web services. Keynote currently controls virtually all of the revenue for monitoring solutions based on UE client applications.

4.16

Mobidia Mobidia is a pre-revenue startup company that focuses on optimization of the transport layer. The company provides a proxy for TCP sessions that uses software loaded in the user’s device and in the GGSN (or a separate box) to bypass the mobile transport layer and boost efficiency. An estimated 15% to 30% efficiency increase in Layer 4 is typical for existing smartphones and 3G networks. Mobidia has roughly 15 million connection manager clients installed so far with one major global mobile operator.


32


4.17

ABIresearch

®

Motorola Despite a very strong history in mobile communications, Motorola has struggled to capture a major position in 3G and LTE networks, leaving the company with a relatively small but loyal customer group. Motorola plans to deploy an LTE network for KDDI in Japan, along with Zain Saudi Arabia and others. The company has developed a distributed approach for SONs and is pushing to gain advantage through adoption of TD-LTE as a major variant of the LTE standard.

4.18

Nippon Electric Corp (NEC) NEC began shipping LTE base stations to NTT DoCoMo during March 2010, highlighting its continued presence as a strong infrastructure supplier within the Japanese market. In addition, NEC has some early LTE trial deployments with Telefonica and SingTel. The company includes SON technology in its LTE solution, as well as software from Actix as an optimization algorithm. NEC will face a challenge in competing with the low costs of Huawei and ZTE, as well as the political clout and scale of Ericsson/NSN/Alcatel-Lucent.

4.19

Nokia Siemens Networks Nokia Siemens Networks is a joint venture of Nokia and Siemens, with 60,000 employees and a global presence as a supplier of mobile infrastructure. During 2010, NSN has renewed its push to provide service assurance solutions to mobile operators. These solutions cover multivendor networks and bring SON features to market in advance of LTE networks. NSN’s SON uses a multi-level architecture with distributed elements and centralized elements, with some capability to handle multivendor inputs for standardized KPIs. The company differentiates through the use of multiple SON layers in the radio layer, the routing/transport layers, and a new capability to be introduced in the core network. NSN has a good stable of service assurance products and can handle multivendor networks. However, it will be challenged by nimble third-party vendors of service monitoring and optimization solutions as NSN’s GSM footprint becomes less important in vendor selections.

4.20

Procera Networks, Inc Procera is a fast-growing supplier of DPI products for both fixed and mobile networks, with roughly $17 million in annual revenue from fixed and mobile networks. The company claims to have the most comprehensive library of signatures available, which it uses to provide the best high-resolution view of applications running on the network. This claim is supported by network operator comments and the 1,600 applications identified by Procera today. The company supplies both OEMs and network operators and has a focused strategy. Therefore, ABI Research sees Procera as a company to watch.

4.21

Radcom Based in Israel, Radcom is a specialist supplier of network monitoring solutions with expertise in radio optimization. The company provides a system of probes and service monitoring software that operates as a standalone solution, independent of the radio access network it is monitoring. Radcom currently reports roughly $12 million in annual revenue, but sees strong revenue growth coming in the radio optimization space. Its solutions attempt to analyze not just Layers 1 and 2, but also the higher layers. Radcom’s system can pick off information about the origin or type of data transmitted without DPI so that the company can optimize radio parameters. The limited scope of the Radcom solution is likely to be successful, but only in its niche of radio optimization.


33


4.22

ABIresearch

®

Spirent Spirent focuses most of its effort on test solutions, instead of the monitoring and optimization of live mobile networks. The company participates in the optimization of transport layer for mobile networks, capturing information on throughput and packet loss through probes embedded in the network. Spirent’s core expertise in the measurement of jitter and latency is useful in isolating problems. However, in the fast-changing market for mobile monitoring and optimization, Spirent may be left out of key segments if the company cannot move quickly to offer a more comprehensive suite of probes and software.

4.23

Stoke, Inc Stoke is a Silicon Valley startup company that has grown a reputation for effective offload of mobile network traffic. Its session exchange solution diverts IP traffic, bypassing the SGSN and GGSN for Internet traffic and selectively passing other mobile traffic through. The Stoke hardware resides physically between the RNC and SGSN, monitoring the Iups interface to reroute the traffic as needed. Stoke includes a DPI solution or an IMEI (International Mobile Equipment Identity) “sniffer” to detect the application and user type, thus allowing new business models and savings in backhaul equipment. ABI Research expects rapid growth in the offload segment, and Stoke will likely move quickly to take a significant share.

4.24

Tektronix Communications Tektronix Communications is a wholly owned subsidiary of Tektronix Inc (which is owned by Danaher), giving the organization a focus on monitoring and optimization of mobile networks. The company collects roughly $300 million per year in annual revenue for test equipment, probes, and monitoring software, most of which comes from the mobile market. Tektronix Communications acquired Arantech in 2008, adding roughly $35 million in OSS revenue and an expanded product focus. The company supplies to both OEMs (13% of revenue) and network operators (87% of revenue). Tektronix commands a leading position in probe-based monitoring and optimization software, and its relationships with almost all key players ensure a strong place as LTE network monitoring ramps up.

4.25

Tellabs, Inc Tellabs generated $1.5 billion in revenue during 2009, with business areas ranging from mobile networks and backhaul solutions to optical networks and Ethernet networks. The company supplies a “Smart GGSN,” which performs DPI to enable optimization in the core network and routing/transport layers. Tellabs’ equipment typically does not replace network elements like the GGSN. Rather, the company’s equipment makes the elements more efficient through by providing a dedicated DPI engine and feeding information back to the GGSN or other network elements for optimization savings of up to 15%. Tellabs also gets involved with policy engines to block traffic or prioritize certain applications.

4.26

Venturi Wireless Venturi Wireless is a private company based in California with multiple technologies for mobile data optimization. The company offers server-based optimization for video and data applications, as well as a client server architecture with software on the mobile terminal. Venturi Wireless uses a proprietary protocol to replace TCP for higher efficiency without compressing data.


34

ABIresearch

®


4.27

Zhong Xing Telecommunications Equipment Company Ltd (ZTE) Founded in 1985 by state-controlled companies involved in China’s Ministry of Aerospace, ZTE has grown very quickly to become a solid global contender in GSM, 3G, and LTE networks. The company is now traded publicly in Hong Kong and Shenzen, and has established customers in more than 140 countries. ZTE offers SONs for the LTE standard, and has integrated some of the self-diagnostic features of this technology into its existing GSM and 3G network solutions.

4.28

Overall Market Share Estimates A look at the overall market for independent monitoring and optimization solutions (excluding SON solutions and OSS software provided by major OEMs) reveals that just a few major vendors have the overall breadth of product capability to capture major market share. Specialist vendors in individual areas such as DPI, OSS software, or specific monitoring areas without the breadth of a wider product line are not able to compete for the larger contracts involving test equipment, probes, analyzers, and software. As network operators get buried in complexity, the ability to handle this end-to-end functionality will become more and more important.

Chart 4.1

Market Shares for Independent Monitoring/Optimization Vendors, 2009

Agilent 13% Others 29%

Anritsu 20%

Astellia 5% Tektronix Comms 27%

EXFO 6%



35

ABIresearch

®


Section 5. INDUSTRY COLLABORATION 5.1

3GPP Self-Organizing Networks SON use cases are defined in the LTE (E-UTRAN) standards by the RAN3 working group, beginning with Release 8. Additional SON functionality will be added to future releases. The primary purpose behind the standardization of use cases is to ensure the interoperability of SONs in multivendor environments. Standardized SON features will expand in scope as the expected LTE network evolves over time. Release 8 covers configuration features, as well as: •

Automated neighbor relations

•

Automated software downloads

•

Automated inventory tracking

•

Automated Peripheral Component Interconnect (PCI) assignment

Release 9 is expected to address issues related to more mature LTE networks, including: •

RF optimization for coverage and capacity

•

Random Access Channel (RACH) optimization

•

Load balancing

•

Energy savings

•

Operation and Maintenance interfaces to control femtocells

•

UE reporting for end-to-end radio optimization

Technical report TR-36.902 codifies the proposals for Release 8 and Release 9 SON features. http://www.3gpp.org/ftp/specs/html-INFO/36902.htm

5.2

SOCRATES The SOCRATES project (Self-Optimisation and Self-Configuration in Wireless Networks) brings together a non-profit consortium of companies, including Ericsson, Nokia Siemens Networks, Vodafone, and others, with the aim to accelerate the adoption of technologies to enable SONs in mobile networks. This organization creates a forum for discussion of use cases and architectures outside of the political environment of the 3GPP standards process. The SOCRATES project is supported by the European Union under the 7th Framework Program, and is currently scheduled to run from January 1, 2008 to December 31, 2010. The EU’s goal for SOCRATES is to assert European leadership in mobile networks, coordinating a strategy for global standardization to benefit European suppliers and network operators. http://www.fp7-socrates.org/


36

ABIresearch

®


Section 6. MARKET FORECASTS The overall outlook for wireless infrastructure involves slow growth, with severe price erosion cutting into the profitability of RAN equipment and hardware for both 3G and LTE. However, the market for network monitoring and optimization solutions includes several bright spots, as strong growth is expected in technologies that improve efficiency and quality of service. Mobile Monitoring & Optimization Equipment Revenue by Segment, World Market, Forecast: 2009 to 2015 1,400 UE Clients 1,200

Radio Test Radio Probes

1,000

Routing/Transport OSS

($ Millions)

Chart 6.1

800

DPI Offloading

600

400

200

0 2009

2010

2011

2012

2013

2014


In particular, tools that provide good end-to-end visibility (such as DPI and some OSS software) and highly focused solutions for specific cost reductions (such as offloading or transport optimization) will be adopted most quickly in both 3G and 4G networks. On the other hand, SON solutions will compete with independent radio optimization solutions, stunting the growth of independent probe-based monitoring tools in the radio layer. UE clients will see solid growth from almost zero in 2010, with the potential to become a mainstream solution if lightweight client applications are proven to be effective in end-to-end optimization.


37

ABIresearch

®


6.1

Regional Outlook Countries with the most advanced data networks will lead the way in monitoring and optimization. For mobile operators, monitoring 3G networks is currently viewed as a defensive measure. Smartphone and laptop users are changing their data patterns so quickly that operators fear the kind of public embarrassment that AT&T Wireless has experienced during 2009 and 2010. Mobile Monitoring & Optimization Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 1,200 Asia-Pacific North America 1,000

Western Europe Eastern Europe

800 ($ Millions)

Chart 6.2

Africa South America Middle East

600

400

200

0 2009

2010

2011

2012

2013

2014


Currently, Japan and Western Europe represent the strongest markets for advanced monitoring equipment, with the United States catching up quickly. Developing markets lag behind, as users in Eastern Europe, Africa, South America, and the Middle East use less mobile data than power users in more mature networks. North America’s growth wave is taking place earlier than other regions and should flatten out in 2014-2015, when the basic monitoring hardware for most LTE networks has been fully deployed. Over the next five years, ABI Research expects the Asian market to overtake North America and Western Europe for advanced monitoring and optimization equipment. This forecast is due to the higher expected subscriber base and higher growth in mobile Internet data in Asian countries without widespread broadband service. The five-year outlook for developing markets in Africa, South America, and the Middle East lags behind Asia. Quality-of-service requirements in these emerging markets will not demand the robust monitoring solutions that Japan, Korea, and China will implement.


38

ABIresearch

®


6.2

OEM Market vs. Direct Sales to Operators Independent suppliers of network monitoring solutions are currently moving more quickly than major OEMs. The result is a trend toward more direct independent business with network operators. In particular, DPI solutions, network probes, and OSS software are currently trending toward outside suppliers despite the product offerings of a few network equipment manufacturers. Over time, ABI Research expects this trend to reverse. Network OEMs will increase the sophistication of SON solutions, and as LTE is deployed, the number of use cases and functionality for self-optimization will rise. OEMs will buy many of the best solutions from independent suppliers, and will implement pieces like the radio optimization and self-configuration functions internally. By 2015, revenue for independent suppliers of monitoring/optimization solutions will be roughly 30% through OEMs and 70% direct to operators.

Chart 6.3

Mobile Monitoring & Optimization Equipment by Customer Type, World Market, Forecast: 2009 to 2015 100% 90% 80%

Netw ork Operator

70%

OEM

60% 50% 40% 30% 20% 10% 0% 2009

2010

2011

2012

2013

2014



39

ABIresearch

®


6.3

Outlook by Network Generation (2G/3G/4G) Monitoring/optimization of 3G networks will dominate the market for the next five years. As LTE networks are deployed, supporting monitoring and optimization equipment will always lag behind, addressing problems as the system gets loaded (and sadly, in some cases, addressing problems after the system gets overloaded). ABI Research’s forecast includes radio test equipment used for monitoring emissions and power levels for GSM and 2G CDMA networks, which represented roughly 40% of the 2009 market. Excluding radio test equipment, almost all of 2010 monitoring and optimization revenue will be tied to 3G networks, with 90% of the solutions monitoring U-TRAN (UMTS/HSPA/HSPA+) networks and 10% monitoring EV-DO networks. Operators are clearly focused on mobile data optimization, and spend far less effort on mobile voice optimization nowadays. In addition, many independent solutions such as OSS and DPI are in deployment on 3G networks today, but will be used for LTE networks later. ABI Research’s forecast categorizes the monitoring/optimization solutions according to the initial primary usage, even if the same solution is used for multiple generations over time.

Chart 6.4

Mobile Monitoring & Optimization Equipment, 2G/3G/4G, World Market, Forecast: 2009 to 2015 100% 90% 80% 4G

70%

3G

60%

2G

50% 40% 30% 20% 10% 0% 2009

2010

2011

2012

2013

2014



40

ABIresearch

®


6.4

Smartphone Client Outlook Solutions that utilize client software on the mobile terminal itself offer an intriguing alternative to pure network solutions, addressing the network operator’s desire for a true “end-to-end” solution. The revenue associated with these solutions sits at practically zero today, with early solutions just reaching the market. As a result, even with very strong growth, the market for client-based solutions is likely to remain small through 2015. It is unclear today whether every smartphone will need a monitoring client installed. In cases such as the Venturi Wireless optimization protocol or the Mobidia TCP proxy, the step toward transport optimization requires a client on every smartphone. Simpler monitoring concepts may not need every user to act as a mobile “drive test.” As the market for client-based solutions develops, the choice between monitoring and full optimization will dictate the scale of growth. Mobile UE Clients for Network Monitoring Revenue by Region, World Market, Forecast: 2009 to 2015 60 Middle East 50

South America Africa Eastern Europe

40 ($ Millions)

Chart 6.5

Western Europe North America

30

Asia-Pacific 20

10

0 2009

2010

2011

2012

2013

2014


Note: Enterprise sales of clients to monitor mobile web traffic or web site performance are not included in ABI Research’s forecast. The above chart refers to client-based monitoring and optimization solutions and includes sales of the smartphone client software, as well as any accompanying server software in the network.


41

ABIresearch

®


6.5

Radio Test Equipment Outlook Spectrum analyzers, vector network analyzers, and scalar analyzers today often sit in the back of a cell technician’s truck, ready for the endless site tests and drive tests that accompany changes to 2G and 3G networks. While these portable instruments are not always characterized as “monitoring and optimization” solutions, in fact these are the primary tools for optimizing the physical aspects of the radio layer. LTE-capable test equipment is available in the market today, so sales are shifting quickly from 2G/3G test gear to the full capability. Even if the equipment is used for 2G or 3G systems, operators planning for the future want LTE functionality built in. RAN Portable Test Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 250

200 Middle East South America ($ Millions)

Chart 6.6

150

Africa Eastern Europe Western Europe

100

North America Asia-Pacific

50

0 2009

2010

2011

2012

2013

2014


Western Europe and North America have represented the majority of the RF test equipment market for many years, due to the attention paid in Western countries to emissions regulations. Essentially every base station and every NodeB is tested for compliance to spurious emissions targets. Over the next five years, ABI Research expects the Asian, Eastern European, and emerging markets to begin monitoring compliance with emissions regulations, as well as optimizing power levels and antenna tilt in the field for improved coverage and capacity. The market for portable RAN test equipment will be essentially flat because the market has already grown during 2G and 3G operations to involve multiple test boxes for each technician. New sales will include new functionality for LTE/WiMAX testing, but the price differences and volume increases will be flat. At a radio level, the new standards represent business as usual, with new features added to the box.


42

ABIresearch

®


6.6

Outlook for Probes in the RAN Probes in the radio access network essentially provide a duplicate or parallel source of information, at times with better granularity and visibility than the network elements provide through the original OSS software. Operators buy RAN probes today in order to access a more user-friendly and intuitive troubleshooting interface provided by a third-party vendor. As operators consider deployment of LTE networks with self-organizing features, the penetration of probes in Layers 1 and 2 for 2G and 3G networks has grown less rapidly than other network features. Operators are unsure of the eventual scope of self-optimizing LTE networks and how their future systems will be integrated. LTE networks may, in fact, need fewer probes for monitoring and troubleshooting than 3G networks. Thus, the likely outcome over the next five years will be slow growth for RAN probes at 5% to 10% per year. The dominant Western European/American market for RAN probes will grow incrementally, with other regions increasingly investing in improved diagnostics as operators in developing countries dive into multivendor networks. RAN Monitoring Probes Revenue by Region, World Market, Forecast: 2009 to 2015 160 140 120 Middle East ($ Millions)

Chart 6.7

100

South America Africa

80

Eastern Europe Western Europe North America Asia-Pacific

60 40 20 0 2009

2010

2011

2012

2013

2014



43

ABIresearch

®


6.7

Outlook for Transport Monitoring/Optimization Layers 3 and 4 (network layer and transport layer) represent a stronger market for stationary probes, troubleshooting solutions, and optimization schemes than Layers 1 and 2 (radio layer and data link layer). In the RAN, the air interface is generally not changed to streamline traffic due to the wide variety of handsets and data terminals on the network. However, in setting up data sessions and setting packet sizes for transport, both ends of the communications link can be controlled, and several vendors have found ways to improve on the efficiency of TCP and are finding willing customers. Monitoring 3G data sessions has also become a major concern for network operators, as iPhones and other web-friendly smartphones are driving a huge increase in data signaling traffic. Nobody expects SON to address this issue for 3G networks or even early LTE networks, so the independent market for monitoring solutions should see strong growth over the next five years. Growth in Layer 3 and 4 solutions should scale along with data traffic growth, as the capacity handled by monitoring/ optimization solutions will grow quickly in conjunction with the growth in data usage. Western Europe, North America, and Japan are the strongest markets for transport optimization. Growth in China and India will take place later as mobile broadband begins to comprise a significant portion of traffic. Transport Monitoring/Optimization Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 700 Middle East 600

South America Africa

500

Eastern Europe Western Europe North America

($ Millions)

Chart 6.8

400

Asia-Pacific

300

200

100

0 2009

2010

2011

2012

2013

2014



44

ABIresearch

®


6.8

Mobile OSS Outlook The total OSS market for mobile networks totals is in the billions of dollars annually, with the majority of sales bundled together with network infrastructure. The independent market for OSS to augment visibility and management of the system will reach $450 million in 2010, with 20% growth over the next five years driven by increasing complexity in the network. OSS software provided by major OEMs will increasingly be replaced by independent software, which can adapt more quickly and flexibly to unique aspects of each network’s vendor base and architecture. OSS Monitoring/Optimization Tools Revenue by Region, World Market, Forecast: 2009 to 2015 1,200 Middle East South America 1,000

Africa Eastern Europe

800


($ Millions)

Chart 6.9

600

Asia-Pacific

400

200

0 2009

2010

2011

2012

2013

2014



45

ABIresearch

®


6.9

DPI Outlook One of the strongest growth segments in mobile infrastructure centers on Deep Packet Inspection tools. Because DPI is proven in Internet networks, mobile operators can readily adopt the technology from established vendors and expect high levels of reliability. The dozens of DPI tests and trials underway now should bear fruit during 2010 and 2011, driving 70% to 100% growth over the next few years. Asia, Western Europe, and North America all represent strong markets for high-capacity DPI tools that examine every packet. Due to the exponential growth of mobile data traffic in these regions, DPI revenue will also grow quickly. LTE deployment in highly developed countries should accelerate DPI adoption even more as operators prepare for their customers using LTE as a broadband replacement.

Chart 6.10 Mobile DPI Infrastructure Revenue by Region, World Market, Forecast: 2009 to 2015 1,400 Middle East South America

1,200

Africa Eastern Europe

1,000

($ Millions)


800

Asia-Pacific 600

400

200

0 2009

2010

2011

2012

2013

2014



46

ABIresearch

®


6.10

Offloading Outlook Mobile network offloading will expand quickly during the 2011-2012 timeframe due to the growing gap between transport bandwidth and end-user demand. Rather than “optimizing” transport, an offloading solution moves the most bandwidth-hungry applications off the operator’s core infrastructure. This approach should find a timely market window for the most data-intensive networks now and acceptance from a wide number of global networks over the next three years. Femtocells (not included in the forecast for “offloading” as an optimization solution) are distinct in that femtocells will not sense the content and divert individual packets to the Internet; instead, femtocells divert all traffic to the Internet. Offloading is likely to be most popular with network operators that feel capable of implementing third-party solutions. In Asia and developing countries, operators do not have the expertise to feel comfortable in implementing a new solution, so the growth of offloading solutions will be slower in those regions. North American operators and multinational operators based in Europe and advanced Asian economies are conducting trials now, and ABI Research expects strong growth in those regions through 3G networks over the next three years. LTE networks are likely to be outfitted with offloading solutions on deployment, so the 3G market is the primary focus for now.

Chart 6.11 Mobile Backhaul Offloading Infrastructure Revenue by Region, World Market, Forecast: 2009 to 2015 180 Middle East

160

South America 140

Africa

($ Millions)

120

Eastern Europe Western Europe

100

North America 80

Asia-Pacific

60 40 20 0 2009

2010

2011

2012

2013

2014



47

ABIresearch

®


Section 7. RECOMMENDATIONS 7.1

For Network Operators Every network operator has seen the tremendous rise in mobile data consumption, but very few operators have adequate visibility to pinpoint the source of bottlenecks in their networks. Operators need to develop a comprehensive monitoring and optimization strategy in order to avoid costly mistakes and delays. The first operators that push through the painful process of integrating DPI tools, network diagnostics, and new billing systems with tiered pricing structures will reap the benefits of increased customer traction. Operators should stay focused on the end goal: increasing average revenue per user through multiple data options with the infrastructure and policy engines to match.

7.2

For Network Equipment Manufacturers Network OEMs should partner with other companies in multiple areas in order to integrate the best possible flexibility and visibility into their networks. Differentiation in mobile networks is moving toward ease-of-use and scalability, both of which can be multiplied by the use of innovative diagnostic tools and a set of easy-to-use optimization algorithms. Network equipment manufacturers can differentiate by folding all possible optimization solutions into a multivendor, comprehensive optimization suite that has clear priorities. Competing independent solutions will struggle with interoperability issues. Thus, the top-tier OEMs can succeed by testing hundreds of combinations of network equipment and optimization solutions and refining their overall set of algorithms to consistently adapt to complex networks. As network OEMs increasingly move into managed services, each company should pay attention to keeping network optimization solutions lined up with its services business strategy. Network OEMs may find that a unique approach to optimization is a differentiator in their services business.

7.3

For Monitoring/Optimization Solution Vendors Most monitoring/optimization suppliers will see strong growth in the next two to three years. However, in the long term, only the companies with a clearly defined objective and obvious interoperability will be able to penetrate the mainstream mobile markets. Vendors of client-based solutions should focus on optimization, not simple performance monitoring, in order to drive ROI through network efficiency savings. Probe vendors should partner with network infrastructure manufacturers to drive the use of probes into LTE hardware. Close integration with top-tier OEMs will drive the sales of monitoring equipment more quickly than competing for attention with up-and-coming SON solutions. In transport optimization, suppliers should stick to a focused strategy and avoid extending solutions into areas that optimize multiple other layers simultaneously. Interoperability with other optimization systems will be best served when the transport compression or session protocols are predictable.


48


ABIresearch

®

OSS vendors that can adapt quickly to an extremely wide array of network configurations will rise above the pack. Most customers have, by now, experienced limitations in network visibility for mobile data diagnostics. Therefore, the network operators are looking for OSS solutions that can drill down to the root cause in troubleshooting scenarios, regardless of the infrastructure used. Offloading vendors simply need to establish a track record of reliability in the field. The market demand is inescapable, so simple offloading solutions that interfere as little as possible with other network diagnostics and optimization will dominate this new segment.


49

ABIresearch

®


Section 8. COMPANY DIRECTORY Actix Ltd (United Kingdom) www.actix.com

Facebook www.facebook.com

Agilent Technologies www.agilent.com

Huawei (China) www.huawei.com

Alcatel-Lucent (France) www.alcatel-lucent.com

Keynote Systems, Inc www.keynote.com

Allot Communications (Israel) www.allot.com

Mobidia, Inc (Canada) www.mobidia.com

Amdocs Ltd www.amdocs.com

Motorola, Inc www.motorola.com

Anritsu Company (Japan) www.anritsu.com

NetHawk Oyj (now EXFO) (Finland) www.nethawk.fi

Apple Inc www.apple.com Astellia (France) www.astellia.com

Nippon Electric Corp (NEC) (Japan) www.nec.com

AT&T Wireless www.wireless.att.com

Nokia Siemens Networks (Finland) www.nokiasiemensnetworks.com

Azimuth Systems Inc www.azimuthsystems.com

Procera Networks, Inc www.proceranetworks.com

Comarch (Poland) www.comarch.com

Radcom (Israel) www.radcom.com

Continuous Computing Corp www.ccpu.com

Rohde & Schwarz (Germany) www.rohde-schwarz.com

Empirix www.empirix.com

Stoke Inc www.stoke.com

Ericsson AB (Sweden) www.ericsson.com

Spirent Communications (United Kingdom) www.spirent.com

EXFO Inc (Canada) www.exfo.com

Tektronix Communications www.tektronixcommunications.com


50


ABIresearch

®

Tellabs www.tellabs.com

Vodafone (United Kingdom) www.vodafone.com

Venturi Wireless www.venturiwireless.com

Zhong Xing Telecommunications Equipment Co. (ZTE) (China) www.zte.com.cn


50

ABIresearch

®


Section 9. ACRONYMS 2G

Second Generation Cellular Services

3G

Third Generation Cellular Services

3GPP

Third Generation Partnership Project

4G

Fourth Generation Cellular Services

ANR

Automatic Neighbor Relations

ARPU

Average Revenue per User

ATCA

Advanced Telecommunications Computing Architecture

BSS

Business Support System

C/I

Carrier to Interference Ratio

CAGR

Compound Annual Growth Rate

CAPEX

Capital Expenditure

CCPU

Continuous Computing Corp

CDMA

Code Domain Multiple Access

CPRI

Common Public Radio Interface

CSCF

Call Session Control Function

C-SON

Centralized Self-Organizing Network

CTIA

Cellular Telephone Industry Association

DPI

Deep Packet Inspection

D-SON

Distributed Self-Organizing Network

EMS

Element Management System

eNB

Enhanced Node B

EPC

Enhanced Packet Core

© 20q0ABI Research • abiresearch.com The material contained herein is for the individual use of the purchasing Licensee and may not be distributed to any other person or entity by such Licensee including, without limitation, to persons within the same corporate or other entity as such Licensee, without the express written permission of Licensor.

51

ABIresearch

®


E-UTRAN

Evolved Universal Terrestrial Radio Access Network

EV-DO

CDMA Evolution-Data Only

EVM

Error Vector Magnitude

FTP

File Transfer Protocol

Gbps

Gigabits per Second

GGSN

Gateway GPRS Support Node

GoS

Grade of Service

GPRS

General Packet Radio Service

GSM

Global System for Mobile

GW

Gateway

HLR

Home Location Register

HSPA

High Speed Packet Access

HSS

Home Subscriber Server

iDEN

Integrated Digital Enhanced Network

IMEI

International Mobile Equipment Identity

IP

Internet Protocol

IP

Intellectual Property

KPI

Key Process Indicator

LTE

Long Term Evolution

Mbps

Megabits per Second

MME

Mobility Management Entity

MMS

Multimedia Messaging Service

MRF

Multimedia Resource Function

MSC

Mobile Station Controller

NEC

Nippon Electric Corp

NEM

Network Equipment Manufacturer


52

ABIresearch

®


NOC

Network Operations Center

NSN

Nokia Siemens Networks

OAM

Operations, Administration, and Maintenance

OBSAI

Open Base Station Architecture Initiative

OEM

Original Equipment Manufacturer

OPEX

Operating Expense

OSI

Open System Interconnection

OSS

Operations Support System

PC

Personal Computer

PCI

Peripheral Component Interconnect

PCRF

Policy Control Rules Function

PDN

Packet Data Network

PDP

Packet Data Protocol

P-GW

Packet Gateway

PIM

Passive Intermodulation

PSTN

Public Standard Telephone Network

QoE

Quality of Experience

QoS

Quality of Service

R&D

Research & Development

RACH

Random Access Channel

RAN

Radio Access Network

RF

Radio Frequency

RNC

Radio Network Controller

ROI

Return on Investment

RSSI

Receive Signal Strength Indicator

SGSN

Serving GPRS Support Node


53

ABIresearch

®


S-GW

Serving Gateway

SMS

Short Messaging Service

SOCRATES

Self-Optimisation and Self-Configuration in Wireless Networks

SON

Self-Organizing Networks (and sometimes Self-Optimizing Networks)

TB

Terabyte

TCP

Transmission Control Protocol

TD-LTE

Time Domain Long Term Evolution

TD-SCDMA

Time Domain-Synchronous Code Domain Multiple Access

UE

User Equipment

UMTS

Universal Mobile Telecommunications System

UTRAN

Universal Terrestrial Radio Access Network

VoIP

Voice over Internet Protocol

WCDMA

Wideband Code Domain Multiple Access

Wi-Fi

Wireless Fidelity

WiMAX

Worldwide Interoperability for Microwave Access

ZTE

Zhong Xing Telecommunications Equipment Company


54

ABIresearch

®


Table of Contents Section 1. ........................................................................................................................................ 2 Executive Summary ....................................................................................................................... 2 1.1 Introduction ............................................................................................................................. 2 1.2 Market Drivers......................................................................................................................... 2 1.3 Technology.............................................................................................................................. 3 1.4 Outlook.................................................................................................................................... 4 1.5 Recommendations .................................................................................................................. 4 1.6 Conclusions............................................................................................................................. 5 Section 2. ........................................................................................................................................ 6 Market Overview............................................................................................................................. 6 2.1 Networks Evolve into Optimization ......................................................................................... 7 2.2 Key Performance Indicators.................................................................................................... 8 2.3 R&D/Testing/Deploying the Network .................................................................................... 10 2.4 Monitoring the Network ......................................................................................................... 10 2.5 Optimizing Performance ....................................................................................................... 11 2.5.1 Cost Reductions................................................................................................................. 11 2.5.1.1 CAPEX Reduction........................................................................................................... 11 2.5.1.2 OPEX Reduction ............................................................................................................. 12 2.5.2 Quality Improvement .......................................................................................................... 12 2.6 Optimizing Revenue.............................................................................................................. 13 2.7 Test Equipment Market ......................................................................................................... 13 2.8 Probes................................................................................................................................... 14 2.9 Protocol Analyzers ................................................................................................................ 16 2.10 Impact of Managed Services on Optimization .................................................................... 17 2.11 Operations Support Systems .............................................................................................. 18 Section 3. ...................................................................................................................................... 20 Technology ................................................................................................................................... 20 3.1 Self-Organizing Networks ..................................................................................................... 21 3.1.1 Self-Configuring Networks ................................................................................................. 21 3.1.2 Self-Optimizing Networks................................................................................................... 22 3.1.3 Self-Operating Networks.................................................................................................... 22 3.1.4 Self-Healing Networks ....................................................................................................... 22 3.1.5 Trends in Self-Organizing Networks .................................................................................. 23 3.2 Monitoring and Optimization Tools ....................................................................................... 23 3.2.1 Smartphone Clients............................................................................................................ 24 3.2.2 Network Probes.................................................................................................................. 25 3.2.3 Signaling and Protocol Analyzers ...................................................................................... 26 3.2.4 Deep Packet Inspection ..................................................................................................... 26 3.3 Centralized vs. Distributed Optimization ............................................................................... 27 3.4 Interoperability for Multiple Optimizing Algorithms................................................................ 28 3.5 Mobile Offloading .................................................................................................................. 29 Section 4. ...................................................................................................................................... 30 Key Industry Players.................................................................................................................... 30 4.1 Actix Ltd ................................................................................................................................ 30 4.2 Agilent Technologies............................................................................................................. 30 4.3 Alcatel-Lucent ....................................................................................................................... 30 4.4 Allot Communications ........................................................................................................... 30


55


4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18 4.19 4.20 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28

ABIresearch

®

Amdocs ................................................................................................................................. 30 Anritsu Company................................................................................................................... 31 Astellia................................................................................................................................... 31 Azimuth Systems Inc............................................................................................................. 31 Comarch................................................................................................................................ 31 Continuous Computing Corp............................................................................................... 31 Empirix ................................................................................................................................ 31 Ericsson AB......................................................................................................................... 31 EXFO Inc............................................................................................................................. 32 Huawei Technologies.......................................................................................................... 32 Keynote Systems, Inc ......................................................................................................... 32 Mobidia................................................................................................................................ 32 Motorola .............................................................................................................................. 33 Nippon Electric Corp (NEC) ................................................................................................ 33 Nokia Siemens Networks.................................................................................................... 33 Procera Networks, Inc......................................................................................................... 33 Radcom............................................................................................................................... 33 Spirent................................................................................................................................. 34 Stoke, Inc ............................................................................................................................ 34 Tektronix Communications ................................................................................................. 34 Tellabs, Inc.......................................................................................................................... 34 Venturi Wireless .................................................................................................................. 34 Zhong Xing Telecommunications Equipment Company Ltd (ZTE) .................................... 35 Overall Market Share Estimates ......................................................................................... 35

Section 5. ...................................................................................................................................... 36 Industry Collaboration................................................................................................................. 36 5.1 3GPP Self-Organizing Networks........................................................................................... 36 5.2 SOCRATES .......................................................................................................................... 36 Section 6. ...................................................................................................................................... 37 Market Forecasts.......................................................................................................................... 37 6.1 Regional Outlook................................................................................................................... 38 6.2 OEM Market vs. Direct Sales to Operators........................................................................... 39 6.3 Outlook by Network Generation (2G/3G/4G)........................................................................ 40 6.4 Smartphone Client Outlook ................................................................................................... 41 6.5 Radio Test Equipment Outlook ............................................................................................. 42 6.6 Outlook for Probes in the RAN.............................................................................................. 43 6.7 Outlook for Transport Monitoring/Optimization ..................................................................... 44 6.8 Mobile OSS Outlook.............................................................................................................. 45 6.9 DPI Outlook........................................................................................................................... 46 6.10 Offloading Outlook .............................................................................................................. 47 Section 7. ...................................................................................................................................... 48 Recommendations ....................................................................................................................... 48 7.1 For Network Operators.......................................................................................................... 48 7.2 For Network Equipment Manufacturers ................................................................................ 48 7.3 For Monitoring/Optimization Solution Vendors ..................................................................... 48 Section 8. ...................................................................................................................................... 50 Company Directory ...................................................................................................................... 50 Section 9. ...................................................................................................................................... 51


56


ABIresearch

®

Acronyms...................................................................................................................................... 51 Scope of Study ............................................................................................................................. 58 Sources and Methodology .......................................................................................................... 59 Notes ............................................................................................................................................. 59

Please be aware that an Excel worksheet containing all market forecasts accompanies this document. When downloading this report as a PDF from the ABI Research web site, please check to see if the Excel worksheet is also available for download. If you have any questions regarding this, please contact our client relations department. TABLES Table 1-1. Mobile Network Monitoring & Optimization Equipment Revenue, World Market, Forecast: 2009 to 2015 Table 3-1. Applicability and Impact of Various Monitoring Solutions Table 6-1. Mobile Monitoring & Optimization Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-2. Mobile UE Clients for Network Monitoring Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-3. RAN Portable Test Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-4. RAN Monitoring Probes Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-5. Transport Monitoring/Optimization Equipment Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-6. OSS Monitoring/Optimization Tools Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-7. Mobile DPI Infrastructure Revenue by Region, World Market, Forecast: 2009 to 2015 Table 6-8. Mobile Backhaul Offloading Infrastructure Revenue by Region, World Market, Forecast: 2009 to 2015


57

ABIresearch

®


SCOPE OF STUDY In researching and reporting on mobile monitoring and optimization, ABI Research limited its scope to the solutions available as products in the merchant market. Network equipment manufacturers with bundled solutions for infrastructure equipment and OSS software are not covered. However, adaptations of the OSS solutions offered separately by network OEMs for use with multivendor networks are included in the body of the report and the market forecasts. A wide scope was used in looking at multiple solutions and segments, ranging from Layer 1 through Layer 7 and from smartphone clients to hardware and server-based software. ABI Research compiled a comprehensive overview of the technologies used for network monitoring and optimization. Portable test equipment was included to the extent that the equipment is used for diagnostic monitoring and troubleshooting of an operating network and not simply for initial system setup. Self-organizing networks are included in the scope of the technology description in this study since they represent an important piece of an operator’s overall strategy for network optimization. Note that the self-organizing network is generally sold to a wireless operator by a network equipment manufacturer. Therefore, ABI Research did not include a forecast for SONs in the merchant market quantified in this report. The technology scope covers descriptions of monitoring and optimization techniques, including the interfaces and network elements to be monitored and the actions taken for optimization. Problems with interoperability between optimization solutions are described at a high level, with a few specific examples for illustration. Details on the specific implementation of individual products are omitted for the sake of brevity. Note that business solutions, including business support system software and other tools for tracking and billing the wireless subscriber, are not included in the scope of this report, though these solutions can involve some monitoring and optimization.


58

ABIresearch

®


SOURCES AND METHODOLOGY An analyst was assigned to coordinate and prepare this Research Report. Research and query specialists helped lay the data and information groundwork for the analyst, who also developed a focused interview strategy. ABI Research teams follow a meticulous process when examining each market area under study. The three basic steps in that process are: information collection, information organization, and information analysis. The key element in ABI Research’s information collection process is developing primary sources, that is, talking to executives, engineers, and marketing professionals associated with a particular industry. It is from these conversations that market conditions and trends begin to emerge, free from media hype. Analysts use secondary sources as well, including industry periodicals, trade group reports, government and private databases, corporate financial reports, industry directories, and other resources. Analysts’ conclusions take several forms. The text addresses hard data and well-defined trends and is supported by forecast tables and charts. The text also addresses issues and trends that are difficult to quantify and present in neat, tabular form. Lying at the margins of an industry, they are often precursors of the next technology wave. For this report on mobile network monitoring and optimization, the ABI Research analyst interviewed multiple network operators, network equipment manufacturers, and suppliers of specific test equipment, software, monitoring solutions, and optimization solutions. Each company was asked to project a view of its market and the driving forces behind future growth. The forecast was derived from company reports and network operator comments. It is segmented generally by Open System Interconnection (OSI) layer, since most solutions address the RAN in Layers 1 and 2, the core network efficiency in Layers 3 and 4, or application efficiency in Layers 5 through 7. Regional forecasts and other market segmentation were derived from specific comments from suppliers and regional operators.

NOTES CAGR refers to compound average annual growth rate, using the formula: CAGR = (End Year Value ÷ Start Year Value)(1/steps) – 1. CAGRs presented in the tables are for the entire timeframe in the title. Where data for fewer years are given, the CAGR is for the range presented. Where relevant, CAGRs for shorter timeframes may be given as well. Figures are based on the best estimates available at the time of calculation. Annual revenues, shipments, and sales are based on end-of-year figures unless otherwise noted. All values are expressed in year 2010 US dollars unless otherwise noted. Percentages may not add up to 100 due to rounding.


59

ABIresearch

®


Published 2Q 2010 ©2010 ABI Research PO Box 452 249 South Street Oyster Bay, NY 11771 USA Tel: +1 516-624-2500 Fax: +1 516-624-2501 http://www.abiresearch.com/analystinquiry.jsp

ALL RIGHTS RESERVED. No part of this document may be reproduced, recorded, photocopied, entered into a spreadsheet or information storage and/or retrieval system of any kind by any means, electronic, mechanical, or otherwise without the expressed written permission of the publisher. Exceptions: Government data and other data obtained from public sources found in this report are not protected by copyright or intellectual property claims. The owners of this data may or may not be so noted where this data appears. Electronic intellectual property licenses are available for site use. Please call ABI Research to find out about a site license.


60

Mobile Netwk Opt for 3G and 4G Netwks

Recommend Documents