Latest Trends in Optical Networking: 100G & Next-Generation ROADMs
Rajan VaradarajanFriday, June 18th, 2010
There has been a renewal of interest in optical networking companies recently with the anticipated shift to 100G technologies poised to happen soon. Various players are positioning themselves for this big industry shift. Along with this interest in all things 100G, there is also an industry evolution from the early generations of ROADM technology (Reconfigurable Optical Add Drop Multiplexer) to newer versions that provide a significantly higher degree of flexibility and reconfigurability.
Reconfigurable Optical Add/Drop Multiplexers (ROADMs) technically refer to a network element that has the capability of adding/dropping selected wavelengths for local traffic as well as redirecting express traffic to other directions in a multi-degree node. The “Reconfigurable” part in ROADM refers to the ability to do this local add/drop as well as the redirection of express wavelengths in a dynamic fashion from a remote network operations center and without manual re-fibers of line cards and with minimal pre-planning to account for uncertain traffic patterns. This reconfigurability is the key to efficient network adaptation to unforeseen demand patterns and customer connects/disconnects without costly service disruptions. The term ROADM, has been used more broadly than just the network element itself and has been used to refer to (a) an optical transport system that incorporates ROADM technology, and (b) optical components and sub-systems (made by companies such as JDSU) that form the building blocks of the ROADM system.
100G networking and associated optical products can be broadly classified into (a) client side, short-reach, standardized, pluggable optical modules that are used to connect short reaches between routers/switches and transport equipment or among switches, and (b) line side models, DWDM interfaces that are proprietary, vendor specific and cover distances of 1,000 - 2,000 km.
Early low-volume shipments of client side interfaces, which are standardized by IEEE, have commenced and the key players for these interfaces are Finisar, Santur and Opnext. The client side models have not yet reached attractive price points compared to 10G modules, hence slow uptake of the models is projected until the cost-volume positive feedback cycle kicks into higher gear. While early parts are sampling this year, rapid market adoption is highly predicated on the cost points of the modules.
On the DWDM line side that covers geographical reaches across regional, long-haul or ultra-long haul distances, there has been a concerted effort by all major optical equipment vendors to introduce 100G capable systems. The advantages of 100G on the line side are very compelling, as it increases the fiber capacity to 8 Tb/s and the preferred technology path using coherent optical transmission provides a number of additional advantages of simpler link design and inherent compensation of some fiber impairments such as chromatic and polarization mode dispersion. The advantages of 100G systems and reasonable price points relative to 10G systems will facilitate line adoption of 100G; it is expected to occur rapidly once systems are available in early to mid 2011. While early versions of 100G line side technology are available now, optimized and field deployable systems are expected in 2011. The major players in this space are Ciena/Nortel, Alcatel-Lucent, Nokia-Siemens, Huawei and Infinera. Each of the vendors has announced slightly varying flavors of the technology and approaches. With the R&D focus on 100G and coherent technology and the commoditization of 10G technology, it appears that 40G technology is being squeezed out in carrier applications (for both line side and client side applications). This was reinforced by Infinera’s recently announced decision to discontinue 40G non-coherent PIC (Photonic Integrated Circuit) and focus resources on 100G coherent technology in a PIC.
Along with the evolution from 10G to 100G discussed above, the other major area of interest in the optical networking space is the emergence and adoption of next-generation ROADMs that provide colorless, directionless and contentionless capabilities. The first generation of ROADMs introduced in 2003-2004 enabled the carriers to add/drop some wavelengths without disrupting other wavelengths. While this provided a huge improvement over the previous method of operation, it still had a number of limitations in terms of a fully automated reconfigurability. Newer optical building blocks and lower costs are enabling the realization of next-generation ROADM architectures which remove these limitations. Directionless ROADMs enable a common bank of transponders to connect to any direction in a multi-degree node. Colorless ROADMs enable a transponder to flexibly connect to any mux/demux port and contentionless ROADMs enable use of the same wavelength in different segments of a network with a common node. The key building blocks for these new levels of reconfigurability is higher port count WSS (Wavelength Selective Switches) and smaller, lower cost WSS switches. These advances are expected to catalyze the ROADM market and continue the high growth trajectory of this sub-segment of the optical component space and continue to benefit ROADM market leaders like JDSU.
![[Post to Twitter]](http://www.pg-research.com/blog/wp-content/plugins/tweet-this/icons/tt-twitter.png){kind=icon}
Posted in Internet, Networking, Rajan Varadarajan, Technology, Telco | No Comments »
Femtocells
Rajan VaradarajanFriday, March 26th, 2010
Femtocells have seen renewed interest due to announcements from carriers as they try to increase their coverage within office buildings and homes where the cell strength is an issue. At CTIA this week, AT&T announced that it is offering a residential femtocell in collaboration with Cisco for $149.99. T-Mobile, Verizon and Sprint already include femtocells within their suite of offerings. In Europe, Vodafone is one the carriers offering femtocells.
Femtocells are indoor devices that operate in licensed spectrum and connect to the cellular backend infrastructure using the broadband connection in the home or office. They guarantee “5 bars” of coverage within your home and can be used to carry voice and data sessions over the broadband network. In addition, you can seamlessly move the voice call from the femtocell on to the “regular” cellular network when you move out of your home.
The femtocell (also know as the Femtocell Acess Point or FAP) is one of two components that enable the home user to connect to the cellular network. The second component is the femtocell gateway which is used to aggregate femtocell traffic from multiple FAPs. This node is part of the carrier network and can aggregate thousands of FAP connections. The FAP and the gateway are often from different vendors. For example, the Airvana UMTS FAP is certified to interoperate with the Nokia Siemens Network femtocell gateway.
FAPs can be simple or complex with high degree of functionality. FAPs can be lightweight or include a high degree of functionality. For example, the Alcatel Lucent 9365 Base Station Router Femto product line includes the 3G Node B (i.e. base station) and Radio Network Controller (RNC) functionality. In addition, the FAPs act in a collaborative manner across the enterprise when multiple FAP devices are deployed. This includes an auto configuration/plug and play feature. FAP vendors include Airvana, Alcatel-Lucent, Ericsson, Huawei, Samsung and Ubiquity, among others.
Some phones use the home or office WiFi network for connecting to the data network when the cell phone signal is weak. This is more common for data traffic though it is possible to use VoIP over WiFi for voice calls. FAPs, in contrast, are “micro cellular base stations” – so the phones continue to operate over the licensed cellular frequency when connecting to FAPs. Carriers claim that their spectrum licenses make them the only permitted providers of these femtocells, though YMax, the parent company of MagicJack has indicated that the carrier spectrum licenses do not extend into the home.
The disadvantage of femtocells is that they tie up the bandwidth on the home user’s broadband network. Add to this the original reason for femtocells – poor carrier’s coverage inside the home or enterprise – and you end up paying more for a carrier shortcoming. Third, it is not clear if this is a good financial proposition for the carriers, especially if they end up subsidizing the cost of the femtocell. Finally, truck rolls to help debug installation issues with femtocells might end up costing the carriers much more than they anticipated.
All that being said, femtocells are increasingly being seen by carriers as an area of revenue, while fixing the technical issue of cell coverage. For the femto equipment vendors, close supplier relationships with the carriers is the key to success.
Posted in Author, Rajan Varadarajan | No Comments »
Mobile Packet Core — Ready to Take off
Rajan VaradarajanWednesday, February 10th, 2010
Renewed interest in packet core networks over the past several months has generated new product announcements from infrastructure equipment vendors and touched off a round of acquisitions in the space as well.
I should say that packet core networks are not new. They’ve been used to backhaul IP data traffic to servers and other nodes on the Internet since the days of 2.5 G/GPRS networks. The central player in the packet core network is the Gateway GPRS support node (GGSN) in GSM/UMTS networks — a powerful router that performs multiple functions including user authentication, traffic forwarding and management, support for thousands of sessions, data records/billing, tunneling across multiple GGSNs, etc.
That said, the packet core network’s new prominence is the result of a phenomenal increase in mobile data applications and traffic driven largely the result of a multitude of all-you-can-eat data plans from carriers. Another significant factor driving demand for packet core routers is WiMAX, which is being adopted at a great rate for fixed and nomadic IP data applications, especially in developing nations. The ASN gateway packet node is the key component of WiMAX’s core and, although not a cellular network substitution or replacement, there is a sizeable overlap in the functionality of the ASN gateway and cellular networks’ packet core nodes.
Looking ahead, as the industry moves through the transition to Long Term Evolution (LTE), often categorized as 4G technology, the packet core needs equipment that not only meets today’s amped-up capability requirements but must accommodate future expansion as well. Interestingly, several packet core node vendors are targeting sales to the LTE core network, also known as the Evolved Packet Core (EPC). Unlike today’s 3G networks, EPC provides an “all-IP” aspect that carries voice traffic over IP.
Girding up for what’s next
Packet core equipment vendors include the traditional mobile infrastructure market leaders, among them Ericsson, Nokia, Siemens, Alcatel-Lucent, Hitachi (which recently acquired Nortel), Huawei and ZTE. But smaller startup companies are also making their presence known. For example, WiChorus saw initial success in the WiMAX space and enhanced its product line with offerings for the LTE Market. Tellabs recently acquired WiChorus and will integrate its product into the highly successful 8800 Multiservice Router.
Cisco also acquired a small startup, Starent, in response to its recent success with Verizon. Starent’s ST-series Multimedia Core Platforms connect to multiple types of access networks, including UMTS, WiMAX and LTE EPC. The core software runs StarOS, a variant on Linux, and the platforms incorporate hot swap capability, redundancy, and a variety of other carrier class features. Unlike the larger vendors, Starent and WiChorus offer software systems that are easier to scale for various uses and capacities. Their core products are purpose built and designed from the ground up.
A number of equipment vendors, including Hitachi, NEC, NSN and ZTE, offer modified Advanced Telecommunications Computing Architecture (ATCA) platforms as the basis for their packet core nodes. These “Big Iron” packet core nodes have a multi slot chassis populated with high capacity cards and often have network processors/ASICs/switching devices. The network processors/ASICs are useful for deep packet inspection (DPI) that, allows the node to shape/police/report traffic based on embedded content although no operator will publicly admit it. Companies such as Ericsson and Starent rely on custom-built platforms to address this market.
Ready and not
It will be interesting to see Juniper’s roadmap for the packet core network as it is fleshed out over the next few months. Juniper was supposed to partner with Starent, but instead was acquired by Cisco. Juniper also partnered with Ericsson for the 3G packet core (GGSN), but won’t be involved on 4G equipment. And Ericsson’s 2006 acquisition of Redback appears prescient as it provides credibility for the company’s IP technology with the service provider network.
Some reports have Alcatel-Lucent addressing the packet core by adding functions to the 7750 service router platform, which they gained in the TiMetra acquisition and saw significant success with in the metro Ethernet space. This also has high credibility with service providers.
Lastly, Huawei’s success in Europe and Asia does not make it a major player in the North American market. Observers indicate Huawei’s presence depressed bids and profits of European manufacturers of next generation packet cores.
Tags: Alcatel-Lucent, Ericsson, Hitachi, Huawei, Juniper, NEC, Nokia, Nortel, Siemens, Starent, WiChorus, WiMAX, ZTE
Posted in Author, Networking, Rajan Varadarajan, Telco | No Comments »
Calix: Positioned Well Enough on the Road to IPTV?
Rajan VaradarajanMonday, January 4th, 2010
Calix, a strong player in the growing wireline access (GPON/DSL/IPTV pipes) segment has a singular focus on Tier-2/Tier-3 markets, which in the near-term bodes quite well as the Tier-2/3 segment is in line to receive most of the funds made available through the federal government’s Broadband Stimulus Program to build out the nation’s broadband network, part of the multi-billion-dollar American Recovery & Reinvestment Act’s (ARRP). Tier-1 companies largely passed on Broadband Stimulus.
Calix focuses on wireline access solutions of all types: DSL, GPON and ActiveEthernet. The company’s systems allow service providers to go beyond mere connectivity and reach for the brass ring of extended communications services and revenues that is driving the transformation from legacy circuit to packet; from narrowband to broadband; and from copper to fiber networks.
Wireline is the true grail. Industry headlines may all be about next-generation wireless but wireline access deployment is the key to the reliable, cost-effective, very-high-speed pipes that are the foundation for Telcos’ entry to the IPTV space.
Calix provides the entire solution ecosystem for the wireline systems. Though it started as a DSL/next-gen voice company called Broadband Loop Carrier, Calix added GPON capabilities through the acquisition of Optical Solutions Inc., and carried out a technology refresh to evolve its product portfolio from ATM to IP/Ethernet. There’s a lot to like: Calix enjoys a strong market position and is the market leader in the Tier-2 segment with good customer traction. The company has a strong, comprehensive product portfolio that is well positioned to address all wireline access market needs. As noted they are a beneficiary of Stimulus funds over the next 12-18 months and Calix is aggressively addressing growth markets such as IPTV and GPON.
But Calix also has significant challenges. The company has not succeeded in expanding into Tier-1 or international accounts where strong, well-established players maintain deep, long-standing relationships with their customers. As a result, its upside potential is limited because once ARRP infrastructure funds to Tier-2/3 carriers are deployed, Calix has little opportunity to grow at a rate higher than the organic growth of the Tier-2/3 segment in the U.S. Add to that concern over the long-term competitiveness and operational viability of many Tier 2 & 3 carriers, Stimulus funds notwithstanding, and the wireline access market is fiercely competitive with margins under 30% and relatively high R&D costs.
Overall, this is a growth market characterized by a few technology shifts. Verizon’s bold FTTP initiative has spurred a deep-fiber build out by many other carriers, enabled to a much greater degree by the ARRA Broadband Stimulus program. Calix stands to gain significantly from this build out as it has focused on Tier-2 and Independent Operating Companies (IOCs)/Rurals, again, those that are in line to receive the bulk of the funds.
The next-gen wireline access market is segmented by types of customers, and different players have strengths in different segments. Also, with almost no international presence, Calix is viable only in North America. Still, Calix’s focus on the Tier-2/Tier-3 market of North American wireline carriers brings a long customer list to bear that includes CenturyLink, Windstream, TDS Telecom and many other smaller companies.
Looking at the competitive landscape, the Tier One customers (AT&T, Verizon, Qwest) are largely the domain of Alcatel-Lucent, Motorola, Adtran and Ericsson. The Tier-2 customers (CenturyLink, Windstream, Frontier etc.) are serviced by Calix and Adtran and Tier-3 customers (IOCs/ Rurals, etc) see Calix, Enablence and Occam.
Tags: ActiveEthernet, AT&T, Calix, CenturyLink, DSL, Frontier, GPON, Qwest, Verizon, Windstream
Posted in Author, Rajan Varadarajan, Telco | No Comments »
Docsis 3.0 and its implications
Rajan VaradarajanFriday, October 30th, 2009
The Data-Over -Cable-Service-Interface-Specification version 3, more widely known as DOCSIS 3.0 or D3, is a cable industry standard for deploying broadband Internet service over HFC (Hybrid Fiber Coax) cable plant. The standard is managed and updated by an MSO (Multi System Operator) supported consortium called CableLabs based in Colorado. CableLabs defines the specification and certifies vendor products for interoperability. D3 specifications define technologies such as channel bonding and IPv6 which allow MSOs to deploy 100Mbps+ Internet service and cable IPTV. The products that must be D3 certified are CMTS (Cable Modem Termination System) at the operator premise and cable modem at the customer premise.
It has been more than three years since CableLabs issued the D3 specification in August 2006. D3 has changed the playing field in broadband Internet services, giving life to the cable infrastructure once thought to be out-dated. With D3 deployment well underway, cable has a defendable position against Telco FTTH/VDSL and potential for new advanced services. Cable Operators worldwide are gaining more subscribers through D3, and vendors stand to benefit from this growth for many years to come.
D3 was borne out of the need to answer Telco’s VDSL2 and FFTH (ex: Verizon FiOS). Prior to the D3, the channelized nature of the cable spectrum limited broadband speed to 40Mbps. Channel bonding in D3 allows MSOs to pump up speed in multiples of 40Mbps. In theory, MSOs can get 80Mbps, 160Mbps, 320Mbps, and even reaching 1Gbps. To deploy the service, MSOs need to either upgrade existing CMTS with new blades and software, or buy brand new D3 CMTS chassis. On the customer premise side, there is no need to change cable modem if the subscriber does not want D3 service. D3 CMTS is backwards compatible and supports older DOCSIS cable modems.
Deployment started even before CableLabs issued the specifications in 2006, and unlike previous Cable technologies, this one started overseas in places like Korea and Japan where the wire-line competition is fierce. Today, D3 installations are covering a vast number of homes in North America, Europe, and Asia. The tier 1 MSOs that have deployed D3 are Comcast (Cisco, Arris), TWC (Cisco, Arris, Moto), Liberty Global (Casa, Cisco, Motorola), Cox (Cisco, Motorola), Cablevision (Cisco), Virgin (Motorola), NumeriCable (Cisco), Ono (Cisco), and Rogers (Cisco).
Strangely, the D3 CMTS vendor community has shrunk over the years. Since the inception of the D3 concept, vendors like Juniper, Terrayon, and BigBand have shut down their respective CMTS businesses. The heavy engineering investment and the delay in D3 specifications made it difficult to maintain a reasonable P&L, particularly for vendors with a limited cable portfolio. However, D3 deployment in the last two years has rejuvenated the industry, and today we have a community of three dominant players (Arris, Cisco, Motorola) and a startup (Casa Systems). In the coming years, new CMTS vendors will surface from the QAM community with companies like Harmonics and RGB, and from the IP community with companies like ALU, Huawei, and Ericsson. As D3 subscription picks up, cable modem vendors will see upside also, but this is unlikely to happen until 2011. Today’s D3 subscription is less than 1% in companies such as Cablevision and Comcast.
In the coming years, MSOs will purchase even more D3 ports to beef up capacity for Cable IPTV. Cable IPTV, or Video over DOCSIS, demands much more network capacity than the current data service which is mainly tuned for emailing and surfing. MSOs will double or even quadruple the current DOCSIS capacity in order to deliver new service like “TV-Everywhere.” Operators like Comcast are reportedly working on a new architecture whereby DOCSIS will be the primary delivery protocol in the HFC spectrum. Today, QAM uses more than 90% of the cable spectrum to deliver TV channels to homes while DOCSIS varies from 5%-10%.
What is next after D3? PON has been in discussion but MSOs will likely focus on upstream bandwidth. This is the Achilles Heel for Cable and even with D3 upstream channel bonding, it is limited to 120Mbps, while FTTH can get as high as 1Gbps. This weakness will hurt cable in the growing enterprise subscribership. Work is underway to look at how to re-design the 4Gbps cable spectrum to enable more speed on the upstream side.
Posted in Author, Rajan Varadarajan | No Comments »
Identity Management Challenges and Best Practices
Rajan VaradarajanThursday, September 24th, 2009
Introduction
With the wide array of web-enabled devices and increased usage of contingent work-force, identity management has become a huge business challenge. This challenge is further amplified by disparate security devices and lack of coordination between these devices. Simply identifying the user by login credential and the device by its IP address is no longer enough.
Identity enabling the network not only ensures that the outsourcers and contractors are able to access the network resources they need to fulfill what the company hired them for, but also makes sure that these users are not able to access sensitive data, which is critically important for company’s long term sustained competitive advantage. Identity enabling the network also ensures that trusted users behave appropriately and in any unfortunate event of insider breach, proper logs and audit trails are available for forensic experts to look into.
The ideal identity management solution should capture and correlate all the access parameters and provide a holistic view to the security administrator who can take an informed decision. Having identity management as the center piece will also help administrator configure policy information based on access parameters.
Challenges
Access Parameters: Access parameters can be classified as follows:
- Who is accessing the network? : Includes username or network identification of the person accessing the network. Typically this information is captured by AAA server infrastructure.
- How the network is being accessed? – The ubiquity of web-capable devices makes it important to understand how the network is being accessed. Is it via a laptop or handheld? Is it a clean system?
It’s also important to understand whether the device is a corporate asset or not. To understand the vulnerability of the machine it’s important to determine the operating system. Typically, a Window-based device is more prone to malware and viruses than a Mac or Linux device. This information about the client device helps IT risk managers evaluate organization’s exposure.
- What is being accessed? – It’s important to understand what resources are being accessed. With more and more resources getting virtualized or moving to the cloud, identifying the resources by IP address is not enough.
- Where is the network being accessed from? – Is the user on office LAN or remote? If the user is remote it becomes critically important to ensure the confidentiality of data. The security administrator must be able to ensure that a secure channel exists between the client and the server.
- When is the network being accessed? – This area is often overlooked by organizations. The time stamp associated with the access of the network helps identify whether network access is during office-hours or not. Specifically, personal web-browsing during business hour can be very dangerous and could bring down the entire network to halt.
Organizations find it difficult to tie these parameters together into a single, secure Identity Management system. Sometimes the very organization of the IT department hinders a security administrator’s ability to combine these parameters usefully but, more often; it is lack of coordination among the organization’s various security entities that pose the greatest obstacle to identity enablement.
Typically, organizations deploy a firewall as intrusion detection and prevention system to block unauthorized access and to detect and prevent attacks against a specific host (host-based IDS) or the network (network-based IDS) and an AAA (Authorization, Authentication and Accounting) infrastructure to authenticate and authorize any user.
A lack of coordination between these security devices hinders a security administrator’s ability to act quickly in case of any attack. Identity enabling the network ensures that not only every person and device accessing the network is identified but also that disparate security system act in a coordinated fashion to thwart attack.
Industry Implementations:
Identity enablement objective can be achieved by either deploying commercial identity management software or by rolling out Network Access Control (NAC).
Typically, software vendors such as Microsoft, IBM, Oracle, Sun Microsystems and Novell offer Identity management software which also offers single sign-on (SSO). These are cheaper, offer deployment flexibility, provide robust single SSO functionality and tight application integration. But they do not provide network visibility and difficult to mitigate against insider threat
NAC Vendors such as Cisco and Juniper offer identity enablement by providing integration among various network elements - Firewall, Intrusion detection system or 3rd party network appliances. Cisco has taken the approach of providing NAC APIs to 3rd party vendors whereas Juniper is focused on integrating its Switch, firewall and IDP product-lines. These are easier to mitigate against insider threat since network vendors can control all access gateways. For example, NAC appliance can get information from IDP and block the user by dynamically configuring firewall or switch. And they are easier in terms of getting a top-level view of security status. However, they are costly and provide lesser deployment flexibility and vendor specific.
Tags: identity enabling, identity management
Posted in Rajan Varadarajan | No Comments »
State of the Telecom Optical Components Industry
Rajan VaradarajanFriday, August 14th, 2009
The genesis of the telecom optical components industry, as we know it today, was seeded in the research labs of major equipment telecommunications vendors of the 1980’s, companies such as Nortel, Lucent and Alcatel. The optical components that were developed were the key to winning the race in DWDM and OC-192. Those that had the components technology or the scale to drive the supply chain, dominated the burgeoning optical transport market of the late nineties into early 2000. These companies drove component business groups to serve the internal market and feed into the system - level product lines – thereby ensuring market differentiation. Some companies drove these component groups as cost centers and the profit margin went to the system business.
When capex budgets started to dry-up by the middle of 2001 and the frenzied deployment of 10Gb/s DWDM long-haul systems reduced to a trickle, the obvious and immediate fix to improve finances were to reduce fixed costs associated with component design centers and optical chip wafer Fabs. Deals were quickly made to divest and merge the captive component divisions with either existing merchant suppliers or with start-ups that were able to catch the IPO window. Since the burst of the telecom bubble, the industry has slowly clawed its way back to some level of recovery even with the emergence of new component suppliers such as Optium, Opnext and Finisar. However, with remnants of VC-funded start-ups sputtering along, the emergence of low cost manufacturers from the Asia-Pacific region along with general oversupply in the industry, the return to profitability has slowed with some companies suffering occasional losses. Even the minimally profitable companies continually face the challenges of restructuring and off-shoring while addressing the need to further invest in leading edge optical products.
Not all is doom and gloom however. Bandwidth growth has allowed service providers to expand their infrastructure on real revenues coming from bandwidth demand. Spending levels in 2005-2008 were at a high in both terrestrial and submarine networks and component company valuations did recover modestly. Behind this trend the telecom optical components industry has improved palpably. At trade shows and executive forums, the topic at the top of the list is industry consolidation, which many agree is a long-term solution. Several deals were made to this effect, but not at a pace or degree to produce lasting effect.
Then, in late 2008, strains in the financial industry brought a new round of difficulties that resulted in a freeze in spending patterns. Only now are we seeing infrastructure spending sustainably on the rise. In fact, we expect to see a continued recovery with CAGR back at 2005-2008 rates before long.
Going forward we think the fundamental problem that plagued the optical industry will have been addressed. In our view there are three key developments underway that will move the industry forward at better rates, starting now and lasting into 2010.
1) Service providers and system vendors recognize that the health of component suppliers is necessary to ensure stable and reliable supply with continued innovation coming from the bottom of the supply chain.
2) Component companies are right-sizing their businesses to run at >30% gross margin either by fully utilizing Fabs and/or outsourcing and streamlining operations.
3) Much like in 2000, M&A and consolidation will come out of necessity more so than luxury. But a second shakeout may be in the cards as operating winners and losers become more apparent. This will lead to a much healthier future.
Tags: telecom investments
Posted in Rajan Varadarajan, Telco | No Comments »
Back Up Your Truck to Backhaul Space
Rajan VaradarajanThursday, July 16th, 2009
As the mobile communications industry continues to grow at a significant pace, the greatest capex expense facing mobile service providers today is the overhaul of the backhaul (BH) network. This is the part of the mobile network that encompasses the cellular tower and the base station; the radio controller, the equipment that aggregates data at the base station and transports it to the metro access point, and the elements in the metro network that aggregate data from the multiple base stations. This infrastructure is currently comprised of leased lines or point-to-point microwave links. As much as a quarter of a mobile service provider’s opex is associated with this infrastructure. This has a direct impact on the mobile operator’s operating margins. If this infrastructure cannot be maintained and upgraded in a capital efficient manner, the operating margins of the mobile carrier will deteriorate rapidly.
The total number of mobile subscribers worldwide is expected to grow to 5.2 billion by 2011. Wireless operators have to also respond to the increased usage and availability of mobile data devices such as laptops, 3G handsets, smartphones and PDAs. Two trends are driving up bandwidth (BW) requirements in the BH network. The first is the introduction of mobile devices with interfaces that provide the user with an enjoyable experience for data applications. The Apple iPhone was the first device to break this barrier, which led to rapid growth in data application usage, and BW demand, followed by such devices as the LG Dare and Samsung Instinct. This has exponentially increased the demand for mobile data, multimedia and video services. The second trend is the introduction of mobile technology. The demand for mobile data services will only accelerate with the deployment of 4G technologies like WiMAX (Worldwide interoperability for Microwave Access), HSPA (High Speed Packet Access) and LTE (Long Term Evolution). The key challenge faced by most mobile operators is how to support the increasing BW requirements per subscriber.
While the average tower used to produce less than 10mb/s of traffic to be backhauled, 4G wireless technologies will produce hundreds of mb/s per tower. This rapid increase in BH bandwidth requirements will quickly overwhelm existing TDM and ATM backhaul networks.
Yet exciting as that growth may seem, the huge ramp-up in data traffic means more cost for wireless network operators, who aren’t necessarily getting significant additional revenue from such services. The introduction of bundled - all-you-can-use voice and data - packages is a clear sign that the wireless operators’ ability to charge for the extra BW is limited. The expected growth in data and evolution to such plans mean wireless network operators need to find a way to handle traffic in a more efficient and cost-effective manner.
Therefore, the mobile BH access, aggregation and transport infrastructure is undergoing a massive overhaul. This presents significant opportunities to companies that offer solutions in the space. There are a number of them jostling for various pieces of the pie. Most notable of the players are: Cisco, Juniper, Huawei, Alcatel-Lucent, Ericssson, Ciena, and ECI Telecom.
It is widely accepted that long term, all of the backhaul will be accomplished using Ethernet links. However, mobile operators will be unwilling to completely write off their current investments and upgrade completely to Ethernet based backhaul. In the short and midterm, a hybrid backhaul approach that deploys Ethernet in parallel with TDM/ATM links will dominate.
Tags: backhaul network, mobile communications industry
Posted in Rajan Varadarajan | No Comments »
Cisco’s Blade Server Strategy
Rajan VaradarajanWednesday, July 1st, 2009
Cisco already has had dominant market share in its core businesses. In looking for new sources of revenue and continued growth, the company saw the datacenter virtualization trend as a great opportunity. Perhaps it was a challenge as well, to enter a server market dominated by its traditional partners such as HP, Dell and IBM and, hence, antagonize them. This is inevitable, because wherever Cisco goes for growth, it is going to step on somebody’s toes.
Datacenter design is changing rapidly to meet the demands of massive amounts of data being driven largely by video applications. With the benefits of virtualization at hand, designers are beginning to take a top-down view of the datacenter. Cisco has teamed up with the leader in the space, VMWare, and storage behemoth EMC to just do that. In addition, Cisco also has partnered with NetApp in its grand vision of the USC- Unified Computing System.
The basic question may be, why does Cisco really care about server virtualization?
One possible answer: In a datacenter, each virtual machine mimics a physical server. Obviously, Cisco cares about any technology that increases bandwidth usage and thereby helps sell its mainstay routers and switches. Some examples are video like TelePresence, WebEx, and now Flip Video. Cisco wants to accelerate the adoption of 10Gb/s port usage so it can start building the 40Gb/s and 100Gb/s gears of the future. Datacenters are the first places that can happen. However HP, DELL and IBM are neither in a hurry nor motivated to provide their server blades with 10Gb/s ports yet. They appear to be content with the existing 1Gb/s ports. So, Cisco saw an opportunity to push its 10Gb/s adoption in server virtualization. Instead of waiting for other companies, Cisco is now building blade servers for the datacenter that are customized for server virtualization. These blades provide up to 1Gb/s pipe to each virtual machine, which would mean more 10Gb/s ports at the backend of the datacenter, helping in Cisco’s 10Gb/s push.
An alternate possible answer as to why Cisco cares about server virtualization: HP has recently stepped up its efforts in building networking gear, stepping on Cisco’s toes. HP’s Procurve product line seems to be gaining traction among its enterprise customers. In addition, with the acquisition of one of Cisco’s largest customers, EDS, HP is beginning to challenge Cisco’s dominance in the market. By some accounts, HP is the second-largest networking gear vendor.
Perhaps the reason is a combination of both the answers or there may be a third one. However, Cisco definitely has the wherewithal - money, technical prowess, innovative spirit, and great leadership in John Chambers - to deliver on the promise of its vision - UCS.
Conclusion: Admittedly, this is a very big vision. In hard times, Cisco re-invents itself as it has done many times in the past. I believe it has the fundamentals, savvy management and competitive spirit to pull off this one as well.
Tags: data center virtualization
Posted in Data Center, Networking, Rajan Varadarajan, Semiconductor, Software, Solar, Technology | No Comments »