Each new generation of standards is at least an order of magnitude more complex than its predecessor. As a result, the cost of developing standards has increased, in most cases, beyond the level that an individual company would be comfortable shouldering. The growing size and complexity of standards also means that more companies are needed to provide the specialized skills required. Because international markets are now a target of most standards, involving companies from different countries helps with regulatory and market acceptance.

By its very nature, standards development requires a nimble approach; every time the standard consortium adopts a change, models must be updated and re-run. More partners in the consortium can mean more changes.

The traditional approach to standards development, using C models, is ill-suited for this kind of rapid and iterative development. Model-based design provides an alternative method of modeling communications standards with hierarchical, block-based models that drastically reduces the time required to make modifications.

Weaknesses of current method The current approach to standards development is based on defining the specifications in text and writing models in C to verify the specifications and evaluate performance. Typically, each contributor to the spec must model the complete communications system, including the mobile unit, base station, and channel path, to evaluate the performance of the part of the standard they're developing. Then, they'll evaluate various algorithms or subsystems and determine which works best. Normally, companies share the text defining the standards with other collaborators but not the models that they use to validate the standards.

C has proven to be a low-productivity environment for communications system simulation. It lacks built-in constructs for expressing concurrent algorithms and the connections between them. This weakness presents particular problems with frame-based multi-rate systems. C also forces users to expend energy on low-level issues, such as pointers and semicolons, rather than higher level issues. There are no supported standard libraries of signal processing workhorses such as filters, channel models, channel coding, source coding, etc. Finally, it's not easy to represent time-domain algorithms and frequency domain RF behavior in the same model (Fig. 1).

The collaborative nature of the standard communications process brings to light these weaknesses. In the typical case, a standards body invites submissions on technologies to incorporate into a new standard. Company X presents its results. Company Y confirms those results. The standards body votes to incorporate the new technology. The rest of the companies working on the standard then scramble to incorporate the new technology into their computer models. The standards development process is divided into different development groups, ensuring almost constant change. C-based standards development makes it difficult and time-consuming to implement these changes.

Movement to model-based design Today, system architects are using model-based design to rapidly develop systems and algorithms for new standards. Model-based design is the creation of executable models in a block diagram design environment using blocks to represent algorithms or subsystems. A key to the success of model-based design is the existence of libraries of pre-built standard communication algorithms (or blocks), such as convoluters, interleavers, modulators, spreaders, scramblers, and libraries of visual blocks enabling BER plots, constellation diagrams, and eye diagrams to be quickly and easily added to models. These libraries enable systems to be built quickly, literally by dragging and dropping the algorithms (or blocks) from a library into the model and then joining the blocks. This flexibility also enables models to be changed easily and quickly.

Users many not find every block they need pre-defined in a library, so model-based design lets users create new blocks. Users can create new blocks in C, or other languages, using a standardized interface and share these blocks with other users through the libraries.

Blocks are parameterized, which means it's possible to change their parameters using a block GUI. On a multi-path block for example, it might allow for the underlying multi-path model to be changed. An additional benefit is provided by block-diagram environments that provide a built-in scheduler or solver. This is particularly useful for standards with multiple rates (e.g., UMTS) or where the user wants to simulate analog components. Writing these solvers in C, or developing multi-rate C models, would take considerable development resources (Fig. 2).

Click here for Fig. 2

Model-based design advantages During the standard development process, model-based design offers several key advantages. The biggest is that models can be changed more easily and rapidly than C code. The blocks' behavior can be changed quickly by altering their parameters. Structural changes can be made promptly by adding or removing blocks from a model by dragging and dropping, without recompiling the model. The results of changes can be examined using the visual blocks, for example to show the BER after a coder has been added to a model. This allows different designs and algorithms to be evaluated quickly. C-coded models can't offer this level of flexibility.

Because users who create their own blocks in C use a standard interface, it's easy to share blocks using libraries. The intent of collaborators is clear and it's easy to incorporate new blocks into models. This is in stark contrast to wholly C-based systems where including code from collaborators can be difficult due to differing interfaces.

Model-based design can help bridge the gap between the time-domain tools used by systems architects and the frequency-domain tools used by RF engineers. It provides a platform that's suited for developing the systems architecture, yet can easily accept or generate the frequency-domain data used by RF engineers during the verification and testing process. So the same models used for specification can be used in the verification and validation process.

Design tools A number of block-based tools let users gain the benefits of model-based design, but the key requirement of such tools is that they allow the system being designed to be simulated or executed within the design environment. Simply put, the design must be executable. The tools must also have available libraries of components that can be used to quickly and easily build systems. Without such libraries, the benefits of this approach can't be fully realized. The final requirement is perhaps the most obvious, that the tools must allow for a visual system design. Tools such as Simulink are increasingly being adopted in the communications industry to allow for more rapid simulation development, a feature that's particularly attractive for groups working on standard research and development.

4G standard development The intense research efforts currently underway toward the definition of physical layers for 4G cellular systems is an example of how these tools are used for next-generation standard development. For example, Samsung's Advanced Technology Group is developing a multiple-carrier code division multiple access (MC-CDMA) system model that represents one of the contending 4G visions. Unlike the other main 4G alternative, orthogonal frequency division multiplexing (OFDM), MC-CDMA divides the available spectrum into sub-bands and share these sub-bands to more efficiently use common spectrum. Entities wishing to use part of a frequency band for transmission will first scan for currently unused portions of spectrum and then quickly establish temporary occupancy of these bands. Ideally, these sub-bands should be orthogonal, or independent of each other. One of the primary challenges of spectrum sharing is multiple access interference (MAI) which is caused by deorthogonalization, primarily as a result of multipath propagation effects such as reflections off buildings and vehicles.

Evaluating spreading and coding schemes Samsun's task was to evaluate various spreading and coding schemes to increase the robustness of MC-CDMA against the effects of MAI under various scenarios, including wide-area and short-range cellular, feeder links, and short range terminal-to-terminal.

The Samsung researchers have evaluated both single user detection (SUD) and multiple user detection (MUD) schemes to address MAI while at the same time considering the impact of various error coding schemes and different error coding rates. They've taken advantage of the modeling environment's ability to provide pre-built blocks for common error coding schemes and to change the coding rates by modifying parameters in the blocks. The modeling environment plots out the performance results for each scheme that they evaluate, calculating BER as a function of signal-to-noise ratio (Eb/N0) and as a function of the number of users on the channel.

The results of the simulation have already shown that the right combination of spreading and coding can increase the robustness of MC-CDMA against MAI, and in this way improve the affordable system load. The Samsung researchers have determined that MC-CDMA is a good candidate for the 4G physical layer interface, although its cellular performance needs further evaluation.

About the author
Mike Woodward is the communications industry marketing manager at The MathWorks. He has degrees in Physics and Semiconductor and Microwave Physics. He can be reached at mike.woodward@mathworks.com.