RTC Magazine

Consumers are demanding increasingly sophisticated telecommunications services from their service providers SP. The “handheld” is no longer simply a device for carrying on remote conversations. A growing number of people use their mobile phones as converged devices for a variety of real-time and asynchronous communication functions including creating, saving or sending pictures and videos; instant messaging; push-to-talk; and even online multiplayer gaming.

As users of and uses for the mobile phone continue to grow, network operators must respond to ensure their networks are capable of delivering the new converged services quickly and cost-effectively. This requires upgrading their networks from those consisting of disparate elements that provide voice, data, signaling and control functions to an all-IP network that is built around the emerging set of standards and architecture defined by the Internet Protocol (IP) Multimedia Subsystem (IMS).

Telecommunication service provider requirements are putting considerable pressure on telecom equipment manufacturers (TEMs) to provide new functionality, reduce cost and shorten time-to-market. Adding to the challenge, TEMs need to quickly evaluate emerging technologies to determine which will impact their business and how best to meet these new requirements. With increasing competition and tightening budgets, many TEMs are faced with unprecedented business challenges to meet customer expectations while maintaining profitability.

Challenges

There is a direct correlation between service provider (SP) spending on telecom gear and the revenue that TEMs can generate. Increased competition and tighter budgets create unique business challenges for TEMs wanting to both meet shifting customer expectations and maintain profitability. In recent years, network operators have put significant pressure on TEMs to meet a variety of challenges—provide equipment with new and improved functionality, reduce their acquisition and operational costs, and reduce the complexity and time required to bring new capabilities to market. The rapid pace of innovation in the marketplace coupled with the need for TEMs to rapidly evaluate the most effective emerging technologies, further complicate the situation.

In order to offer converged services such as multiple play services—voice, text, video and mobility bundled as one package—the SPs must choose the right business model. That depends on a variety of factors such as the economic environment, the subscriber density, customer care needs, etc. These factors determine whether such services can be offered profitably.

Furthermore, to protect profitability and return on investment in challenging market conditions, SPs must continue to look for ways to reduce their capital expenditures and operating costs. Through recent consolidations, SPs are enjoying increased buying leverage over their suppliers, presenting TEMs with mounting price pressures. Today’s network elements are selling for less than half the amount of those with similar or less functionality only a few years ago. This trend is likely to continue for the foreseeable future. SPs continue to demand increasing functionality and performance at decreasing cost.

Another factor contributing to the price pressures is that some of the fastest growth in wireless adoption is occurring in price-sensitive emerging markets, especially India and China. Wireless subscribers in both countries have long surpassed landline subscribers. Market penetration remains fairly low (especially in India), presenting equipment providers with excellent opportunities for growth. Currently, prepaid service is keeping tariffs low. The resulting average revenue per user (ARPU) is around U.S. 10 dollars in India and China. This compares to an ARPU of 57 dollars in the U.S., and 40 dollars in Europe. This clearly indicates that the telecom equipment developed for high ARPU regions (North America and Europe) is not suitable for these emerging markets. TEMs must adjust their cost structures to address such price pressures.

Network operators must upgrade their legacy networks to equipment that can effectively move from commodity services to new revenue-generating services. Mounting competition between wireline, wireless and cable operators is turning voice and best-effort data services into commodities. Voice services are continually migrating from circuit-switched wireline to VoIP and mobile networks. Consequently, even though declining voice services generate cash, they are not necessarily generating profits. This is driving SPs to seek out new sources of revenue and profit.

A multiplay offering of Internet, phone and television service—combined with mobility—provides a compelling set of services that promises a significant increase in ARPU for SPs that have the capability to deliver converged services. At the same time, many telecom and Internet SPs have little experience delivering the kind of multimedia content that cable and satellite providers have offered for years. Encouraging news for telecom SPs, however, is that technologies incorporated into the IMS framework will enable the delivery of multimedia services over their carrier infrastructure. To make the most of the multiplay opportunity, SPs and TEMs must meet the quick and cost-effective implementation challenge head on.

Offering converged services on an all-IP network poses unique challenges of its own. For decades Ma Bell set a pretty high standard for service reliability and availability—the dial tone is there every time we pick up the phone! The traditional circuit switched networks, primarily built to deliver voice services, were designed for high reliability of service with five to six nines (99.999% - 99.9999%) availability. Even many of our video content providers—the cable companies – operate over hybrid fiber-cable networks capable of delivering four nines (99.99%) availability. These traditional services have set a consumer expectation of service availability that must be met by the all-IP networks if multiplay services are to be adopted widely. This is a significant challenge if one considers that in general Internet provides less than three nines (99.8%) availability at best!

SPs generate revenue from applications and services running on systems provided by the TEMs. Historically, TEMs have built proprietary platforms in-house because of the lack of availability of standards-based commercial off-the-shelf (COTS) components that met their requirements. At a high level, these systems generally consist of four layers: hardware, operating system, middleware and applications (Figure 1). The hardware, operating system and middleware layers must be acquired and perfected before differentiated end-user applications and services are developed and deployed. Working with proprietary platforms often means encountering many of the following challenges:

• Long development and integration cycles involved with ensuring proprietary functionality and integration of third-party and legacy components

• Product commercialization cycles measured in years rather than months

• Missed deadlines resulting from underestimated development and integration effort

• Lost revenue or market segment share from being late to the market

• Significant resources hit whenever a change is made in any of the layers

Several market realities have caused systems developers to move away from proprietary development. Budget constraints, aggressive time-to-market requirements, increasing cost pressures and fewer resources are prompting TEMs to acquire hardware and middleware from third parties while focusing on their core competence of developing revenue-generating applications. At the same time, emerging standards are allowing designers the flexibility to build systems by combining sets of interoperable, off-the-shelf hardware, operating systems and middleware building blocks from several competing vendors. As the telecom world moves away from proprietary systems in favor of those built using commercially available standards-based components, the challenge for TEMs becomes deciding where to concentrate their efforts. They must choose where to rely on other suppliers to allow them to quickly address the service, cost and time-to-market requirements of their customers.

The Solution

The need to address the challenges of reduced time-to-market, reduced capital and operational expenditures, and increased opportunities for service offerings over IP is driving the demand for modular communications platforms comprised of standards-based hardware, operating system and middleware. By incorporating such industry-standard building blocks as AdvancedTCA (ATCA), carrier-grade Linux (CGL) and Service Availability Forum (SA Forum) interfaces, the modular platforms are intended to help drive new service offerings based on COTS components. With key standards such as ATCA, CGL and the SA Forum specifications gaining increasing acceptance in the market, the ecosystem continues to mature with more companies providing the sophisticated pre-tested, pre-integrated components that TEMs require to quickly build cost-effective, application-ready platforms. The transition from all-proprietary systems to standards-based systems is well underway and is expected to accelerate as rapid adoption of these standards continues. The migration to standards-based integration is illustrated in Figure 2. Let us look at the various layers of such application-ready platforms in a bit more detail.

Standards-based hardware can provide significant cost savings to TEMs. Among the most important factors driving standardization of the hardware layer is the ATCA standard. A natural evolution of the PICMG specifications, ATCA is the first open standard targeted primarily at developers of telecommunication systems. It provides specifications for creating carrier-grade hardware architecture to provide the reliability, performance and scalability demanded by telecommunication applications. ATCA is quickly gaining industry acceptance, and major TEMs have already announced plans to provide network elements based on this standard. Although revenue estimates vary widely (the 2007 estimates range from $4 billion to $20 billion), this standard is clearly gaining popularity with OEMs.

Another important development is the Hardware Platform Interface (HPI) specification from the SA Forum. HPI specifies a rich set of hardware platform services which, when implemented by the hardware OEMs, provide significant ease of integration with HPI-compliant commercial middleware from a variety of providers. This specification has quickly gained widespread acceptance in the COTS ecosystem, and an increasing number of hardware providers and ISVs are implementing the HPI services.

Carrier-grade Linux (CGL) continues to be adopted by the telecom market with several TEMs already delivering systems based on CGL. CGL is one of the four working groups of what was until recently called the Open Software Development Laboratory (OSDL)—an industry body dedicated to accelerating the adoption of the Linux kernel across multiple markets. Recently OSDL merged with The Free Standards Group to form The Linux Foundation. The CGL Working Group is defining feature roadmaps and specifications for use in telecommunications architectures. MontaVista, Red Hat and Wind River are among several commercial vendors providing different distributions of Linux software based on CGL specifications.

The next logical step up the standardization chain is middleware. If the middleware layer can provide abstraction between layers, the potential benefits to TEMs are huge, allowing them to focus on telecommunication services—their primary added value—without having to worry about the underpinnings. Such standardization is well underway. The SA Forum has delivered interface specifications that help middleware vendors write software conforming to established application program interfaces (APIs) at the hardware and application layers. The Application Interface Specification (AIS) establishes a common interface between the application layer and middleware components. These specifications aim to facilitate portability of middleware and applications across multiple platforms, reducing startup costs and integration efforts. Several middleware and systems vendors have announced support for AIS, creating and marketing middleware for high availability, systems management and database development.

The application layer is where SPs generate their revenue. The hardware, OS and middleware layers constitute enabling technologies that support applications and services provided at the application layer. This is also where the TEMs have the greatest opportunity to differentiate themselves and provide compelling converged services to meet or exceed SP requirements. In other words, when TEMs can minimize their cost and effort in the lower layers by using COTS components to build an application-ready platform, they have more resources to focus on their core business in the application layer.

An Implementation Example

GoAhead Software has been working with Intel to ensure its high-availability (HA) software platform takes full advantage of Intel’s high-performance silicon on ATCA boards. GoAhead SelfReliant software running on an Intel processor-based ATCA platform utilizes a fully redundant design controlled by active HA software to monitor subsystem performance and automatically fail over in detected fault conditions. The software also provides a single, integrated view of the whole system by using the SA Forum HPI to the Intel Chassis Management Modules (CMM). GoAhead and Intel have created a solution to specifically address the IMS market. This solution illustrates how a complete IMS-compliant system can be constructed on a single ATCA shelf. It describes a stateful failover using the HA software running on an ATCA platform that features Intel processors and wire-speed network processors. It includes a fully redundant design controlled by the HA software that monitors and automatically fails over from faulty subsystems.

The architecture diagram depicted in Figure 3 shows several single board computers (SBCs), also called blades, with instances of session-initiated protocol (SIP) being monitored by SelfReliant on AdvancedTCA. The SIP bulk call generator makes calls to the proxy server, which routes calls to active instances of the SIP applications. If one of the instances of the SIP application fails, the highly available software restarts the standby application that will pick up the call load at no loss. If one or both of the active SBCs fail, the standby SBC assumes all traffic. Once the failed SBC is back online, the standby SBC automatically resumes its standby role. The HA software provides instantaneous switchover at failure to ensure uninterrupted service for the end user. It allows users to actively monitor failover effects on processed calls, as well as manage and control instances of the SIP application using a Web console. The CMM communicates with the HA software through open HPI, open intelligent platform management interface (IPMI) libraries and APIs. HPI and IPMI allow it to monitor and manage the status of various components of the chassis.

Whereas IMS offers new and compelling opportunities for SPs and TEMs alike to bring converged multiplay services to market and create new sources of revenue, it has unique challenges associated with it. Not only must they seek and implement new business models, they must also find ways to upgrade their legacy networks quickly and cost-effectively. The emergence of key standards and a vibrant COTS ecosystem provides an attractive opportunity for SPs and TEMS to meet both sets of challenges.

GoAhead Software
Bellevue, WA.
(425) 453-1900.
[www.goahead.com].