�ݺ�ߣ

Editor's Notes

• #2: Hi everyone, my name is … and I am … at Telco Systems. Our webinar today will be about LTE mobile backhaul with SDN and NFV. We will look at the opportunities and challenges it presents, and answer the question “Why do operators need it?” [Next slide]
• #3: In this webinar we will discuss the following topics: current LTE backhaul – how its different from 2G and 3G backhaul, where the main challenges lie, and what the requirements are. Then we will move on to describe SDN briefly, and we’ll see how it can help the mobile operator. The next topic will be the “small cell” use case. We will focus on this use case as one of the most challenging use cases, and as one of the major technologies that will help mobile operators face increasing bandwidth demands. And last but not least, we will present an innovative approach that can help operators solve the challenges of small cell mobile backhaul. [Next �ݺ�ߣ]
• #4: The phrase ‘small cells’ refers to any cell size smaller than the traditional Macro-cell deployed for 2G/3G or LTE coverage. In other words, small cells are cells that offer a cell range from a few tens of meters to perhaps 500-1000 meters or more, depending on conditions like the type of area to be covered – including residential, enterprise, indoor or outdoor, higher or lower building density etc. Small cells allow for a very high deployment density of 4G/LTE and multi-mode (3G/LTE/Wifi) cells and thus provide very high capacities per geographical unit. There is a very large technical difference between the various cells listed here. While femtocells for homes are considered Consumer Premises Equipment (or CPE), that, in essence is similar to a home wifi hotspot or DSL or Cable router. Picocells, microcells and metrocells are carrier-grade units developed and delivered by the 4G/LTE mobile providers. They interface directly to the 4G/LTE core network over S1 interfaces and network management by the operator.
• #5: 2014-01-17
• #6: So the rules of the game are changing, and small is the new big, playing a bigger role in the investment of the mobile operator and in the network overall. And with this change, new challenges are being introduced. The amount of cells building the network, incorporating the small and the macro-cells, increases the scale of the network, which poses new management problems. This then requires higher levels of automation, due to sheer numbers, manual management becomes virtually impossible. Another element affected by the growth of the network is the signaling overhead. In LTE technology, on top of the connection between the base station and the MSC, there is a signaling interface called S1 interface. But there is also base station-to base station connectivity, called X2 interface. This means that at each site there is 1 S1 interface and few X2 interfaces creating higher overhead on which the backhaul network needs to support. Availability is another aspect that needs to be looked at. While many of the small cells will reside within macrocells they have, in a way, backup coverage in case of failure. However, the load and the timing of that load, for example at an important sports event, will make it harder for the network to serve as backup. All of the these challenges, together with the availability of the connectivity to the small cells and the ability to actively and remotely manage and monitor the network will dramatically affect the backhaul.
• #7: In a typical deployment we will see a few small cells installed in each macro / metro cell site. Usually they will be deployed at the network edge, especially in busy places like shopping malls and business centers, but also in poor “reception holes”. These are the places where a low reception condition causes low radio utilization, and by installing a small cell the operator can improve its RAN efficiency. The small cells will be using different access technologies to connect, mainly wireless and fiber. The mix between wireless and fiber-based access is often debated in the industry, and in many cases will be dictated by the availability of infrastructure, the local conditions and location of the small cells vs. the macrocell. Even in the wireless domain, few alternative technologies exist that vary by frequency, bandwidth, range, and PtP vs. PtMP nature. The deployment scenario will also be affected by the entity providing the backhaul: Different infrastructure owners like the mobile operator own infrastructure, wholesale providers or even a dark fiber providers. As it is today, we believe that at the macro site, the wholesale providers market will continue to grow and even expedite its growth, allowing new players like utility and small city carriers, as well as the incumbent service providers, to increase their mobile backhaul market share. Some of the leading wholesalers like Zayo in the US, and Colt in Europe have already identified this market as an opportunity and are expanding their small cell offering even more. Installation-wise, most of the small cells will be installed outdoors, some installed on a light pool, some on high building roof tops. Real estate and space is not trivial, and therefore requires the small cell physical dimensions to be very small.
• #8: © Ericsson AB 2012
• #9: © Ericsson AB 2012
• #16: [Next �ݺ�ߣ]
• #17: But first let’s explain some of the acronyms: Network Functions Virtualization or NFV is the name of a new and evolving networking technology. This technology utilizes the virtualization concept and the increasing computing power of the general processing CPU, to run some networking functions (for example, routing, or firewalling) at a different network location where they are run today, using COTS servers and software instead of purposely designed appliances and hardware. Of course not every device and every network position can be virtualized, as the CPUs still find it hard to compete with the dedicated ASICs, especially in high speed and capacity. Still, there are enough applications that can be virtualized. VNF or virtual network function, is a specific networking function we want to virtualize. So a specific platform or CPU can run several VNFs, in parallel. An example for such a function is routing. A virtual router is a software product processing the L3 packets and assigning them into a virtual interface, that is later mapped into a physical one. The last definition I want to go over is distributed NFV. Since NFV initiated around heavy processing functionalities, a lot of the applications were focused around the network core and virtualizing it into the data center. However, out of this trend even more interesting technology has emerged. Distributed NFV is all about utilizing the NFV concept in a distributed manner, or in other words, placing the computing resources at the edge of the network in a distributed architecture, running the different VNFs on these platforms. Locating the processing resources at the edge of the network can have huge benefits in terms of the locality of the service, security and network load. So how does this relate to the small cell L3 mobile backhaul challenges? [Next slide]
• #18: But that’s not the all of it. If we are installing a distributed NFV platform as a cell site router we are really opening up the cell site to innovation by allowing multiple VNF to run and provide many types of services. For example, we can install a virtual CDN allowing to cache in some frequently used content thus reducing the networking load and dramatically improving the customer experience. Another application to keep in mind is virtual tunnels. For example, creating IPSeC tunnels from the main cell site to the EPC (and offloading this functionality from the small cells or enabling it to non IPSeC supporting small cells). So by utilizing the processing resources for running several VNFs, we can improve the business case dramatically, and enable more innovative usage of your networks. [Next �ݺ�ߣ]
• #19: Installing a Distributed NFV platform as the cell site router provides an innovative solution that utilizes the distributed NFV to deliver the L3 transport in a virtual manner. With this architecture, the operator can install a small cost-efficient entirely outdoor switch and run a virtual router instance. This way the data that comes from the base station is sent via L2 transport to the macro cell site. Once the packets reach the cell site router, it is then sent into the additional processor, mapped to the right vRouter instance, then it is getting routed into the right virtual interface and transmitted into the physical network. This innovative solution can save a lot on the CAPEX but also greatly simplifies network operation costs. Another benefit this solution provides is that the L2 backhaul can utilize a lot of the standard L2 OAM tools that do not exist in L3 (like EFM, CFM, Y1731), and provides a clear SLA visibility into the small cell backhaul. In the next slide we will explore how this new concept changes the small cell business case [Next slide]
• #20: [Next �ݺ�ߣ]
• #26: So to conclude, mobile backhaul is becoming a major challenge in 4G / 5G networks due to the packet based technology, the capacity, resiliency and QoS requirements. SDN is a new and emerging technology that can help mobile operators overcome this challenge by allowing a centralized network view, optimizing the network resources and enabling “best of breed” product selection on every segment. Distributed NFV and an innovative new approach can create a true revolution in the mobile backhaul, and presents a very clear CAPEX and OPEX advantage over traditional approaches. With that I want to thank you all for attending…..
• #28: One of the interesting questions that is asked is what will be the predominate mobile backhaul technology. I mean beside the obvious fact mobile backhaul is turning completely packet based, still many struggle with the question should they choose L2 or L3. L3 transport looks more natural for some of the mobile operators when transitioning to all IP mobile network, while other prefer the consistent manner and simplicity of the L2. there is obviously a price difference as L3 devices are more expensive from L2. when small cell deployment is considered, this price difference is becoming critical and can endanger the small cell business case. To demonstrate it we’ve taken an averaged price estimation based on price lists (not the transfer price), of small L2 and a high density L2 switch and small and high density L3 router. As you can see the total site backhaul solution is almost 4 times more expensive when going L3. [Next slide]
• #29: As you can see the distributed NFV solution is 40% of the cost of a native L3 solutions. It is clearly a path towards implementing the L3 functionality into the small cell mobile backhaul. [Next �ݺ�ߣ]