The use of pre-fabricated building blocks can be an extremely cost-effective approach to system design, especially when using off-the-shelf products that are based on open industry standards. Because standards-based blocks have broader utility than custom designs, vendors can leverage the efficiencies of volume manufacturing to reduce unit cost. System designers save development time and effort, further reducing costs as well as speeding time to market.
System development companies also gain the ability to make a more focused investment. By not having to create these standard blocks themselves, design teams free resources to concentrate on their company's unique value add to the system design. This is especially beneficial when the company does not have the resources required to create everything from the ground up anyway. Adopting standards-based design provides such resource-limited companies the opportunity to enter a market that would otherwise be closed to them.
Adopting the standards-based design approach can further reduce costs over the design's lifecycle by leveraging multi-vendor competition in creating the blocks. Competition naturally helps keep down component costs for both initial purchase and maintenance replacements. Multi-vendor competition also fosters innovation that provides designers with the opportunity to make system upgrades at low cost by simply replacing blocks.
Experience has shown, however, that the promise of standards-based design is easily undermined. The COTS offerings may include capabilities not needed in the end system, resulting in wasted resources. An even greater problem, however, is that ambiguities, allowed options, and gaps within the standards that can lead to variations among vendors' product offerings. These variations, in turn, prevent building blocks from different vendors from working with one another, despite the fact they all conform to the standard. This situation not only undermines the benefits of standards-based design, it can turn the approach into a liability. Interoperability among multiple vendor products, then, is an essential requirement of standards-based design.
Communications Standards Ecosystem
Despite these risks, the benefits of standards-based design are so compelling that the communications industry has gone to considerable effort to foster the approach and develop the appropriate standards. The standards now available address all elements of a communications system design (See Figure 1). One set of standards addresses the design of the communications platform hardware. Another defines the operating system software that runs on the platform. Further sets of standards address the hardware/software interface, applications programming interface, and middleware for high-availability operation and essential system management services.
These standards do not all reside under one roof, however. Three different groups are actively involved in the definition of standards, with additional groups providing guidance and coordination. The hardware platform standards, for example, fall under the auspices of the PCI Industrial Computer Manufacturers Group (PICMG). The PICMG has defined the Advanced Telecommunications Computing Architecture (ATCA) and the derivative MicroTCA platforms, developing standards for card cages, power systems, and plug-in modules including aspects such as system management, cooling, and inter-module communications. The organization has also defined a mezzanine card architecture (AdvancedMC) that both modularizes the design of ATCA boards and serves as the base board for MicroTCA platforms.
The operating system building block is the domain of the Linux Foundation. This organization has two roles. One is the standardization of the Linux operating system for general computing use. The other is to define a carrier-grade Linux that meets the real-time, fault-tolerant, high-availability needs of communications system operation.
Because many of a communications system's functional requirements are well established and common among all systems, an opportunity for standards-based system software has arisen. One of the key software requirements is that system operation be highly reliable as well as fault tolerant so that systems maintain operation 99.999% of the time. The Service Availability Forum (SA Forum) has defined the functional requirements of a software base, called Service Availability Middleware that will provide this capability and created the OpenSAF reference implementation. The SAForum has also defined standards for the middleware's interface to other system elements--the Hardware Platform Interface, the Applications Interface Specification, and the Systems Management Interfaces.
Each of these standards-defining organizations arose independently from the vendor community serving communications system developers. In order to ensure that these independent efforts closely reflect developer needs the SCOPE Alliance, a consortium of network equipment providers, arose. The SCOPE Alliance has created profiles that define the functional requirements on platforms and systems built from standard hardware and software blocks. These definitions give the standards bodies a target for their efforts.
The Role of the CP-TA
Ensuring that these various standards-based hardware and software building blocks will all work together once assembled falls outside the scope of any one standards organization, however. To fill that gap, the Communications Platforms Trade Association (CP-TA) arose. The CP-TA seeks to accelerate the adoption of standards-based COTS for communications system design, while preserving the developer's freedom of choice among vendors, by tightening and clarifying standards to address interoperability issues.
Where the SCOPE Alliance provides the standards organizations with concrete targets, the CP-TA works to ensure that the results all mesh together without requiring substantial integration effort. The CP-TA's unique value is that it bridges between organizations to keep the standards compatible, defines compliance levels that resolve potential interoperability issues, and provides tools and test procedures that validate those compliance levels.
In addressing system interoperability issues, the CP-TA has focused on three main domains: thermal, data transport, and system manageability. The domains represent what experience has shown to be the most problematic issues in building systems from standards-based building blocks. Thermal issues, for instance, include the presence of hot spots on one board that can affect adjacent boards and how board designs affect the flow of cooling air in a chassis, topics not addressed in the standards that can create problems during system integration. Data transport and system manageability issues arise as the result of vendors choosing different allowed options or having differing interpretations of the standards, also creating system integration challenges.
The CP-TA has created a tiered program that helps developers mitigate such interoperability issues. The first tier is identification of which standards requirements have significant impact on system integration efforts. At this tier, the CP-TA has released its Interoperability Compliance Document--ICD 1.1. This document defines objective and verifiable interoperability criteria and serves as the foundation of the CP-TA compliance testing.
Test procedures that offer a uniform measure of compliance with these standards form the second tier. These procedures specify what attributes to test, how to perform the test, and the degree of interoperability that the results demonstrate. The CP-TA has released its Test Procedure Manual--TPM 1.1. The procedures in the manual have a one-to-one mapping to ICD 1.1, fully defining the tests that will measure the interoperability criteria.
The third tier is the development of industry-harmonized test tools that can generate directly-comparable compliance results for evaluating vendor offerings. The CP-TA has worked with member companies to create specialized test tools for rapid performance of key tests. These tools include an ACTA Manageability Tester, a Chassis Scan Tester for evaluating cooling air flow, and a Blade Profiler (See Figure 2) that evaluates the thermal attributes of an ATCA board.
A recently-concluded proof-of-concept study demonstrates the value of the CP-TA program for mitigating interoperability issues. The study explored the integration effort required when creating a system from ATCA modules. One set of modules was evaluated for compliance to the PICMG integrated platform management interface (IPMI) specification using the ATCA Manageability Tester before system integration. The second set of modules was integrated without any testing. Integration without testing required 400 man-hours of effort while integration based on test results took only 104 hours, a 75 percent savings in integration effort.
Goals for 2009
Working from this foundation, CP-TA established several goals for further addressing interoperability in standards-based communications building blocks this year. One is to revise both the ICD and the TPM to reflect changes in the hardware platform standards with the release of PICMG 3.0. These documents were recently released and are available on the CP-TA Web site. Other goals include defining the interoperability requirements on implementations of the SAForum's Hardware Platform Interface and the hot-swap features of PICMG's AdvancedMC modules.
Ultimately, the design of communications systems using COTS standards-based building blocks will become increasingly common. According to market analysts VDC, 50% of Tier 1 and 86% of Tier 2 and 3 network equipment providers are already in the process of implementing an xTCA-based system design. Further, the current economy is expected to increase adoption as companies seek cost-cutting alternatives and opportunities to focus resources on value-added efforts.
The major barrier to fully achieving the promised benefits of COTS-based design, however, remains interoperability. Developers see interoperability as a criterion equal to product quality in making their make-versus-buy decisions, and interoperability issues are one of the main reasons other standards-based designs have declined in popularity. CP-TA is actively working to provide the guidance and tools that will help the industry resolve such issues and ensure that COTS-based communications systems provide more capability at lower cost in less time than proprietary design efforts can achieve.
About the Author
Sven Freudenfeld is President of the Communications Platforms Trade Association (CP-TA) and has more than 15 years of experience with voice, data, and wireless communications.