Top 20 Frequently Asked Business Questions About IMS
Dr. Jerry Lucas
10/01/2006
Telecoms invest in their network infrastructure for two reasons: first, to increase revenues and hopefully profits; and second, to increase operational efficiency and hopefully reduce expenses. This tutorial addresses IP multimedia subsystems (IMS) and increasing revenues.
IMS is a hot topic for conferences and magazine articles, because there seems to be consensus on what the questions are, but no consensus on the answers. So here are my top 20 frequently asked business questions about IMS, with answers from both an optimistic and pessimistic perspective.
Question 1: What’s IMS?
IMS is a standard architecture being developed by almost all tier 1 telecom operators and their vendors. It provides a service creation and delivery platform for new session initiation protocol (SIP) based multimedia services independent of network access technology.
From a technical perspective the IMS architecture is a new standard that reuses previously developed IP protocols and introduces new application servers. So what is meant by standardized? It means that functions and interfaces can be located in multiple nodes (computers), or multiple functions can exist in a single node. Several IP protocols are reused, such as SIP; Diameter for authentication, authorization and accounting; and common open policy service (COPS) for gathering usage information from routers. What is new about IMS is the many new SIP servers and the additions to existing IP interface protocols.
Question 2: Why IMS?
IMS is an important architecture for the telcos for three reasons. First, it monetizes the IP network. Today’s telco IP residential revenues are generated by broadband access and ISP service. Both are flat-rate, all-you-can-eat. The reason is that the Internet was designed for routing and connectivity, not for usage monitoring. IMS gives telcos control over IP networks so they can offer and bill for usage, applications and QoS-based services.
Second, the telco can create new services and businesses—for example, seamless wireline, wireless and IP networking—and more importantly it can unbundle customers from the network and introduce non-telco products that have a telecom component.
Third, IMS enables operational efficiency. An IMS architecture can lead to the removal of OSS/BSS smokestacks. The bottom line is that all future services will only need one OSS/BSS.
Question 3: Why do service providers need IMS?
Service providers need IMS for three reasons. First, it enables QoS. The Internet is best-effort service, and the quality varies. IMS synchronizes session QoS with service provisioning.
Second, IMS allows for accurate accounting. Today the Internet is designed to route packets, and the basic signaling protocol is SIP peer-to-peer. With IMS, service providers are aware of what customers want to do or what they have done, because all signaling (SIP) is monitored by the service provider.
Third, IMS enables service integration. Because IMS architecture unbundles the user from the network, it’s easier for third-party services to add applications. Also, all services use packet technology, making it easier to add or customize service offerings.
Question 4: Who doesn’t need IMS?
The arrival of the Internet has been a disaster for wireline telcos’ bottom-line revenues, in large part because of VoIP. Worse yet, all the great new services and revenue streams created by the Internet—advertising, e-commerce and more—benefited non-telcos. The portals (Yahoo!, Google, MSN) love the Internet the way it is now. Flat-rate pricing and best-effort is just fine. All they want from the telcos is “dumb pipe,” net-neutral access for their customers; eBay and other e-commerce companies don’t need IMS (or SIP, for that matter) for continuous growth. Microsoft already has the equivalent of IMS, called Connected Service Framework. And in the time it takes you to read this article, more non-IMS hardware and software will be sold (think global GSM) than all of the IMS stuff that will be sold in year 2007 or more.
Now for the more controversial questions. Here is my take on how both an IMS optimist and pessimist would answer the remaining questions.
Question 5: What’s the IMS killer app?
An optimist would say that IMS allows the telecoms to offer presence, push-to-talk, digital content and, finally, “if you build it, they will come.” A pessimist would say, if you asked a person on the street what presence means, they would think you were talking about Christmas. Or ask a sophisticated user if they wanted presence, and they would say they already have it—it’s called a cell phone. Push-to-talk is here today; same with digital content. Technologies try to solve problems, but users have already found answers with simple solutions. So, if you built IMS, will they come?
Question 6: Will wireless, wireline and cable IMS networks be the same?
The IMS optimist would say that all the major standards groups—ITU, ETSI, 3GPP, 3GPP2 and so on—have endorsed IMS, all are working to harmonize IMS standards among themselves, and eventually you will see but one IMS architecture. The pessimist would say that 3GPP is GSM-focused, 3GPP2 is CDMA-focused, and TISPAN is PSTN/ISDN-focused. All of these standards groups are supported by different telecoms (or divisions with a telecom), each is going after the same customer, and all will want service differentiation. Just like you have dozens of different wireless networks, you will see dozens of different IMS networks.
Question 7: What’s an IMS device?
The optimist would say that an IMS device is a SIP-enabled terminal, a dual-mode phone (CDMA or GSM, and Wi-Fi or WiMax) or an IMS client framework in a device to which a user can add IMS-type applications. A pessimist would respond by stating that they don’t know what an IMS device is, because they haven’t seen one yet, or that out of curiosity they would buy one, if only they could find one.
Question 8: Will IMS reduce churn?
An optimist would say that IMS will reduce churn. It’s common industry knowledge that the more services customers subscribe to, the less likely they are to leave their service provider. In addition, IMS makes it much easier to add a new service. For example, the same buddy list can be reused for new services, rather than going to another service provider for it. Services can also be personalized; there is one-stop shopping, one bill, free service trials, smart service bundles and more.
The IMS pessimist would say it is unlikely that IMS will reduce churn, because the more services that are provided, the greater the chance for decreasing overall customer satisfaction because some of those services could be poor in quality. Also, the free trial services can degrade the overall QoS and thus the paid critical services. Bottom line: There are too many risky unknowns with IMS to assume it’s going to lead to higher customer satisfaction.
Question 9: What services does IMS support today?
Again, here the optimist would say push-to-talk, presence, fixed-mobile convergence and many more services that are under trial. The pessimist would respond by saying that everything that is IP- or SIP-based today is being labeled IMS by the zealots and hype-mongers.
Question 10: Is VoIP service the stepping-stone to full IMS deployment?
The IMS optimist would say absolutely yes! IMS is simultaneous SIP-based voice with data or video. The pessimist would say yes as well, but add that VoIP would have to be bundled with something else in order to move IMS along. But that something else is likely to be content that has to be delivered by a non-telco. Also, an IMS OSS/BSS will be required beforehand, and that is not here yet.
Question 11: What is the status of IMS among tier 1 service providers today?
Before delving into this question, some background. Introducing new technology into tier 1 networks goes something like this: First, a vendor will demonstrate the product at its own site; second, the testing will move into the service provider’s lab; third, service trials are launched (usually by a service provider’s own employees); and finally the product is fully deployed in a revenue-producing mode.
Now, the IMS optimist would claim that more than 50 percent of the tier 1s will have IMS deployed in 2007. The pessimist would counter this statement by saying yes, all tiers are looking at IMS, some are lab testing, and no one is doing a field trial as of yet. 2007 is basically the year for IMS tire kicking.
Question 12: Can small IMS network equipment providers (NEPs) succeed?
The optimist would argue that service providers will want “best of breed” suppliers. The pessimist would claim that only the big NEPs can provide an end-to-end solution. IMS is a “think big, start small” kind of play—operators will want to “start small” with someone who can deliver the “think big.”
Question 13: What is the challenge for IMS application developers?
The optimist would answer this question by stating that application developers will see an IMS network as an enhanced, QoS-enabled network that can support many new applications. They will develop new stuff for IMS networks like entrepreneurs (eBay, Yahoo! and Google) did with the Internet.
The pessimist would respond by saying that no one owns the Internet, and the terminal (the PC) is owned by the customer and has a well-established connection standard (IP). With IMS the developer doesn’t know the service provider’s bandwidth policy, given that the IMS device might be doing other thing like non-IMS IPTV and more. IMS is nowhere near ready for third parties or application developers to prosper.
Question 14: Are there promising resale opportunities with IMS?
The optimist would say that there are resale opportunities, because regulators will support IMS network resale just like they did for the PSTN. For an IMS network, all a reseller would need is a home subscriber server (HSS) and a billing system. The pessimist would say, no way! The FCC ordered PSTN resale in a 1970s-style business environment, and then took the competitive teeth out of the Telecom Act. Things are different now. Regardless, if a tier 1 is ordered to allow IMS resale, they won’t build it.
Question 15: What’s A-IMS?
Before going over the different viewpoints on this one, let’s first look at some background on just what A-IMS (Advances to IMS) is. Verizon Wireless identified shortcomings in current IMS standards—specifically, a lack of support for non-SIP applications, security weaknesses and more. It is working with Cisco, Nortel, Qualcomm, Lucent and Motorola to fix these problems. Currently, A-IMS is before the 3GPP2 standards group for consideration.
The optimist sees A-IMS as an improvement to IMS—it’s technology-agnostic (not just for CDMA), making it a win-win situation. The pessimist sees this as a first step toward a standards disaster. It’s now GSM’s IMS versus Verizon Wireless A-IMS versus Cingular, Sprint or Vodafone IMS. It’s a fracturing development that further reduces the chances of seeing a single IMS architecture any time soon.
Question 16: Can IMS monetize the “long tail”?
The long tail is a business model that has been promoted by Chris Anderson of Wired Magazine. His theory is about why the future of business is selling less of more. The IMS optimist thinks that the long tail is only possible because the Internet makes IMS the perfect match for what the long tail needs.
The IMS pessimist thinks this is a classic case of too little too late. Legitimate P2P networks like SNOCAP and micropayment companies like Javien and PayPal are already monetizing the long tail without IMS.
Question 17: Is net neutrality good or bad for IMS?
The optimist would say that net neutrality is good for IMS. A net-neutral Internet created many new businesses and business models. A net-neutral IMS network will do the same. The IMS pessimist thinks net neutrality is bad—IMS is not about a net-neutral environment for a service provider, it’s about monetization, QoS, walled gardens and getting a piece of the Google and Yahoo! pie. If service providers are ordered to make their IMS networks “neutral,” they won’t build them.
Now here is where a couple of questions get more complex.
Question 18: What are the top 10 technical challenges with IMS?
1. Security
2. Integrating SIP and non-SIP applications
3. Vendor interoperability
4. Carrier interoperability
5. Standards readiness
6. Mapping standards to product profiles
7. Measuring the customer’s QoE
8. OSS/BSS
9. General complexity
10. The unknowns
1. Security: With IMS, every SIP-based terminal is a source of network attack. For IMS networks to be secure, service procedures will have to support reverse firewalls. For example, even if Verizon Wireless solves this problem with its A-IMS, there are bound to be other security holes.
2. Integrating SIP and non-SIP applications: Other than the Internet and enterprise intranets, there is no end-to-end IP networking application creating a revenue stream. Non-SIP applications will be driving most of the tier 1’s revenue for the next 10 years or so. The challenge will be to make the user experience (applications) seamless across SIP and non-SIP infrastructure.
3. Vendor interoperability: Vendors love being sole-source suppliers, and tier 1s are looking early for turnkey IMS vendors. But tier 1 telecom networks today have hundreds of network components supported by vendors who compete with each other. Further complicating matters is that IMS infrastructure is a combination of telecom technologies with IT. The service provider will be playing a very large role in integrating products from different vendors and industry segments (IT and telecom).
4. Carrier interoperability: The vision of IMS is that a subscriber will be supported seamlessly over any access technology (DSL, cable modem, wireless, etc.) and anywhere (home, office or travel). This is going to require operators to support roaming with QoS support, even if IMS-based carriers want to interoperate; no one has a solution to date.
5. Standards readiness: Back when companies like the Bell System or AT&T were monopolies, standards weren’t an issue. The Bell System would dictate network standards, and if you wanted to sell a product or interconnect with their network, you had to do their way. With IMS, no vendor, operator or NEP can declare itself a leader. Combine this with the activities going on with Verizon Wireless’s A-IMS, and you get the point.
6. Mapping standards to product profiles: It is one thing to say that IMS is based on existing standards (SIP, Diameter, COPS, etc.), but it’s another thing to make IMS standards fit existing IP/SIP products. Further complicating the technical issue here is that IMS functions and interfaces can be scattered over many nodes (computers) or rolled up into one node. One vendor’s IMS box is not going to look like another vendor’s IMS box.
7. Measuring the customer’s QoE: The IMS architecture promises QoS-enabled applications, as well as differentiated service support based on application QoS requirements. Measuring the QoS being delivered to a customer is challenging enough if you are only dealing with lost or delayed packets, when you have multiple networks connected together. But measuring the customer quality of experience (QoE) is even more challenging. IMS is about telecom technology and IT. Where’s the metric for QoE, when it’s more than just QoS?
8. OSS/BSS: Almost every Billing World and OSS Today issue for the last year has covered OSS/BSS challenges of IMS architected networks. In order to save on OPEX in the long run and remove stovepipes, most of today’s OSS/BSS must be replaced by IMS OSS/BSS. Within legacy tier 1s, this replacement is not likely in the near term. So the challenge is to integrate IMS networks with legacy OSS/BSS without IMS deployment creating even more stovepipes.
9. General Complexity: The IMS architecture is like quantum theory in physics. Most of the groundwork for the quantum theory was laid out by 1928, but only a handful of physicists could see the big picture. It took decades to apply it (making transistors, bombs, etc.). Today, many telecom technologists understand their own piece of IMS but don’t necessarily understand the other pieces. Case in point: Verizon’s A-IMS is being developed with a watchful eye out for legacy CDMA service integration, mobile wireless security and more. The bottom line is that most IMS technologists can’t see the forest beyond the trees.
10. The Unknowns: The most challenging aspect of IMS is positioning the deployment to be able to deal with the technical unknowns, such as the impact of lost SIP messages, security flaws, and the list goes on.
Question 19: What are the top 10 business challenges with IMS?
1. Developing a business plan with near-tem ROI
2. Merging/managing IT, network operations and marketing
3. Device availability
4. Competing against free services (portal voice, video, etc.)
5. Meeting today’s regulatory mandates (CALEA, E911, etc.)
6. Unknown legislation (net neutrality)
7. Developing killer apps
8. Intercarrier interconnection
9. Understanding the business of alliance partners
10. Timing (thinking big but starting small)
1. Developing a business plan with near-tem ROI: Today, the telecoms have to see a business plan with an ROI within a budget cycle. No one has discovered the IMS killer app as of yet, and it’s not feasible to see a near-term OpEx reduction with IMS because of the CapEx required to replace OSS/BSS.
2. Merging/managing IT, network operations and marketing: To succeed with IMS an operator is going to require IT, network operations and marketing to cooperate as never before. One of the three has to be in charge, because an IMS operation run by a committee is doomed to failure.
3. Device availability: Even if a telco CEO would write a billion-dollar-plus check to build an IMS network, it would produce no meaningful revenue. Why? Because there are no IMS devices in the hands of the customer.
4. Competing against free services (portal voice, video, etc.): The Yahoo!s and Googles and e-commerce companies like eBay don’t need IMS, nor do they need voice or video payments from their customers. In this Internet ad-based world, services are given away for free (such as Skype by eBay) to attract people. Making IMS networks compete against free is a tough business proposition.
5. Meeting today’s regulatory mandates (CALEA, E911, etc.): Solutions to regulatory mandates are being placed in IP networks, including IMS networks. Lawful intercept (U.S. CALEA) is doable, but it will cost money. E911 in an all-IMS environment is doable, but it will cost money as well.
6. Unknown legislation (net neutrality): Nothing scares investors or telecom boards of directors more than regulatory uncertainty. What are the U.S. Congress and foreign legislators going to do about net neutrality? Why would a telecom invest in IMS, if it has to make the benefits available to the competition? The telecom lobbyists have a challenge on their hands—selling net non-neutrality isn’t easy.
7. Developing killer apps: The IMS optimist says there is no single big killer app, but many small ones. But in reality, you don’t need IMS to support SIP-based services.
8. Intercarrier interconnection: Other than cable CEOs (with the help of CableLabs), C-levels in the wireless and wireline spaces don’t seem to be talking the same way with each other. Application developers will want to see an IMS network that looks like the Internet in terms of user access (for example, it doesn’t matter which broadband or ISP is providing service). Same with content developers: they don’t like exclusive deals with service providers, and neither do the politicians. So for IMS to succeed in generating new, big sources of revenue, IMS networks have to be interconnected.
9. Understanding the business of alliance partners: The telecoms will need content and application developers as business partners to succeed with IMS. Working deals with Hollywood producers means investing in more projects like “Pirates of the Caribbean.” It will be a challenge for a telecom to figure out how to leverage such investments beyond program transmission fees or subscriptions.
10. Timing (thinking big but starting small): IMS deployment is a classic “think big but start small” venture. The challenge for a telecom is how to start slow with IMS but be able respond quickly if a competitor develops the killer app.
Question 20: Who is going to win big with IMS?
The companies best positioned to win big with IMS are IT and network operations outsourcing companies serving large enterprises or Fortune 500 companies. Why? IMS makes sense in the enterprise space, because they are the first adopters of VoIP allocations—particularly fixed mobile convergence—and have budgets for new telecom applications that save money and create a competitive edge.
Why IT/networking outsourcers? Because they manage not only IT applications but network security. Note that IMS is about merging IT with IP networking; also, a problematic area for IMS is network security. Who better to manage corporate IMS network security than outsourcing network security experts? The best part is that outsourcing operators also are application service providers: they “rent” their apps to their clients.
If you want to find out if the telecom analysts are either optimistic or pessimistic on these 20 IMS questions—and which outsourcing companies will be the big winners with IMS in the enterprise space—plan to register for a special webinar, “Top 20 FAQs on IMS: Why the Enterprise Market, and Who Are the Likely Winners?” on Oct. 24, 2007, 1 p.m.–2:30 p.m. Eastern time. For more information on this webinar or to register, go to www.telestrategies.com.
Email this article
Add a comment
Printer version
Order reprints
RSS Feed
Bookmark article