MITRE | The Edge | Communications Layering

The military envisions a future "network-centric warfare" where unrestricted communication between any allied combatants is possible. It is widely recognized that the military's current wireless communication systems are inadequate for this task. The underlying problem is both pervasive and subtle; it is not one of technology, but one of design. Military systems were not–and, largely, are still not–designed in a way that allows extensibility and reuse.

Extensible and reusable communications systems do exist, and a study of their architectures is revealing. One successful architecture is demonstrated in the revolution in phone-line modems. Each generation provided a very precise and very limited function, interfacing an RS-232 digital data port to an analog phone line. These devices "encapsulate" the modem function and present standard interfaces (such as the HAYES modem protocol) to their hosts. Encapsulation and interface standardization allowed developers to innovate, differentiate, upgrade, and interoperate with other products. It is the functional encapsulation and standardized interfaces, not the underlying technology, that afforded this flexibility. In fact, throughout their evolution, phone-line modems have freely changed technology, the most dramatic shift being from a hardware-intensive implementation to the primarily software implementations of today.

The Internet protocols offer a second, larger example. The Internet protocols loosely partition communication into distinct functional layers. This functional encapsulation provides the extensibility and flexibility that have led to new protocols, applications, and hardware. Again, the acceptance of several inter-layer standard interfaces have enabled innovation and competition. As with the modem example above, the Internet protocols show a technology neutrality which, for example, allows hardware and software implementations of routers. While the Internet's layering may be imperfect, the power of the layered approach is obvious; the Internet is a very large, rapidly evolving, heterogeneous, ubiquitous, interoperating communication system.

We believe that a functionally encapsulated, layered architecture is a critical requirement for future military communications systems.

Are "Software Radios" the Answer?
With computational power continuing to increase exponentially, truly flexible "software radios" are becoming reality. However, the malleability of software can not overcome the shackles of an inflexible architecture; spaghetti code is no more maintainable or extensible than custom hardware, and faster processors won't improve this. Second, in computing technology, "flexible" and "efficient" are antonyms, and some applications such as hand-held radios will place the greater emphasis on efficiency. As with our modem example above, a technology-neutral architecture has an advantage as it allows either custom hardware (efficient) or software (flexible) solutions.

Software radios, while an important part of future military communications, are not the whole answer. We believe that technology neutrality is a far more important architectural goal.

Architectural Considerations
We have discussed two criteria of an extensible radio architecture: functional encapsulation and technology neutrality. There are several others as well. A good architecture must also provide an intuitive functional decomposition, reduce interface complexity and number, allow reusable components, allow system extension through the insertion of unplanned components, and not rely on resource sharing. We believe that a radio architecture based on a layered, functional decomposition meets these criteria. This layered architecture, which has proven its mettle in networking, is directly applicable to radio design as well. The layered radio architecture and its relationship to the seven-layer model is shown in Figure 1. While the granularity of the decomposition can be debated, the concept should be clear. Figure 1 also suggests a key point in our decomposition: we believe that the radio is a media adapter much like an Ethernet card or phone-line modem. As such, the radio should not include functions of the network layer or above: those functions should be relegated to the host and a common, media-independent interface provided by the radio. This is precisely the approach now being taken in the commercial sector: WLAN adapters appear to their hosts as wired Ethernet, and some low-speed wireless devices have standard modem-like interfaces.

Conceptually, in a functionally layered architecture, individual functions such as media access, channel coding, and modulation can be selected independently and integrated into a working system. Such flexibility requires clearly defined interfaces in addition to a concise functional decomposition. The functional decomposition and design of these interfaces is not trivial, and several commercial standards are emerging.

Standardizing inter-layer interfaces is more problematic than clean functional decomposition. There are two reasons for this. First, the design of the interface affects the complexity and flexibility of components. Second, interface compatibility requires standardization on several levels: one interface may have mechanical, electrical, framing, protocol, and semantic requirements. This is shown in Figure 2.

Fortunately, the least abstract levels of interface are both the most likely to change and the easiest to modify. For example, conversion between mechanical interfaces (9-pin to a 25-pin connection) is easy and inexpensive. The adaptation becomes more difficult and costly at the more abstract levels of interface which, perhaps oddly, are the ones frequently implemented in software. Thus, the higher levels of interface (command format and semantics) are the most critical; these are the emphasis of our work at MITRE in communications layering.

Conclusion
Current military communications systems are inadequate for use in future network-centric warfare. The root cause of this problem is becoming obvious as extensibility and reuse comes about through forethought and planning absent from legacy designs. Recently proposed military architectures, notably Joint Tactical Radio System (JTRS), attempt to address this but overlook the importance for a technology-neutral architecture and easily separable, individually reusable components. We are investigating an alternative architecture based on functional encapsulation and layering.

For more information, please contact Michael Butler using the employee directory.

	Solutions That Make a Difference.® Copyright © 1997-2013, The MITRE Corporation. All rights reserved. MITRE is a registered trademark of The MITRE Corporation. Material on this site may be copied and distributed with permission only.
	Privacy Policy \| Contact Us