LOGO LTE语音解决方案

汇 报 提 纲 业界LTE Voice方案介绍 IMS&SRVCC方案分析 CSFB方案分析 中兴通讯LTE Voice解决方案 2

移动运营商的目标网络架构是纯PS网络 高级NGMN 系统结构 CS网络将会消亡，目标网络将是纯PS域的业务网络。 ngmn(nextgenerationmobilenetwork)成立于2006年9月，由中国移动、英国vodafone和orange、日本nttdocomo、德国t-mobile、荷兰kpn、美国sprint等全球六大电信运营商共同成立，旨在推动下一代移动网络技术发展。该组织是以运营商为主导的移动通信标准化组织，研究可在2010年～2020年商用的下一代移动网络的需求，制定未来宽带移动网络的系统性能目标、功能要求和演进方式，为相关标准化组织、设备制造商开展下一代移动网络的标准化和产品开发提供明确指导。这将使全球移动通信产业链聚集在统一需求之下，从而降低产业风险、提高产业效率，为移动通信行业构建共赢的和谐生态环境。 在ngmn举办的2008ngmnindustryconfernece上，ngmn宣布了董事会对lte(3gpp)、wimax、umb(3gpp2)进行技术评估的结论：lte/sae是第一个广泛满足ngmn白皮书的技术。由此，ngmn联盟批准lte/sae作为它的第一个适用技术。 在NGMN的目标网络中将只包含分组域接入网络和基于分组的控制网络，电路域中的会话业务将由的分组域端对端的IMS VoIP方式提供，用户感受要求同现有电路域相同，电路域网络最终将退出市场。 由于LTE网络是逐渐发展的，电路域网络不会立即退出，在很长的一段时间内还将配合LTE网络为用户提供语音业务，包括传统用户和LTE覆盖外的LTE用户。 为降低LTE终端的技术难度、减少LTE终端的耗电量，以及LTE部署的方便性，NGMN组织决定LTE与传统2G／3G CS无线之间为Single-Radio制式，即双模单待方式，终端不能同时在LTE和CS域收发信号。为此需要提供相应的技术方案。 CS网络将会消亡，目标网络将是纯PS域的业务网络。 语音业务将在PS域开展，用户在LTE网络上的通话业务感受应与传统CS相同。 在LTE和GERAN/UTRAN无线网络之间做切换时，语音中断时间应小于300ms。 LTE VoIP业务对传统GERAN/UTRAN网络应该没有改动需求。 LTE VoIP业务对传统CS核心网络修改尽可能小。 IMS是业界公认的面向多种分组接入接入的统一语音和多媒体业务控制平台。

LTE 无线部署对现有CS网络的影响 Single Radio意味着LTE和CS业务不能同时使用，类似于“双模单待”。 3GPP/NGMN 确定LTE与2G／3G之间为Single Radio技术,部署LTE需要考虑对CS业务的影响 Single Radio意味着LTE和CS业务不能同时使用，类似于“双模单待”。 在LTE网络注册时，只能接收LTE无线信道中的信息，收不到GERAN／UTRAN的CS寻呼信号、广播信号。在GERAN／UTRAN CS注册时，也无法监听LTE的广播信道，无法收发LTE的数据包。 LTE初期热点地区孤岛式部署，不能像CS一样做到完全覆盖，CS网络需要作为LTE Voice的补充。 Dual Radio Concept the candidate technology can work in parallel with the legacy radio technology in the terminal, at least when handover happens. Less complex network procedures。 Single Radio Concept only one radio technology can be activated at a time. Less complex terminals Less radio interference Less impact on battery life Less transmission power If frequency bands are close to each other, single radio handover is needed. If frequency bands are not close to each other, dual radio handover is possible.

LTE分阶段部署对语音业务有不同影响 LTE的语音和多媒体成为主流 LTE语音和多媒体引入阶段 移动的Internet阶段 纯数据卡阶段 Phase 4 Phase 3 LTE的语音和多媒体成为主流 LTE覆盖大部分地区，基于LTE/IMS的语音/多媒体业务充分发展，CS网络将逐步萎缩，最终CS退出历史舞台。SRVCC技术使用越来越少。 LTE语音和多媒体引入阶段 通过移动终端不仅能够通过LTE享受宽带业务，而且可以打电话，享受丰富的IMS 多媒体业务。 LTE无线覆盖外仍使用CS提供的传统语音业务。 IMS＆SRVCC方案能够解决语音控制和移动到CS网络切换时的语音连续性问题。 Phase 2 移动的Internet阶段 随着终端的能力增强，用户可以通过小型便携的LTE移动终端随时随地享受移动宽带业务，语音业务则仍然由已经部署的传统2G/3G 电路域网络提供。 CSFB方案解决终端在LTE注册时，无法接受GERAN/UTRAN网络信号的问题。 by AT&T, Orange, Telefonica, TeliaSonera, Verizon, Vodafone, Alcatel-Lucent, Ericsson, Nokia Siemens Networks, Nokia, Samsung and Sony Ericsson. Phase 1 纯数据卡阶段 基于纯LTE数据卡开展移动宽代业务，与语音业务无任何交集。 LTE网络在2010－2011年逐步开始有商用部署案例，大多数运营商初期将满足用户的移动宽带数据业务方面的需求，再开展能够替代CS语音业务的会话类业务。

LTE Voice Service Solutions PSTN/PLMN PSTN/PLMN Almost all operators consider IMS as ultimate mobile CN all-IP development direction, and traditional voice network finally becomes an access technique for IMS. Taking IMS-based LTE VoIP as long-term evolving solution, CSFB is taken as transit technique in intermediate stage, and SRVCC and CSFB can be co-existed. Some operators have limited spectrum resources, and expect to assign CS exclusive spectrum to LTE packet service, to provide voice service over LTE based on packet network. INTRENET INTRENET IMS MGCF/MGW CSCF SCC AS MSCS MSCS SGs MSCS MSCS MME SAE-GW MME SAE-GW CS CS Sv SGs EPC MGW EPC MGW MGW MGW RAN/GERAN eNodeB RAN/GERAN eNodeB handover CSFB- UE speech fall back to 2/3G, No VoIP on LTE When UE initiates MO and receives MT calls, it needs hand over to CS wireless network. Suitable for overlapping deployment scenes of 2/3G CS domain and LTE wireless network. Update all VMSCs which have overlapping wireless coverage with LTE, to support SGs interface to perform joint location updating , paging and short messages ,etc functions, like Gs interface. Simple network structure, no need to deploy IMS. SRVCC- LTE VoIP with IMS control Voice and multimedia services are based on LTE, LTE wireless coverage can be considered as the supplementary of 2G/3G wireless network. Enhance MSC control based on SCC AS and SRVCC, LTE voice is able to hand over to CS network. Deploy IMS network as unified control platform of multimedia service . Deploy independent SRVCC to enhance MSC, support Sv interface and SIP interface to avoid all MSCs updating. CSFB和IMS&SRVCC是不同阶段的需求,CSFB是过渡阶段,IMS&SRVCC是目标方案. 原VoLGA方案删除, 因为主推VoLGA的TMO在2011年2月世界移动大会已经宣布放弃VoLGA. VoLGA方案曾经在TMO主导的VoLGA中定义，3GPP未接受此方案。 TMO已经宣布放弃VoLGA，转而支持CSFB和SRVCC。 VoLGA利用新增GAN IWF网元，再利用传统CS网络控制LTE语音，不需要部署IMS，也不需要升级MSC。终端需要支持VoLGA 如果主推IMS&SR-VCC，策略和理由： 1.在热点地区部署LTE网络； 2.VoLTE减轻CS网络压力，为高端用户提供基于LTE的高质量语音和数据业务； 3.在LTE下可以数据和语音并发，如视频共享，并且IMS还可以提供更加丰富的IMS融合业务，例如多媒体会议，一号通等（CSFB数据和语音不能并发）； 4.也符合IMS提供LTE语音的终极目标方向。

VoLTE－业界共同的目标方案 One Voice Profile was released on Nov,2009. The One Voice Initiative was a collaboration between AT&T, Orange, Telefónica, TeliaSonera, Verizon Wireless, Vodafone, Alcatel-Lucent, Ericsson, Nokia, Nokia Siemens Networks, Samsung and Sony Ericsson. One Voice profile use current open standards to define the minimum mandatory set of functionality for interoperable IMS-based voice and SMS over LTE. GSMA Leads Mobile Industry Towards a Single, Global Solution for Voice over LTE 15 February 2010, Barcelona: The GSMA today announced it has adopted the work of the One Voice Initiative to drive the global mobile industry towards a standard way of delivering voice and messaging services for Long-Term Evolution (LTE). The GSMA’s Voice over LTE (VoLTE) initiative support the principle of a single, IMS-based voice solution for next-generation Mobile Broadband networks. (IR9210, IR9220) The GSMA will also lead the development of the specifications that will enable interconnection and international roaming between LTE networks, and will complete that work by Q1 2011. The GSMA has widespread industry support for its VoLTE initiative Mobile operators supporting the initiative include 3 Group, AT&T, Bell Canada, China Mobile, Deutsche Telekom/T-Mobile, KDDI, mobilkom austria, MTS, NTT DoCoMo, Orange, SKT, SoftBank, Telecom Italia, Telecom New Zealand, Telefónica, Telenor, TeliaSonera, Verizon Wireless and Vodafone. VoLTE is a key enabler for the success of LTE Vicente San Miguel, CTO of Telefónica Telefonica plans to deploy LTE to provide an improved Mobile Broadband service for our customers. It is vital that we also deliver a high quality voice and messaging service. Telefónica has led the work with the GSMA on VoLTE and we strongly support this initiative to drive a common voice and messaging solution for the mobile industry, as it is a key enabler for the success of LTE. http://www.telecoms.com/18588/tight-squeeze The surge in mobile data and broadband traffic in advanced markets over the last couple of years is something of a double-edged sword. While it has finally validated carriers’ long-held strategic thinking, it has exposed gaps in their network performance. The surge in smartphone sales has fueled a dramatic increase in the demand for mobile data, which in turn has brought some networks to their knees. This demand will increase exponentially over the next few years as phones become more sophisticated and connectivity comes to a wide range of devices. The world of popular storytelling is full of tales that caution the reader to be careful what they wish for. After all, they might just get it. Since the 3G auctions began a decade ago amid a cycle of hype that, ten years on, looks like a kind of mass hysteria, the wish of mobile operators in the most advanced markets has been for sufficient uptake of mobile data services to justify the enthusiasm with which they snapped up 3G licences. It’s been a long time coming and, now that it’s here, it is proving to be something of a mixed blessing. While the arrival of genuine mobile broadband capability is a cause for celebration, some carriers have been caught short by the sheer scale of uptake. Victims of their own marketing push-offering ‘unlimited’ data usage for relatively low subscription charges-and the unforeseeable impact on users of the iPhone and the devices it has subsequently influenced, carriers have been overtaken by events. High profile headlines have seen US carrier and iPhone pioneer AT&T castigated for failing to deliver a sufficiently robust network to cope with the handset’s uptake, and UK player O2 making public apologies for similar shortcomings. Figures from Informa Telecoms & Media show that total mobile traffic in 2008 measured 1.1 exabytes (trillion megabytes) worldwide. This is forecast to grow to 2.5EB in 2010 and 18EB in 2014. While some of that will be traditional, circuit-switched voice traffic, mobile broadband is set to explode, surpassing 14EB during 2014. By the end of 2010, mobile broadband subscribers worldwide will number 450 million, which will represent a 45 per cent share of the global broadband market. In 2011, though, the mobile broadband market will overtake the fixed segment, growing to near 670 million subscribers and a 51.8 per cent market share. Despite a healthy CAGR of 12 per cent between 2008 and 2013, Informa says, the fixed market’s share of global broadband subscribers will drop by half over the forecast period from 70 per cent to 35 per cent. Carriers have won what they wished for, and the problem of provisioning capacity for the data boom is clearly not going to disappear any time soon. Mobile users vote with their feet and poor network performance is something to which they assign great value. Application performance specialist Compuware recently conducted a survey of senior executives at 22 mobile broadband operators from around the world in a bid to gauge their visibility of customer experience and satisfaction. The results, according to Jerry Witcowicz, product manager for Compuware’s Vantage for Mobile solution, make for a sobering read. “All of the executives we interviewed said that experiencing a poor performance in data service is the top reason for their customers to churn,” Witcowicz says. One interviewee revealed that, “when we introduced the iPhone our traffic spiked by 30 per cent and the capacity did not keep up.” Another revealed: “We expect the network traffic to increase 50 times in the next few years.” The two mobile broadband segments- dongle/embedded laptop (portable) and smartphone (mobile)-are each problematic. Portable usage accounts, and will continue to account, for the overwhelming majority of mobile broadband usage. Informa figures for global mobile network traffic show portable usage at 1EB in 2010, almost 90 per cent of traffic. By 2014 this will have grown to 14EB, representing 93 per cent of traffic. But this does not mean that portable traffic represents the only challenge. UK carrier O2 has been deliberately slow into the dongle market but has been hard hit by uptake of the iPhone. And mobile broadband-capable handsets are going to grow in number enormously over the next four years as technology gets pushed further down device ranges. According to Dave Nowicki, vice president for marketing and product management at US network and femtocell provider Airvana, smartphones have their own particular issues. Studies conducted by Airvana on EVDO networks in the US and Asia, he says, have found that smartphones place a far higher signalling burden on the network than dongles and embedded laptops-a problem he says has just as much relevance to carriers using 3GPP standards. “We found that there was this amazing acceleration in signalling traffic. On measuring it what we noticed was that while a dongle or data card was using 25 times more data than a smartphone, the signalling ratio was only three to one,” he says. “Which means that, for every bit it sends, the smartphone is pumping out eight times more signalling traffic.” Nowicki argues that this requires a shift in the way networks are dimensioned, with carriers needing more radio network controllers. “It’s not so much a spectral efficiency issue,” he says, “it’s more about the cost per megabyte increasing. The cost to deliver a megabyte to somebody is no longer defined just by how many megabytes you’re sending. It’s no longer a common number, it’s now influenced by the mix of your traffic.” It’s almost a tradition in this industry to look to the dawn of the next major technological era as the answer to everyone’s problems. But in these straitened times many operators have conceded that LTE, the champion of the fourth generation, may be slower to arrive than originally anticipated. So what can carriers do? A variety of options are open to them, some network based and some more strategic than technical. Perhaps the most obvious solution is simply to put in more cell sites. It’s the kind of response that enables an operator to show contrition to disgruntled users as well as a willingness to roll up its sleeves and solve the problem. At the end of 2009, O2 UK made just such an announcement in the wake of its iPhoneinduced network issues. The firm promised to build out 1,500 new network sites in 2010, with 200 planned for London, where the iPhone (along with critics from the press and analyst communities) is present in the greatest concentration. The firm said its investment would run to hundreds of millions of pounds, and followed a £500m payout over the previous two years in order to meet increased demand for data. Such a response has its detractors, though. John Spindler, vice president of product management at ADC argues that traditional behaviours require a rethink. “Things have to change in the way that carriers look at the network architectures and network topologies,” he says. “The old days of throwing up cell towers and doing what we used to refer to as ’spray and pray’ just isn’t going to be adequate. It’s becoming impossible.” Even those who back the effectiveness of such a strategy concede that there are obstacles. Land is a finite resource and, in the kind of dense, urban areas in which capacity boosts are most sorely required to accommodate data surges, sites are at a particular premium. Such build-outs are also arguably the most costly form of improvement upon which an operator can embark. But, says Mike Roberts, senior analyst at Informa Telecoms & Media, the cost need not necessarily be prohibitive. “When you get down to the network itself, it can just be ten or 15 per cent of your base stations that are generating the vast majority of your traffic,” he says. “So it might not be a massive investment programme that you’ll need to undertake to add capacity to the base stations that are overloaded. From that point of view it can be a manageable problem,” he says. Roberts’ assessment of geographical load differentials chimes with a statement made late in 2009 by Ralph de la Vega, chief executive at US carrier AT&T. Q309 saw record activations of the iPhone for AT&T but, said de la Vega, just three per cent of his customers were generating 40 per cent of his data traffic. But some argue that although major network deployments may offer a solid capacity solution, the time involved in making them happen lessens their effectiveness in dealing with a problem that requires an immediate solution. Capacity in other forms is already in place, they say, and can be used to alleviate the pressure more promptly. These people are proponents of offload strategies, using either wifi or femtocell technologies as a release valve for congested mobile data networks. It’s an irony particular to the mobile data boom that wireless carriers, for so long motivated by drawing business away from the fixed line, are now scrambling to dump their own traffic back onto it wherever it’s available. “Offloading to wifi is great for us and great for the customer,” says Matthew Key, CEO of Telefónica Europe (read the full, exclusive interview with Key on p18). “The customer gets a better experience and it takes the traf- fic off of our network,” he adds, in the kind of statement that might almost have been unthinkable from a large cellular carrier just a few years ago. “We’ve got agreements with The Cloud and BT in the UK but we’re also going through an educational process with users about how to make use of wifi in their home. The vast majority of people now have broadband in their home, with a wireless router,” he says. Key’s last point highlights the relevance of offload strategies in a world where a great deal of mobile broadband usage is happening in the home. Airvana’s research, says Dave Nowicki, shows that the peak time for mobile data traffic is the evening. Given that that peak traffic remains on the one cell, and that most people are home in the evenings, that traffic must be coming from the home, he says. The total share of home-based mobile broadband could be as high as 50 per cent, he adds. Most laptops today ship with wifi embedded and most smartphones, too. As Stephen Rayment, chief technical officer at Belair Networks, a specialist in wifi offload, argues: “The problem and solution lies in the palm of our hand. More than 1.5 billion wifi chipsets have been shipped to date and we’re looking at one billion run rates in four or five years. Wifi is pervasive and is a very effective way of getting some of that data traffic off the mobile networks.” Rayment argues that wifi makes sense for wide area capacity relief for two reasons likely to prick up the ears of any carrier with a capacity issue; the ease and speed with which it can be deployed, and the cost. Historically the firm’s customers have been owners of large venues like sports arenas and shopping complexes, as well as fixed broadband players looking to add nomadic wireless access to their portfolio as a sweetener. Rayment says wireless carrier deals are in the offing for the first half of this year. “The big advantage for mobile carriers is that they can deploy this offload at anywhere from one third to one tenth the cost of deploying wholesale upgrades to their mobile network,” he says. “You can deploy wifi in a targeted fashion; where the 3G carriers’ subscribers are congregating and where the cell towers are glowing red hot. Plus the devices are already here. We all know LTE is coming but we don’t have to wait for it.” Furthermore, he says, the wifi networks can be plugged into a carrier’s network operations centre, and can be managed using existing tools. Not everyone is convinced, though. “The thing about wifi,” says ADC’s John Spindler, “is that it was never designed for wide area use. From the attempts that we’ve seen to put in metropolitan wifi networks we’ve seen that it doesn’t work. Most have failed miserably. It’s really designed for smaller coverage areas and has a limited number of channels on the access points that you can actually deploy. It’s good as a hotpsot technology for coffee shops, but not for wide area. Plus it operates in unlicensed spectrum so you get more interference.” Rayment concedes that, three or four years ago, it was all but impossible to take a wifi offload solution to a mobile carrier “without being ridiculed”. Central to carriers’ objections, he ways, was the problem of operating in unlicensed spectrum; something he says has now been overcome. “I think they’re beginning to see that you can do some pretty good networks with unlicensed spectrum. Plus there are so many pressures on the mobile carriers-the loads on the networks are driving them to do this.” The offload alternative to wifi in the home is the femtocell. Having occupied the ‘next big thing’ slot on the mobile networks agenda for some time, femtocells are now starting to gain traction. SFR, Vodafone, Sprint and AT&T are just some of the operators to have made high profile moves in the space. But to date most carrier uptake has been driven by coverage issues in a bid to address indoor signal performance, says Dave Nowicki of Airvana, supplier to Sprint. “Coverage is the pain spot that everyone really understands at the moment,” he says. “But people are realising now that there’s a capacity pain point as well, and I expect that capacity femto deployments will be the next phase.” But femtocells, like all solutions, generate issues of their own. Some detractors suggest that femtocells could create problems with interference, and that users on the macro network could experience performance problems when in proximity to a femtocell. But it seems the most serious sticking points could be commercial rather than technical. The last thing cellular carriers need is another expensive device to subsidise into consumers, and asking subscribers to pay out for something to improve a service for which they are already paying is a difficult proposition to work. Vodafone, which was the first European carrier to launch femtocells in July 2009, revised its offering in January 2010, giving added weight to the view that femtocells are a tough sell. The first move was to change the name of the offering from Vodafone Access Gateway- a network-based name if ever there was one-to Vodafone Sure Signal. Prices were cut with the rebrand and now the product costs £50 in a one off charge, or £5/month for 12 months on price plans of £25 or more. If customers spend less than £25/month, the price is £120 in a one off cost, or £5/month for 24 months. While he agrees that femtocells are a solid proposition for domestic usage, ADC’s John Spindler believes that, like wifi, they are not suitable for use as a capacity enhancement in the wide area. ADC favours distributed antenna systems that use fibre to connect smaller cells to the macro base station, generating what Spindler claims is more efficient use of spectrum and a reduction in practical hassles. “When you aggregate a BSS location like this, you’re talking about eliminating a lot of the cost of real estate acquisition, eliminating a lot of the time involved, lowering power and HVAC requirements and enabling easier maintenance because all the upgrades are done at the central BSS location,” he says. What all of these solutions have in common is that they are designed to ease the strain on the radio access network. But the RAN is not the only part of the network that requires attention. “Backhaul is a huge focus for all operators,” says Bradley Mead, vice president for services and multimedia at Ericsson, whose responsibilities include managing the network owned by T-Mobile and 3UK’s MBNL joint venture. “You have to get enough coverage and capacity into the RAN to be able to deliver the service, but then the bottleneck moves up to the backhaul. Because if the backhaul’s not there to support the RAN, there’ll be a big problem.” Lance Hiley, vice president for market strategy at microwave backhaul specialist Cambridge Broadband Networks, takes a similar view. Backhaul provisioning is complicated by the same issues as RAN provisioning, namely the difficulty, involved in predicting where the traffic surges are going to appear, and providing enough capacity to cope with them. The ideal solution for backhaul, Hiley concedes, is fibre, which has limitations far beyond those of the network elements it is being used to connect. The problem, as always, is cost. The answer, says Hiley, is microwave. “You probably have enough credit on your Mastercard to buy a point-to-point microwave link that could backhaul 100Mbps over 2km,” he says. “But you’d probably have to take out a second mortgage on your house if you wanted to do the same thing with fibre. In Western Europe we have fewer than 30 per cent of cell sites connected by fibre. If you think about how many cell sites there are in Western Europe and what it would take to connect the remaining 70 per cent by fibre. It would put the UK trade deficit in the shade.” The enthusiasm for microwave is driving saturation of the spectrum that’s reserved for backhaul, he says, which is why point to multipoint backhaul architectures are coming increasingly to the fore. “With point to multipoint you can deliver a much better quality of service because you’re dynamically allocating resources, while at the same time making the most efficient use of spectrum resources,” he says. The reality is that most carriers will use a range of technical methods to help them deal with the boom in mobile data traffic. In the meantime, the issue of tiered service is rearing its head. Some carriers-Vodafone Portugal is one-are putting software into their networks that throttles back the speed available to subscribers who have exceeded their fair usage caps. Others, AT&T included, are introducing price hikes to dissuade customers from excessive usage. It’s a divisive strategy. ADC’s John Spindler describes price caps as “a crazy business model.” Compuware’s Jerry Witcowicz, meanwhile, says that “what we hear from customers is that the unlimited usage offer is a losing proposition. You can’t afford to offer unlimited usage, operators have to have caps.” Telefónica Europe CEO Matthew Key says he is undecided, and that a policy is being thrashed out within the group at the moment. “How many industry that are relatively capital intensive can live with a model that says ‘pay one fee and it doesn’t matter how much you consume’? People who are using multiple gigabytes each month are paying the same as somebody who’s using one megabyte. The key is to get a balance between the two, and that’s a live strategic debate for us at the moment.” And, one suspects, for many other carriers watching the sudden boom in mobile data usage.

CSFB和SMS over SGs－过渡阶段的选择 NGMN Alliance Delivers Operators’ Agreement to Ensure Roaming for Voice over LTE February 15th 2010, Frankfurt, Germany and London, UK: NGMN stated that the Alliance and its members are firmly committed to IMS Voice as the target solution for Voice over LTE. Furthermore, NGMN recognises the benefits of recommending a migratory non IMS voice solution that guarantees voice roaming with LTE handsets whilst allowing use of LTE data roaming in the same device. Thus NGMN recommends that: In all LTE devices that support GSM or UMTS and that provide the voice service, Circuit Switched Fall Back (CSFB) shall be implemented as a minimum requirement for supporting voice . The CSFB implementation shall be compliant to the 3GPP specifications. If a Mobile Network Operator operates a network comprising LTE plus GSM or UMTS and if this MNO aims to provide a non-IMS voice service as well as an LTE data service to visiting subscribers utilising an LTE device, then it shall, as a minimum, support CSFB for voice. The implementation shall be compliant to 3GPP specifications. NGMN Alliance Recommend implementation of the “SMS only over SGs from Day 1 14 April 2010, Status update Voice/SMS over LTE Task-force: In order to ensure the SMS roaming for the data only devices from Day 1 of the commercial operations of the LTE EPC networks, NGMN agrees the following way forward: – Desire the UE manufacturers implement at least “SMS only over SGs” to enable roaming from Day 1*) – This does not preclude the implementation of “SMS over IMS”. Other solutions are not precluded – Recommend implementation of the “SMS only over SGs” feature in all networks to resolve the SMS roaming issue *) This solution is proposed because SGs requires only a new SW release deployment which seems more realistic than proposing IMS in all networks from day 1 for SMS

LTE Voice方案的比较 方案特点 应用场景 优势 劣势 1 CSFB－中间阶段 2 IMS&SRVCC－目标方案 方案特点 GSMA牵头在3GPP等开放标准的基础上，进一步限定IMS Voice Profile，用于促进多运营商/设备商间的漫游和互通。 需要部署IMS, 语音业务迁移到IMS，由IMS实现业务的控制。新建enhanced MSC和SRVCC AS支持LTE Voice到CS的切换 终端需要支持IMS和SRVCC。 方案已在3GPP中标准化 语音业务仍然在传统CS网络上开展, 属于中间过渡方案 CS网络需要升级或者新增网元 终端需要支持CSFB 应用场景 积极推进基于终端的LTE宽带数据业务 现网CS容量充足，不急于开展基于LTE／IMS的Voice业务 NGMN规定MNO网络需要支持CSFB，保证LTE业务跨运营商的漫游 积极推进基于终端上的LTE宽带数据业务和IMS语音、多媒体业务 热点地区CS容量紧张，需要借助LTE增大语音业务容量 希望调整原CS频谱，更多地用于数据业务 统一固定和移动的会话业务控制网络 优势 不需要部署IMS,新增网元少， 网络部署快 现网提供CS业务，用户业务感受一致 跨运营商接口少，易于实现跨网漫游 业务迁移到IMS/LTE网络，符合网络发展趋势 IMS实现统一的固定和移动业务控制，便于开展IMS创新业务 语音和LTE数据业务能够并行 One voice profile has been produced by AT&T, Orange, Telefonica, TeliaSonera, Verizon, Vodafone, Alcatel-Lucent, Ericsson, Nokia Siemens Networks, Nokia, Samsung and Sony Ericsson. NGMN - LTE voice IMS Voice is the final target of LTE voice solution. Support LTE and GSM/UMTS terminals, must provide CSFB function. Deploy LTE and GSM/UMTS, operators which provide non-IMS voice service must deploy CSFB NGMN LTE SMS solution UE must support SMS only over SGs SMS over IMS is not exclusive Other solutions are not exclusive Network is updated by software, support “SMS only over SGs”. VoLGA方案曾经在TMO主导的VoLGA中定义，3GPP未接受此方案，TMO已经宣布放弃VoLGA，转而支持CSFB和SRVCC。 VoLGA利用新增GAN IWF网元，再利用传统CS网络控制LTE语音，不需要部署IMS，也不需要升级MSC。终端需要支持VoLGA 劣势 IMS网络部署工作量大： CS业务需要向IMS迁移, BOSS系统改造工作量大 需要进行跨运营商的IMS漫游和互通的IOT工作 语音业务回落到CS域,呼叫建立时延略有增加 拨打或者接听电话,LTE数据业务会暂时中止,或者回落到3G PS. 移动网络未来需要向IMS演进,无线资源未来需要重新配置

CSFB和SRVCC共存的Voice业务域选择机制 基于终端和网络的能力指示, IMS签约信息决定UE接入的无线网络和享受数据、语音业务的方式。 语音和数据业务的优先级，用户能够在终端上设置 配置Voice centric的UE，需要保证能够进行Voice呼叫。如果驻留在E-UTRAN中不法开展Voice业务会退出E-UTRAN。 配置了"Data centric"的UE，当E-UTRAN不能提供Voice的时候，可以继续驻留在EUTRAN中。 Priority Data or Voice Centric in UE’s configure 语音业务的提供方式，一般由运营商通过在终端上预置或者设备管理方式修改；用户修改配置需要对运营商的网络有深入的了解 CS Voice Only，这种情况下，UE只用CS域发起voice呼叫，不允许使用VoLTE. CS Voice preferred，IMS PS Voice as secondary：这种情况下，如果CS voice可用，则使用CS域发起呼叫，否则使用VoLTE. IMS PS Voice preferred, CS Voice as secondary：这种情况下，如果IMS voice可用，则使用IMS发起呼叫。否则使用CS Voice. IMS PS Voice only：这种情况下，将使用IMS voice可用. 否则,配置了Voice centric的UE将选择并驻留在CS RAT；配置了Data Centric的UE驻留在当前LTE网络. Copyright © 1996 Northern Telecom Domain selection policy Operator can set UE’s configure by Device Management 网络在UE接入的移动性管理信令中，向UE指示当前接入网络的业务部署能力，供UE选择实际接入的无线域和语音业务的使用方式。 IMS voice over PS Session Supported Indication：网络指示UE拜访网络是否支持基于LTE/IMS的Voice。如果不支持，UE只能通过CS域做语音业务。 SMS only over SGs： 网络指示UE不支持CSFB,只支持SGs的短消息。终端或者使用LTE/IMS的Voice，或者驻留在2G/3G无线网络中，使用CS域语音业务。 CSFB not preferred ： 网络支持CSFB，但不希望voice centric的用户发起CSFB业务。 Network Capability Core Network indicate to UE on registration or TA update

LTE终端状态 Qualcomm支持LTE系列芯片 LG Revolution打通Verizon第一个LTE Voice Call Qualcomm MDM9600芯片 在2010年12月支持CSFB R8 with Redirection, 2011年3月支持CSFB R9 with Handover。 MSM8960 2012 年支持Smartphone. Sv-LTE 支持CDMA1X与LTE数据 的并发 LG Revolution打通Verizon第一个LTE Voice Call Verizon宣布2011年2月在商用网络上完成了实时的、基于IMS的VoLTE（LTE语音呼叫）。首次呼叫采用LG Revolution 4G （Android）智能手机，持续了33秒。在随后的呼叫中，技术人员能在语音通话时同时浏览网页和使用其他数据服务。 HTC支持语音的LTE Phone 在美国MetroPCS发布 美国MetroPCS在13个地铁城市部署了LTE，提供费用低廉的业务。在LTE上开展语音业务，将省出更多的频段用于Internet业务。 HTC Thunderbolt采用 Android 2.2操作系统。 ZTE GSMA移动亚洲通信年会演示CSFB ZTE携手CSL，使用第三方终端演示CSFB业务。 LTE手机终端开始出现 LTE商用终端与2010年相比有较大突破，逐渐从单模走向多模，数量也接近100款，但终端以路由器、USB Modem为主，手机 及其它智能终端数量较少。 根据GSA2011年6月的统计，全球已经有161款LTE终端，其中仅有8款手机终端，8款平板电脑，10款笔记本电脑，而路由 器63款、USB Modem 41款。 高通支持Sv-LTE(LTE/CDMA1X双待)方案, 使用LTE/CDMA1X双待手机, 语音和LTE数据业务能够并发.。

汇 报 提 纲 业界LTE VoIP方案的对比分析 IMS&SRVCC方案分析 CSFB方案分析 中兴通讯LTE VoIP解决方案 12

VoLTE成为业界共同的声音 One Voice Profile was released on Nov,2009. The One Voice Initiative was a collaboration between AT&T, Orange, Telefónica, TeliaSonera, Verizon Wireless, Vodafone, Alcatel-Lucent, Ericsson, Nokia, Nokia Siemens Networks, Samsung and Sony Ericsson. One Voice profile use current open standards to define the minimum mandatory set of functionality for interoperable IMS-based voice and SMS over LTE. GSMA Leads Mobile Industry Towards a Single, Global Solution for Voice over LTE 15 February 2010, Barcelona: The GSMA today announced it has adopted the work of the One Voice Initiative to drive the global mobile industry towards a standard way of delivering voice and messaging services for Long-Term Evolution (LTE). The GSMA’s Voice over LTE (VoLTE) initiative support the principle of a single, IMS-based voice solution for next-generation Mobile Broadband networks. The GSMA will also lead the development of the specifications that will enable interconnection and international roaming between LTE networks, and will complete that work by Q1 2011. The GSMA has widespread industry support for its VoLTE initiative Mobile operators supporting the initiative include 3 Group, AT&T, Bell Canada, China Mobile, Deutsche Telekom/T-Mobile, KDDI, mobilkom austria, MTS, NTT DoCoMo, Orange, SKT, SoftBank, Telecom Italia, Telecom New Zealand, Telefónica, Telenor, TeliaSonera, Verizon Wireless and Vodafone. VoLTE is a key enabler for the success of LTE Vicente San Miguel, CTO of Telefónica Telefonica plans to deploy LTE to provide an improved Mobile Broadband service for our customers. It is vital that we also deliver a high quality voice and messaging service. Telefónica has led the work with the GSMA on VoLTE and we strongly support this initiative to drive a common voice and messaging solution for the mobile industry, as it is a key enabler for the success of LTE.

VoLTE 目标架构分层逻辑图 SIP MAP H.248 GTP-U GTP-C Diameter Others Cr Sh Si ISC IM SSF SCS SIP AS BGCF MGCF Mr Mp Mb Mk Mx Mw Cx Mj Mg Mi Ix Mm Dx Mn IBCF I-CSCF BGCF IM-MGW TrGW P-CSCF S-CSCF SLF IMS Parter Control Plane IMS Parter User Plane MRFP MRFC HSS S3 S12 S11 Gx Gs：BSSAP+ IuCS Gb: BSSAP/BSSGP A S1-U C/D Gc Gr S6a S6d Mc S10 S4 IuPS Gn/Gp S5/S8 S16 SGs: SGsAP Sv Nc: ISUP/BICC Gi: IP Rx S1-MME: S1-AP SGi: IP Nb: NbUP PCRF PGW GGSN MSCS /VLR SIP MAP H.248 GTP-U SGW MME SGSN MGW GTP-C Non-3GPP Access Diameter GERAN UTRAN E-UTRAN Others Media/IP switch

Voice over LTE引入的前提条件 无线技术的发展 E2E的Qos保证 终端的支持 IMS&SRVCC是端对端的VoIP 无线分组技术不断发展，带宽增加, 每bit分组业务承载成本下降 E2E的Qos保证 基于分组网络传输实时业务,系统必须能够提供满足用户QoE要求的IP传输处理能力 终端的支持 必须要有多种商用的终端支持，供用户自由地选择终端 IMS&SRVCC是端对端的VoIP 讲解要点： 该页介绍现有的网络架构和维护手段，不能支撑业务精细化运营，一是传统的GGSN只能做到简单业务识别，不能实现业务的深度识别，缺乏数据业务精细化运营的基础；二是现有的维护工具性能统计不支持用户级别的统计，而话单分析由于缺乏详细的业务信息，也难以实现对业务和用户的感知 完善的业务控制架构 现有的CS网络不是端对端VoIP架构，必须要有基于PS的成熟的业务控制架构 与CS相同的业务体验 为用户提供不低于CS、甚至更加丰富的业务体验，才能够吸引用户在转换到LTE域的VoIP

无线分组技术快速发展 WCDMA Evolution IMT Advance 4G 2001-2005 2006 2007 2008 HSPA+ DL >40Mbps UL >10Mbps IMT Advance 4G GSM/GPRS EDGE 171/384kbps WCDMA R99/R4 384kbps HSDPA Phase I 1.8M/3.6Mbps HSDPA Phase II 7.2/14.4Mbps HSUPA 2M/5.76Mbps LTE DL:100Mbps UL:50Mbps CDMA2000 Evolution CDMA 1X 153kbps 1xEV-DO Rev. 0 DL: 2.4Mbps UL:153.6kbps 1xEV-D0 Rev. A DL: 3.1Mbps UL: 1.8Mbps DO Rev. B (MC DO) DL：46.5Mbps UL: 27Mbps WiMAX Evolution IEEE802.16d 20Mbps IEEE802.16e 70Mbps IEEE802.16m DL:100Mbps UL: 50Mbps 2001-2005 2006 2007 2008 2009 2010 2011 In the US, where local calls are mostly free and mobile telephony has not as successful as in Europe, mobile voice was about 14 percent of all traffic in 2003. 无线接入技术的带宽的量级从10k→100k→1M→10M→100M不断提升； 系统容量和分组业务频谱利用率的不断提升使得每bit分组业务承载成本逐步降低； In 10 MHz of spectrum, LTE VoIP capacity will reach almost 400； 多种无线接入技术共存，LTE扁平化网络架构提高分组传输效率，降低时延。

Real-Time (Live) Video 分组传输网络必须保证数据包传送的Qos Real-Time (Live) Video Loss,bit-error, and bandwidth Sensitive CS R4架构下运营商对承载网络和设备处理能力的要求 网络条件很好的情况MOS>4.0 网络较差条件时（丢包率=1%，网络抖动=20ms，时延=100ms）MOS>3.5 网络恶劣环境下（丢包率=5%，网络抖动=60ms，时延=400ms）MOS>3.0 Real-Time Voice Delay and Jitter Sensitive Video Gaming MTTF and MTTR Concerns Transmission Continuity IMS signal Delay and bit-error sensitive, On-demand Video Loss,bit-error sensitive IM and SMS Interactive message Low bandwidth 级别 MOS值 用户满意度 优 4.0-5.0 很好，听得清楚，延迟很小，交流流畅。 良 3.5-4.0 稍差，听得清楚，延迟小，交流欠缺顺畅，有点杂音 中 3.0-3.5 还可以，听不太清，有一定延迟，可以交流。 差 1.5-3.0 勉强，听不太清，延迟较大，交流重复多次。 劣 0－1.5 极差，听不懂，延迟大，交流不通畅。 不同类型的业务关注分组传输网络的不同指标 数据包的延迟和抖动对语音质量的影响比较大。 通过在接收端增加缓冲器可以有效地降低抖动的影响，但是却增加了网络延时。 随机类型的IP丢包（Random Loss）量少时对语音质量影响小；连续丢包（Burst Loss,指连续一个以上的数据包的丢失)对语音质量的影响是明显.

SAE架构提升数据传输质量 1 2 统一核心网 IP SAE的目标是提高数据传输质量 SAE的架构支持数据传输Qos的提高 统一控制承载分离 统一扁平化的网络 HSPA Non-3GPP LTE SAE-GW MME LTE/EPC IP eUTRAN 1 2 SAE的目标是提高数据传输质量 SAE的架构支持数据传输Qos的提高 提供更高的用户数据速率 减少时延,增加对实时业务的支持 提高系统容量和覆盖率,减少运营成本 多种无线技术互为补充 EPC网络层次扁平化，用户面节点尽量压缩，核心网用户面节点在非漫游时合并为一个。 LTE无线接入扁平化，取消RNC。 EPC对多种接入统一控制 提供端到端的QoS保证

SAE网元能够提供不同Qos等级的承载资源 QCI 资源类型 优先级 时延 丢包率 典型业务 1 GBR 2 100ms (-2)10 VoIP 2 GBR 4 150ms (-3)10 电话会议, 会话视频(直播流媒体) 3 GBR 3 50ms (-3)10 实时在线游戏, 实时工业监控 4 GBR 5 300ms (-6)10 非会话视频(缓冲流媒体) 5 Non-GBR 1 100ms (-6)10 IMS信令 6 Non-GBR 6 300ms (-6)10 视频(缓冲流媒体) 表格Copy自3GPP协议23.203 显示 Qci=1,5D的LTE承载能够满足VoIP 业务 Qos的需求。 7 Non-GBR 7 100ms (-3)10 视频(直播流媒体), 话音业务 交互式游戏 8 Non-GBR 8 300ms (-6)10 E-Mail, MSN, QQ, WWW P2P文件共享 9 Non-GBR 9 300ms (-6)10

PCC－会话类业务的质量保证机制 IMS EPC CSCF完会话类业务媒体属性协商后，向PCRF申请该业务流资源的授权 IMS AS PCRF根据业务类型，用户签约信息，承载网状况等数据生成该业务对应的策略，下发给SAE-GW IMS SAE-GW收到VoIP业务策略，为VoIP业务建立专有EPS bearer，并通过MME通告ENodeB SPR I/S-CSCF PCRF P-CSCF ENodeB接收到专用通道建立消息，将分配专门空口资源用于传输VoIP业务流 HSS eNodeB MME SAE-GW EPC 引入专门的业务QoS控制实体PCRF，依据业务类型，用户签约，承载网状况控制承载面为业务流预留所需的承载资源。 从终端到网关的E2E的专有VoIP语音承载通道（QCI=1，GBR类型）， VoIP承载通道QoS参数从管道（IP-CAN BEAR）开始逐层向下映射，确保每层都能提供所需的QoS。

IMS－提供分组域 Voice和多媒体业务的统一控制机制 VOIP Multimedia Telephone IM/PS/GM MRBT/MRAT All IP, access agnostic, flat network architecture Multimedia Conference File Transfer Web Call X-sharing Service Continuity … Easy for rich multimedia service development Convergent Centrex Enhanced Phone Book Office Collaborate Rich Services Open, standard system allowing for healthy ecosystem Resource control mechanism to avoid being a dump pipe Unified call control Unified subscriber profile Fixed mobile convergence enabling new service scenario PDSN Ideal service control platform for mobile broadband, e.g. LTE xGW GGSN BRAS AC LTE GSM/ WCDMA xDSL/ LAN WiMAX/ WiFi CDMA

IMS SR-VCC 保证LTE Voice能够切换到CS域 Nb and Nc/ Ai Nb and Nc,E PSTN/PLMN MSC RAN/GERAN A/Iu Nb and Nc Nb and Nc,E Nb and Nc/ Ai MSC SCC AS C,D Enhanced MSC MGCF/IMMGW IMS SV SGi SGi PCRF CSCF Voice continuity HSS S11 eNodeB SGi MME S1-U S1-MME SAE-GW MSCS enhanced for SRVCC provides the following functions as needed for support of SRVCC: - Handling the Relocation Preparation procedure requested for the voice component from MME via SV; - Invoking the session transfer procedure from IMS to CS as according to TS 23.237 [14]; - Coordinating the CS Handover and session transfer procedures; Handling the MAP_Update_Location procedure without it being triggered from the UE. MME additional function: - Performing the PS bearer splitting function by separating the voice PS bearer from the non-voice PS bearers. - Handling the non-voice PS bearers handover with the target cell as according to Inter RAT handover procedure as defined in TS 23.401 [2]. - Initiating the SRVCC handover procedure for handover of the voice component to the target cell. - Coordinating PS handover and SRVCC handover procedures when both procedures are performed, EPC Internet LTE初期孤岛式部署，利用2G/3G覆盖区作为E-UTRAN覆盖区的补充，基于IMS实现对LTE语音／多媒体业务的控制，用户在E-UTRAN覆盖区到2G/3G覆盖区的切换时保持语音连续性。 如果部署了3G PS，且终端支持DTM和PS Handover，UE PS域也会切换到3G PS；否则，PS域业务在切换后将被挂起。 要求部署具备SRVCC功能的Enhanced MSC 。MME支持VoIP和非VoIP媒体分离功能，并发起向MSC的VoIP媒体切换。IMS SCC AS完成从LTE承载向CS承载的媒体切换控制功能。

SR-VCC - VoLTE Establish Process PSTN/PLMN IMS Signal CS IMS Media GMSC RAN/GERAN SCC AS Other message MSC MGCF/IMMGW IMS Enhanced MSC I/S-CSCF P-CSCF HSS eNodeB MME SAE-GW EPC 1. LTE UE注册过程中，MME从HSS获取STN-SR and MSISDN 2.VoLTE IMS信令流建立过程 3.VoLTE 媒体流传输路径

SR-VCC-Handover Process IMS Signal PSTN/PLMN CS IMS Media RAN/GERAN GMSC Other message SCC AS MSC MGCF/IMMGW IMS Enhanced MSC handover I/S-CSCF P-CSCF HSS eNodeB MME SAE-GW EPC 1. eNodeB根据UE发出的无线信号测量报告，决策发起切换过程 6. Enhanced MSC向IMS域发起会话建立请求,被叫号码为STN-SR,主叫号码为MSISDN,会话路由到SCC AS 2.MME向Enhanced MSC发起PS向CS的切换请求 7-1. SCC AS发起会话远端UE的媒体地址更新,远端UE切换对端的媒体地址到MGW 3. Enhanced MSC向目的MSC发起切换准备消息 7-2. SCC AS 释放IMS会话分支 4. 目的MSC与无线侧之间的切换请求和响应 8. Enhance MSC发送PS->CS切换响应, eNodeB发出切换命令 5. Enhanced MSC与目的MSC建立承载 9. 终端接入UTRAN/GERAN，与目的MSCS建立承载

eSRVCC Architecture ATCF／ATGW VPLMN HPLMN IMS Signal VPLMN HPLMN IMS Media Other message RAN/GERAN SCC AS MSC Enhanced MSC handover I/S-CSCF HSS ATCF/ATGW eNodeB P-CSCF MME SAE-GW ATCF／ATGW 引入ATCF/ATGW逻辑网元，ATCF在发生切换的终端所在的拜访网络中锚定媒体流，因而媒体的切换不需要通过远端的IMS UE参与，避免因为IMS信令路径过长引起的SRVCC切换时的语音中断。 ATCF/ATGW部署在LTE终端的接入网络，即拜访地网络；如果LTE UE没有漫游，拜访地网络也是归属地。 ATCF是逻辑网元，3GPP标准规定实际可以部署在P-CSCF，IBCF，MSCS等网元中。

eSRVCC Procedure IMS Register IMS Session Setup LTE to CS handover P-CSCF将注册消息发送到ATCF，ATCF插入本网元地址作为V-STN-SR送到SCC AS。 SCC AS首先决定用户可以启动eSRVCC, 然后通过Sh接口修改IMS HSS中的STN-SR. IMS HSS通过内部接口修改EPC HSS中的STN-SR, 再插入MME. IMS Session Setup ATCF 被IMS网元引入信令中，并决定是否引入ATGW锚定媒体。 LTE to CS handover MSCS 通过V-STN-SR向ATCF发起新的IMS会话, CS 媒体锚定到ATGW。

汇 报 提 纲 业界三种LTE VoIP方案的对比分析 IMS&SRVCC方案分析 CSFB方案分析 中兴通讯LTE VoIP解决方案 27

CSFB Technique for the interim RAN/GREAN CS for call Service and LTE for internet service RAN/GERAN A/Iu MSC Nb and Nc/ Ai Nb and Nc RAN/GERAN SGs PSTN/PLMN CS A/Iu SGsAP MSC SCTP IP SGs SGs L1/L2 HSS S6a eNodeB S1-MME S11 MME SGi SAE-GW INTRENET NGMN stated firm commition to IMS Voice as the target solution for Voice over LTE. Furthermore, NGMN recognises the benefits of recommending a migratory non IMS voice solution that guarantees voice roaming with LTE handsets whilst allowing use of LTE data roaming in the same device. Thus NGMN recommends that: In all LTE devices that support GSM or UMTS and that provide the voice service, Circuit Switched Fall Back (CSFB) shall be implemented as a minimum requirement for supporting voice. If a Mobile Network Operator operates a network comprising LTE plus GSM or UMTS and if this MNO aims to provide a non-IMS voice service as well as an LTE data service to visiting subscribers utilising an LTE device, then it shall, as a minimum, support CSFB for voice. Infrastructure manufacturers, chipsets manufacturers and handsets manufacturers are requested to take this into account. S1-U EPC 适用于E-UTRAN覆盖区已经有2G/3G覆盖区的无线部署场景；CS域使用 的Voice，UDI Video业务，CS LCS，USSD等业务时，UE离开LTE网络，回退到GERAN/UTRAN的CS网络。传统CS 短信业务可以通过LTE接收。 MME和MSC之间存在SGs接口，类似SGSN与MSC之间用于CS/PS联合位置的Gs接口。SGs接口传递位置更新、寻呼、短消息等功能。

SMS only SGs in CSFB SMS transmitted in LTE access SMSC MSCS enhance for LTE SMS function SGs HSS S6a eNodeB S1-MME S11 MME SGi SAE-GW INTRENET NGMN stated firm commition to IMS Voice as the target solution for Voice over LTE. Furthermore, NGMN recognises the benefits of recommending a migratory non IMS voice solution that guarantees voice roaming with LTE handsets whilst allowing use of LTE data roaming in the same device. Thus NGMN recommends that: In all LTE devices that support GSM or UMTS and that provide the voice service, Circuit Switched Fall Back (CSFB) shall be implemented as a minimum requirement for supporting voice. If a Mobile Network Operator operates a network comprising LTE plus GSM or UMTS and if this MNO aims to provide a non-IMS voice service as well as an LTE data service to visiting subscribers utilising an LTE device, then it shall, as a minimum, support CSFB for voice. Infrastructure manufacturers, chipsets manufacturers and handsets manufacturers are requested to take this into account. S1-U EPC 要求终端能够通过LTE接收CS域的短消息，而不考虑语音业务做被叫寻呼的需求。 新建一个enhanced MSC for LTE SMS, 而不用对VMSC升级. SGs接口用于联合位置更新和传递基于LTE发送的短消息。MME收到SGs接口的短消息后，发送给UE。

3GPP Ctrl-Plane Location CSFB对各网元设备的要求 UE MSC LTE and GERAN/UTRAN Radio capability Combined Location updation CS fallback SMS over SGs procedures Maintain the SGs association SMS over SGs Roaming retry for CS fall back E-UTRAN Mobile Date Service Short Message OMA SUPL RAN/GERAN Telephone /video Call USSD Emergency Call 3GPP Ctrl-Plane Location MME E-NodeB Maintain SGs interface CSFB and SMS over SGs procedures Derived LAI and VLR number from TAI and IMSI Forward CS paging Direct the UE to the CS capable cell BSS 图需要更新，增加国外运营商LTE。 占据宽带无线接入市场的标准不是由技术体制本身决定的，而是由技术成熟度和市场规模决定的。 从全球运营商网络部署策略可以看出，目前主流的都是HSDPA网络，并计划在2008年升级HSDPA网络的速率，同时开始HSUPA网络的部署 据此分析，2006～2008是HSDPA网络的部署高峰期，2008～2009年将是HSUPA网络的部署高峰期 同时各大运营商在2008年底将开始HSPA+及LTE的测试，2009年HSPA+可能有策略较为激进的运营商开始试商用，LTE根据当前产业链判断商用时间在2011年及以后 If the network supports ISR, Forward Gb Paging message, even in NMO II/III In NMOII/II, the BSS is not configured to use PBCCH 30

1XRTT CSFB Technique 1XRTT CS for call Service and LTE for internet service 1X MSC 1XRTT PSTN/PLMN CS A1 1X MSC A1 S102 1X CS IWS HSS S6a eNodeB S1-MME S11 MME SGi SAE-SGW INTRENET S1-U EPC 适用于E-UTRAN覆盖区已经有2G/3G覆盖区的无线部署场景；UE 通过S1接口隧道传递1X CS的移动性管理和呼叫等消息。1X CS IWS网元与MSC之间建立S102隧道传输A1接口隧道消息。

1X RTT CSFB 替代方案- Dual Rx & Dual Radio 高通芯片支持 1X/LTE Dual Radio 1XRTT CS PSTN/PLMN 1X MSC HSS eNodeB MME SAE-SGW INTRENET EPC Dual Rx UE 特点 UE有2个并发下行控制信道，可以接收2个网络寻呼 CS承载语音，PS 承载数据； 1x优先：来自1x的寻呼和短信，会打断EVDO上的数据业务，用户转到1x； 优点： 核心网简单,不需要S102接口和IWS网元 缺陷：频繁的CS短消息挂起PS业务 Dual Radio UE特点 UE有2个并发无线模块，两个网络相互独立； CS承载语音，PS承载数据，业务并行； CS，PS和IMS同时可以并发使用，没有任何关联； 优点： 数据和语音并发, 核心网不需要S102接口和IWS网元 缺陷： 终端成本可能会略高；

汇 报 提 纲 业界LTE VoIP方案的对比分析 IMS&SRVCC方案分析 CSFB方案分析 中兴通讯LTE VoIP解决方案 33

ZTE CSFB Solution √ No need to deploy IMS MGW GMSC RAN/GERAN PSTN… SMSC MSCS CS call SGs E Short Message PDN(Internet) ZXUN uMAC: MME ZXUN USPP HSS eNodeB EPC ZXUN xGW: SAE-GW Data Service ZTE CS fallback solution highlight √ No need to deploy IMS √ Protect existing CS investment，Similar voice service experience √ 多种部署方案, 避免所有VMSC升级以支持SGs接口短消息功能

CSFB architecture and reference point RAN/GERAN MSC ZXWN MSCS/MGW Forwarding paging request and SMS to the UE Directing the UE to the target CS capable cell. A/Iu Supporting SGsAP/SCTP/IP protocol Maintaining SGs association user information Paging on SGs Supporting SMS on SGs Roaming Retry for mistaken mapping of TAI to LAI MME SGs eNodeB E-UTRAN and 3G UTRAN/GERAN dual mode Combined EPS/IMSI attach, update and detach. CS fallback and special SMS procedures S1-MME ZXUN uMAC Deriving a VLR number and LAI Initiating paging procedure for CS service Maintaining of SGs association towards MSC/VLR for EPS/IMSI attached UE

Long call establish time CSFB key issues High upgrade cost SAE-GW The existing MSC Servers/MSC covering LTE’s overlapped GERAN/UTRAN need to be upgraded for CSFB function. There may be some long-time working equipment difficult to be upgrade. MSCS MSCS MSCS SGs SGs MGW MME MGW MGW Long call establish time GERAN/UTRAN LTE GERAN/UTRAN GERAN/UTRAN HLR 5 3 TA1( under control of MME1) is overlapped with a small part of LA1( under control of MSCS1) and a large part of LA2( under control of MSCS2), but MME can only configures TA1 overlapped with LAI2. While UE triggers combined TA/LA update to MME, MME makes SGs associated LA update to MSCS2(step 1,2,3), so UE stores the LA2 as current LA. While UE fallback to CS( LA1), UE find the LAI information get from broadcast channel is different with LA2. So UE trigger a LA update to LA1 before it originate an MO-call or respond to an MT-page. So new inter-MSC LA update must be progressed in MSCS1(step 4,5）and then the call setup, which causes 1~2 seconds delay compared to an normal CS MO call. As CS terminated call scenario, MT call is first routed to MSC2 failed with no page response, then re-route to MSC1 after UE complete its location update on LA1. The three procedures "roming retry to MSCS2","LA update in MSCS1","re-route to MSCS1", will causes about 3~5 seconds delay. SAE-GW MSCS1 MSCS2 2 一、CSFB需要现有MSCS升级支持SGs接口； 二、Long call establish time,呼叫建立时延: 对于CSFB方案，联合位置更新的过程为MME根据TA推导LA，再推导SGs接口相连的VLR ，通常TA覆盖范围会比LA小，标准组织中建议一个TA尽可能在一个LA的范围之内。 但是实际部署中，如果一个TA刚好在两个LA1，LA2的交界上，MME根据TA推导出来的LA和VLR可能错误，导致UE在回落到CS域后，会在MO出呼和MT寻呼响应前首先进行位置更新。如果这种位置更新是跨MSC的，时延会很长。 <多个CSFB MSC联合组成MSC Pool时，能够扩大服务的LA范围，从而减少局间的位置更新，减少做被叫的二次路由过程。下页胶片介绍> MGW MME MGW LAI1 TA LAI2 4 1 36

CSFB 三种方案对比分析 现网VMSC升级 新建Proxy MSC 一般建议方案一，在现网没有能力升级的情况下，可以推荐方案三。 支持Iu-flex方式新建MSC 新建Proxy MSC GMSC C HSS/HLR GMSC GMSC HSS/HLR HSS/HLR C C Nc/Nb Nc/Nb D／C D Nc/Nb Nc/Nb S6a E S6a S6a D MME MME MME MSC MSC MSC SGs SGs Nc/Nb ProxyMSC A/Iu-cs EPC SGs CSFBMSC EPC EPC A/Iu-cs A/Iu-cs RAN RAN-Support A/Iu Flex RAN eNodeB eNodeB eNodeB 升级现网VMSC支持SGs接口 对VMSC以外的现网影响小 呼叫时延预计比普通CS呼叫长2s左右 如果现网BSC/RNC支持A/Iu Flex功能，可以通过新建CS Fallback专用MSCS和MGW方式支持语音业务 不需要对现网的MSC进行软件升级改造 呼叫时延预计比普通CS呼叫长2s左右 如果现网GMSC／VMSC/HLR支持Roaming Retry，可以通过新建一个Proxy MSC来支持语音业务，减少对现网VMSC的升级 如果现网GMSC／VMSC／HLR不支持Roaming Retry，可将CSFB被叫路由到新建MSC上。 呼叫时延预计比普通CS呼叫长4s左右 一般建议方案一，在现网没有能力升级的情况下，可以推荐方案三。 方案二涉及MSC Pool扩容，工程实施会比较困难。

ZTE SRVCC solution ZTE SRVCC solution highlight ZTE IMS Service Platform S OSA-SCS SCP IM-SSF SIP AS ZXUN RCP ZXUN HSS/HLR/AAA/SPR ZIMS Session Controller ZIMS Media Plane processor ZIMS Session Continuity Control I/S-CSCF MGCF P-CSCF AGCF E-CSCF BGCF MRF MGW eMSC SCC AS ZXUN uMAC ZXUN xGW SGSN GGSN AGW/ePDG SAE-GW MSCS MGW MME LTE GPRS HSPA WiMAX WLAN GSM WCDMA ZTE SRVCC solution highlight √Rich multimedia service on LTE √ Support multiple packet access √Fast voice handover from LTE to CS √Qos control for LTE voice

IMS SR-VCC Deployment Networking Rule Highlight IMS home Network Newly deployed MSCS for SRVCC handover, Multiple load sharing MSCS for failure backup EPC local breakout for VoIP Centralized Deployed for IMS control plane HPLMN IMS network ,P-CSCF may be in VPLMN No enhancement for existing MSCS High availibity SRVCC MGW,IM-MGW and VMGW can be co-deployed on one node VoIP media switched on local network SCC AS HSS MGCF CSCF Centralized Control Network MSCS SRVCC MSCS MSCS Local Network Local Network SRVCC MGW SRVCC MGW IM-MGW SAE-PGW MME MGW MGW SAE-PGW MME IM-MGW UTRAN/GERAN eNodeB UTRAN/GERAN eNodeB handover handover

MSCS enhance for SRVCC//MGW SRVCC architecture and reference points MSC ZXWN MSCS/MGW A/Iu Handling the Relocation procedure requested from MME via Sv Invoking the session transfer to IMS SCC AS Handling inter-MSC handover RAN/GERAN E,Nc MSCS enhance for SRVCC//MGW HSS eNodeB Mg LTE/UTRAN dual capability Support IMS client Indicate the SRVCC capability SCC AS SV S6 IMS S1-MME SGi SAE GW MME ZXUN SSS S11 Support Session Continuity Control Domain selection Indicate MME of SRVCC handover ZXUN HSPP ZXUN uMAC Get STN and MSISDN from HSS Perform bearer splitting function Handle the non-voice PS bearers handover as according to Inter RAT handover procedure Initiating the SRVCC handover with STN to enhanced MSC Insert IMS STN and C-MSISDN to MME

ZTE EPC/IMS智能化业务QoS控制架构 IMS control domain P-CSCF AGCF ZXUN SPR ZXUN RCP Service domain (VAS) meeting Video conference OA AF IM HSS VOIP/Centrex eNodeB ZXUN uMAC ZXUN xGW EPC 部署独立的ZXUN RCP，SPR产品,分层架构,基于标准的扩展，系统容易演进，兼容标准。 ZXUN RCP 支持从DPI设备上报的业务的策略控制；支持SP发起的业务策略控制（SP支持标准Rx接口）；支持none-AF based 业务控制；更支持基于标准AF发起的业务的策略控制 ZXUN RCP支持PCC&RACS融合，可以确保业务在不同网络间切换时的策略延续性，如业务QoS保障