Network Standards, AMI and the Smart Grid

by Harry Forbes, ARC Advisory Group

Facing imminent advanced metering infrastructure (AMI) and smart grid investment decisions, electric utilities realize the value of building out their AMI and smart grid systems based on networking standards.

Utilities do not fully appreciate the wide variation of maturity among relevant standards. Some are mature and global while others are new, unproven and regional. Many fall in between. Many suppliers gloss over a thorough discussion of the standards, their development processes and the maturity of the associated products.

Standards are not binary, yes-or-no propositions. They come with a history and level of maturity that largely determine their utility and business value. With standards, maturity is not the same as age. As a standard matures, more than just the content of the standard evolves. The associated standards-development process and testing processes also grow and mature. Standards exist at different levels of maturity. The existence of a published, ratified standard represents only the midpoint of this maturity process, not the endpoint.

AMI Standardization

Several families of standards are most relevant to the North American AMI and smart grid landscape: American National Standards Institute (ANSI) C12, ZigBee, Institute of Electrical and Electronics Engineers (IEEE) 802 and the Internet Engineering Task Force (IETF). Each organization and work product has distinct strengths and weaknesses.

ANSI C12. While the C12 standards target electric meters, the structure of ANSI C12 limits the flexibility of the organization to respond to commercial needs. The organization also has not affiliated itself with a commercial organization that can serve these commercial market demands.

The C12.22 specification had the scope to “define and interface between ANSI C12.19 devices and network protocols,” but there is no way to verify compliance to the C12.22 standard or the C12.19 data tables it references. The National Electrical Manufacturers Association (NEMA) has declined to enter into a standards compliance role. As a result, vendors may claim compliance with these standards, but such claims cannot be independently tested or verified.

The state of the AMI market provides evidence of this weakness. The collector and head-end systems of major AMI suppliers do not support compliant meters from other meter suppliers. Likewise, the common method for integrating a second source for AMI meters is an agreement between two meter suppliers to integrate a communication module within meters of the second supplier. This need might arise because of proprietary technology at L1-L2, but the monthslong schedule to perform this integration indicates that it is not greatly simplified by the existence of C12 standards. In the IT world, it has been decades since physical layer interfaces were shared this way between vendors.

Finally, the key specification, C12.22, does not define any transport, network or lower-level layers. This is appropriate, given its scope. Utilities should be skeptical, however, of strategies that position C12.22 as the primary AMI standard. At best, C12.22 represents an application-level standardization. Even if conformance to C12.22 could be assured, this would not necessarily address Open System Interconnection (OSI) layers 1-4. Utilities will make major investments in AMI equipment that provide these OSI layers and should choose network technology in these layers understanding that they are doing so.

ZigBee. The ZigBee Alliance is a commercial consortium that deliberately modeled its work processes on the practices of IEEE 802 and the Wi-Fi Alliance. The ZigBee Alliance has tried to borrow the best practices from the network standards world—IEEE 802—and networking commercialization—Wi-Fi. While the alliance has produced only a handful of standards in its seven-year existence, its work processes had already been developed and refined before it began.

These best practices adopted by ZigBee include accepting the role of compliance certification and contracting with independent, external testing laboratories to develop and perform compliance testing. ZigBee device-level compliance involves semiconductors, networking software and applications. Each can come from several suppliers, making testing a challenging necessity to ensure interoperability. Another part of its best practices is to use the testing process as a feedback mechanism for improving ZigBee standards.

ZigBee technology is considerably broader than the Wi-Fi Alliance. It is responsible for the network layer (L3) and rules of network formation and management that are managed by IETF standards in Wi-Fi networks. This work was necessary when ZigBee was the only nonproprietary network stack that ran over IEEE 802.15.4. In 2007, the feasibility of running IPv6 over these networks was demonstrated.

This creates new options for suppliers. The usefulness of IP in place of ZigBee networking remains contentious among suppliers. The availability of IP as an alternative somewhat devalues the ZigBee network layer and network management, but it has little impact on the interoperability model defined by the ZigBee application layer, ZigBee Clusters and ZigBee Application Profiles.

IEEE 802. IEEE 802, the preeminent networking standards organization, is responsible for standards that represent most of the dominant networking technologies in the world. By charter, the IEEE 802 standards are restricted to the lowest two layers (Physical and Data Link) of the seven-layer OSI networking reference model. IEEE 802 has highly mature standards development processes and procedures, but these do not guarantee each 802 initiative will converge to a standard, nor do they ensure a high probability of commercial success.

The 802 organization has informally spawned a group of entirely separate commercial consortiums that manage the commercialization of its most successful standards: Wi-Fi Alliance, ZigBee Alliance, WiMAX Forum, etc. Although administratively separate, these organizations are funded largely by businesses with commercial interests in the success of the standards. The compliance testing and certification services they perform provide knowledgeable feedback to the standards developers to help resolve ambiguities in future versions of the standards. This form of cooperation between a standards body and a commercial consortium has become an established best practice.

From an electric utility standpoint, the activities of the commercial consortiums are more important than are the underlying IEEE 802 activities. These organizations act as brand managers in technological and promotional senses. This form of organization has proven so effective that it has been imitated by others. Notable for AMI was the recent formation of a consortium named IPSO (Internet Protocol for Smart Objects) that seeks to perform this commercial role for IETF standards, especially those pertaining to IEEE 802.15.4.

IETF. In the standards landscape, the position of the IETF is analogous to the IEEE 802, except that IETF operates at higher layers in the OSI model (layers 3-7 for the IETF vs. 1 and 2 for the IEEE). Both organizations have seen their standards become ubiquitous.

Besides its obvious success, the strengths of the IETF standards methodology are its emphasis on developing working code that implements the standard during the standards development process. This practice provides early indication of areas for improvement and improves the quality of its final work products. This is also well-suited to academic participants who engage more with this organization than with IEEE 802. The large number of proposed standards and drafts managed by IETF means that the commercial handoff is inconsistent. While many commercial consortiums exist to build on IETF work, the commercialization model is not as uniform as it is with IEEE 802.

Closest to an IETF commercialization consortium for AMI is the Internet Protocol for Smart Objects (IPSO) Alliance. Its founders recognized the ZigBee Alliance’s effectiveness in managing and promoting ZigBee technology and the lack of an organization performing this role for IETF technologies in sensor and device networking. Some of these firms were investing research and development primarily in IP sensor networking, so they felt a need. IPSO founders include Arch Rock Corp., Dust Networks, Eka Systems and Nivis. IPSO also includes Silver Spring Networks, a builder of IP infrastructure for AMI, and networking giants Cisco Systems Inc. and Ericsson.

Network Strategy Recommendations

There are five overall strategies for planning AMI and smart grid networks:

  • Network planning using the OSI seven-layer model is critical, but in commercial practice, the seven layers are bundled in pairs and effectively are not independent. Thus the OSI model can be simplified to three components: L1-L2 choices (Physical and Data Link layers), L3-L4 choices (Network and Transport) and application-level interfaces. This simplification can help enlighten network architecture decisions and highlight alternatives.
  • Utilities should rigorously identify the technologies and standards they plan to use at each of these three layers for each part of their AMI networks: home-area network (HAN), collector, backhaul, head-end and enterprise. Proprietary technology should be identified as such. Without independent interoperability testing, technology that relies only on supplier claims of conformance—such as ANSI C12—should be classified as proprietary. In addition, the level of investment in each technology should be estimated to highlight the relative investments in particular technologies.
  • Evaluate your AMI architecture’s reliance on proprietary networking technologies. Such use may be appropriate but must be recognized as essentially a vote of business confidence in the relationship with the supplier of the technology. In these considerations, the size and financial soundness of the supplier become increasingly important as the investment becomes larger. This technology must be upgraded and replaced by this supplier acting as a sole source. This situation carries technological and business risk. The technology risk is unavoidable. The business risk can potentially be reduced by innovative negotiation of long-term agreements with the supplier.
  • Use networks based on IEEE 802 at OSI layers 1 and 2 when possible. These technologies will be available from multiple suppliers and will have the best chance of providing a graceful migration path to new and better solutions that will become available during the life of the metering infrastructure.
  • Use Internet Protocols at OSI layers 3 and 4 whenever possible, including HAN applications. Internet technologies offer huge advantages for long-term network deployments, chiefly the vast number of related and field-proven protocols for network provisioning, formation, management and security. Numerous commercial products support these protocols.


Harry Forbes leads the network research team at ARC Advisory Group. His focus areas include industrial networks, AMI and the smart grid, the electric power industry and industrial utilities. His education includes an MBA from the University of Michigan and a BSEE from Tufts University. Reach him at

Previous articleEnergizing Utilities for the Economic Stimulus Plan
Next articleARRA Dominates GridWeek

No posts to display