Commercial-Defense Synergy in Wireless Communications
Wireless communications technology development is a complex process that includes interactions between the commercial and military sectors. An understanding of these interactions, including the opportunities for and barriers to synergy, is crucial to an evaluation of the potential for expanded military use of commercial products. Building on the historical and technical foundation provided earlier in this report, this chapter identifies broader organizational and R&D issues that need to be addressed to ensure that the DOD fields affordable, state-of-the-art untethered communications systems that meet future military needs.
Wireless technologies are often transferred among government, industry, and academia. Such interactions take place through multiple mechanisms, sometimes in a continuing cycle from the commercial to the defense sector and back again (see Box 3-1). The synergy can evolve either during the initial research or after technologies are developed. For example, the DOD's funding of basic academic research on wireless technologies and networking (currently through the DARPA GloMo program) creates an active technology base for use in both military and commercial industries. Similarly, there is overlap within companies that have both commercial and defense divisions. Most large corporations also support academic research to gain access to important new concepts.
This chapter examines how this synergistic process might be leveraged to meet future military needs in untethered communications. Section 3.1 provides a brief overview of military use of commercial wireless products. Section 3.2 identifies the motivations and opportunities for
Handie-Talkies Serve Both Military and Commercial Needs
In 1940 Motorola developed the first handheld two-way radio, the Handie-Talkie, a 2.3-kg AM unit with a range of 1.6 to 4.8 km. Within three weeks of U.S. entry into World War II, Handie-Talkie production exceeded 50 units a day; by 1945 more than 130,000 units had been built. In 1942 Motorola's design for the world's first portable FM two-way radio, the SCR-300 backpack unit, won a competition to replace an older Army Signal Corps radio, the "walkie-talkie." The SCR-300 weighed almost 16 kg, had an average range of 16 to 32 km, and could be tuned to various frequencies in the 40–48 MHz band. Motorola police radios were used in the Army's first radio relay system for behind-the-lines communications and its first radio teletype hookup. After the war, Motorola introduced the first commercially available portable radiophones, the Handie-Talkie radio line. A fully transistorized, VHF pocket transmitter version was developed in 1960. A fully transistorized, portable two-way radio was developed in 1962; its weight of approximately 1 kg was reduced by almost half in 1969. These devices have evolved into Motorola's current line of cellular telephones. Component technologies from commercial communications equipment are now designed into future generations of military equipment, thus furthering the ongoing cycle of commercial-defense synergy.
commercial-defense synergy in the development of wireless technology. Section 3.3 outlines the barriers to synergy posed by mismatches between commercial capabilities and military needs and operating requirements. Section 3.4 examines three broad issues that need to be addressed in the design of future wireless systems for defense applications. Section 3.5 reviews the relevant defense technology policy issues.
Myriad wireless technologies have originated within the government. Satellite programs initiated by the federal government in the early 1960s produced technologies that were quickly adopted for commercial use, starting with INTELSAT in 1965 in the United States and other countries in the 1970s. Another important government-initiated technology was packet switching, developed by DAPRA (then known as ARPA) in the late 1960s. This advance led to commercial and military packet-switched systems worldwide as well as to the Internet. The government also led the work on advanced coding techniques (for recovering data from deep-space probes), spread-spectrum techniques, signal and data encryption, and more recently on-board digital processing. All of these technologies have been adopted by commercial enterprises.1
Conversely, the U.S. military uses many commercial communications products. The military uses a variety of commercial systems, including satellites developed in the mid-1970s to transfer weather data to computer processing centers and disseminate the processed data; commercial satellites and land-based services to transport military-encrypted communications links; VSAT networks operating over commercial satellites to disseminate logistical and weather data; satellite video teleconferencing networks to provide training and distance learning to the National Guard and reserve units and for telemedicine applications; and access-management approaches such as TDMA.
The ongoing synergy between the commercial and defense sectors is readily apparent in satellite communications. The introduction of commercial satellite communications in 1965 was limited to very small satellite payloads and required very large Earth stations to receive the very weak signals (e.g., INTELSAT Earth stations required antennas 100 feet in diameter). A 1971 experiment clearly demonstrated the feasibility of providing satellite communicationsincluding digital voice, data, and fax servicesto ships at sea. However, the business aspects of such services were not strong enough to justify the required investment in satellite and ground control systems. Subsequent events led to a contract between COMSAT, an international industry consortium, and the Navy to provide a GEO satellite system with an added commercial L-band package for ship-to-shore use. This agreement eventually led to INMARSAT, now widely used not only by large ships (e.g., tankers, cruise ships) but also by pleasure craft and mobile users around the world, who can transmit and receive data and voice via low-cost, brief-case-sized terminals. Terminals are also used on transoceanic airline routes for navigation, control, and passenger telephone calls. An INMARSAT spin-off, ICO, is building a MEO mobile telecommunications system using 12 satellites.
Similarly, the introduction of commercial DBS sparked military interest in developing the GBS to satisfy broadband data requirements in all environments, including the battlefield, ships, and logistics. The architectures of Ka-band (superhigh frequency, or SHF), high-speed interactive systems planned for commercial operation by the year 2000 will have an impact on the ultimate GBS design in the near future. Ultrasmall-aperture terminals in these systems will be able to transmit several megabits per second and receive 100 Mbps from a 24-satellite constellation. The GBS has been designed to leverage the current DBS satellites through modifications such as moveable spot beams and different frequency bands.
The opportunities for defense use of commercial off-the-shelf (COTS) products depend in part on the particular characteristics of a military
operation. For the purpose of analyzing communications requirements, military activities can be divided into the following four categories:
3.2 Motivations For Commercial-Defense Synergy
Two key factors currently motivate the DOD to seek commercial products and services. First, the size of the business and consumer markets and the nature of many commercial practices help achieve economies of scale at many levels. Second, commercial approaches to R&D reduce cycle time such that advances in technical performance can be integrated into field operations in a timely manner. The DOD therefore has both economic and functional reasons to adopt commercial products and approaches when they meetor could be adapted to meetdefense communications requirements. The commercial equipment is likely to cost much less overall than would equivalent defense-unique systems. Furthermore, because commercial industry evolves very rapidly in response to a competitive marketplace, the DOD can leverage commercial developments to field equipment that offers advantages in size, weight, power, bandwidth, or performance much more rapidly than is possible using traditional defense procurement practices.
Economies of scale are extremely important in the development and deployment of commercial products. Firms seek a balance of cost and quantity when deciding whether and how to enter business or consumer markets. As examples, business products such as VSATs are built in
DirecTV Receivers: An Example of the Volume-Cost Relationship
A DirecTV receiver consists of an 18-inch antenna and a sophisticated mechanism for receiving a 40-Mbps, digitally mutiplexed data stream. With more than 1 million units sold in the first year, these receivers are among the fastest-growing new product lines in the consumer electronics industry. DirecTV receivers were introduced at a list price of $700; after 2.5 million units were sold, the price dropped by nearly 50 percent because of competitive market pressures and economies of scale.
Impressed by the capability of such a small receiver system, the Navy and other services determined that DirecTV technology could be adapted to meet the military's broadband transmission requirements. However, the quantity needed by the militaryhundreds of terminalsis significantly smaller than the commercial market. If a DirecTV-like receiver were developed as a stand-alone military product, then the cost per unit might be hundreds of times higher than the commercial price because the development, tooling, and manufacturing-setup expenses would be amortized over a smaller production base and optimized for smaller production volumes. Three years after DirecTV was announced, the services were still working to define a military version. Had the features necessary to support military needs been considered before the product design was finalized, the DOD could have taken advantage of the cost reductions enabled by the market growth.
volumes of thousands per month, and consumer products such as DirecTV (see Box 3-2) are manufactured in quantities of hundreds of thousands per month. The cost-quantity relationship forced the semiconductor industry, which originally evolved to support military and space applications, to switch to a commercial focus. As shown in Figure 3-1, the commercial market for semiconductors soared, whereas the defense share declined. In 1975 worldwide military purchases of semiconductors totaled $700 million, approximately 17 percent of the global market (INSTAT/SIA Information Services, 1997). At that time all major semiconductor manufacturers had military-quality product lines, particularly for high-reliability and extreme-temperature applications. By 1995 the military share of the market had dropped to less than 1 percent (INSTAT/SIA Information Services, 1997). Most major semiconductor manufacturers have announced the phasing out or termination of military product lines. Now military contractors must use either commercially available parts or obsolete, but military-quality, semiconductor parts.
The commercial sector far outpaces the defense sector in production rates and volumes, not only for final products but also for subsystems and components. The largest DOD acquisition of communications equipment is the SINCGARS radio: The DOD has purchased 75,000 units over 10
years of production. In contrast, commercial production of land-based mobile radios is in the range of 400,000 units per month, and cellular radios are produced in volumes exceeding 2.5 million units per month for the largest suppliers. To meet such market demands a typical cellular telephone factory might produce 5,000 telephones a day. Another factor distinguishing the two sectors is the open, competitive environment of commercial production. The market pressure for improved quality, pricing, and other features is felt by all commercial competitors, whereas the defense market has typically been limited to a few and sometimes just one contractor.
The next four subsections examine economies of scale manifested in several areas of commercial technology development: design, production, maintenance, and training. The fifth subsection reviews how cycle time can be reduced, thereby moving technical advances into the field quickly and also lowering costs over the life of a product.
3.2.1 Design Reuse
Commercial communications equipment typically is produced with a basic design that has a 2- to 5-year life span. The components used in that design are selected in the 1 or 2 years just prior to product introduction and typically represent the then-current state of the art in performance and cost effectiveness. Thus, during the product life the components
remain cost-effective for the suppliers and are manufactured using state-of-the-art, cost-effective manufacturing facilities. Manufacturers typically anticipate new features in the market by modifying the design to use new components after 2 years of production. They also use components and manufacturing processes that are within a generation of the then-current state of the art, thereby operating close to the optimum level of cost effectiveness. By contrast, military equipment is often outdated: SINCGARS was designed more than 10 years ago, for example.
Commercial firms achieve the initial economies of scale through design reuse. The use of previous hardware and software designs can often save 50 to 80 percent of development time because detailed design documentation can be readily reproduced and design weaknesses can be largely eliminated using experience as a guide. Although the design cycle is not a major contributor to the economic cost of a commercial product, it is typically a large part of defense deployment cost. The DOD typically does not reuse hardware designs, instead relying on independent "stovepipe" systems, which are optimized to solve a specific problem. Some efforts have been made to create software libraries for reuse. Increased reliance on common building blocks could significantly reduce design cycle time (see Section 3.2.5).
3.2.2 Production Learning Curve
The learning curve is a statistical tool used to predict production costs and plan and control production. The curve is based on the assumption that there is a relationship between the time required to build a unit and the number of units that have been built; specifically, the learning process reduces the time needed to produce a unit as the cumulative number of units produced rises. It follows, then, that the less time it takes to build a unit, the lower the cost of that unit. If the cost of producing a unit follows an 80 percent learning curve, then there will be a 20 percent reduction in cost per unit each time the total number of units produced doubles. In the example shown in Figure 3-2, the first unit took 100 hours to build and the second unit took 80 hours, or 80 percent of the time and cost involved in building the first unit. The 10th unit required 48 hours, and the 20th unit required 80 percent of that effort, or 38 hours. The major factors that affect the cost of production are the initial cost, or the starting point of the curve, and the rate of improvement or learning, or the slope of the curve (Anderlohr, 1969).
The implication of the learning curve is that large volumes of standardized items, produced continuously (i.e., without significant hiatuses), reduce the cost per unit. Heeding this message, commercial production is fairly continuous, fluctuating somewhat with the demands of the market
but generally changing processes gradually and with few interruptions. Although customized versions of products are increasingly in demand, the basic platform is usually consistent and the adaptations are minimal. By contrast, defense programs frequently begin with the building of only a few units, perhaps a few hundred, to determine feasibility or fulfill a limited need. Often these units are produced in numerous small batches with interruptions between the production cycles.
Cost management is practiced throughout the design and production of commercial products. For example, production volume typically needs to be known before detailed designs can be completed. The component costs, labor costs, and investments in labor-saving manufacturing devices are factored into the final design of a consumer electronic device. Large production volumes enable the manufacturing of designs that would not be viable at smaller volumes. A significant example is the fabrication of customized ICs with many functions that normally would be implemented in separate ICs. The large volume reduces the overall cost of components, parts, and assembly, even after the nonrecurring investments are taken into account. Another example is the design of electronic equipment for cost-effective manufacturing. These designs typically feature modules that snap together and minimal numbers of wiring bundles, fasteners, moving parts, and different part types. Costs are reduced further through incremental production changes. During the repetitive commercial design and production cycles, the boundaries between system, subsystem, unit, and component begin to blur as automation enables larger and larger subsystems to be treated as components. In this way, what was once a high-technology system (e.g., computer memory) becomes a commodity part.
Following the lead of the commercial sector, the DOD might achieve some economies of scale in production by revising its procurement practices
to make large-volume purchases of basic COTS communications equipment for entire departments at one time. Some isolated efforts have been made in this regard, but there are ample opportunities to expand this approach.
3.2.3 Maintenance and Logistics Support
Economies of scale can be achieved in the maintenance of equipment after it has been developed and fielded. Equipment occasionally fails in the field because of design defects, manufacturing defects, worn-out mechanisms, lightning or power surges, or simply heavy use. Sometimes fielded equipment is upgraded during maintenance procedures to add new features or functions.
In the commercial sector field-failure data are typically analyzed on highly automated equipment, which can trace failures to specific modules and components and automatically update design and component history databases, including any links to environmental factors. Design updates are inserted into the manufacturing process throughout the commercial life of a product, thereby improving its robustness. After approximately one year of production, experience with field failures often has produced the feedback necessary to eliminate most design defects, reduce manufacturing defects to a level consistent with the current state of the art, and generally achieve the best product possible within price constraints.
Consumers rarely, if ever, pay for the maintenance or repair of low-cost communications equipment. Rather, warranties and service contracts are often viewed as a necessity in maintaining complex products that are not easily repaired; products are often replaced if they need repairs after the warranty expires. Viewing warranties as insurance policies, or guaranteed streams of income, specialty maintenance companies have emerged to provide a variety of maintenance tasks, both on site and at the factory.
In the defense sector, communications equipment is often maintained by the acquiring agency rather than the manufacturer, typically at greater expense. Typically a module is replaced and the equipment is retested, a strategy that usually finds the primary defect but sometimes misses marginal problems elsewhere. Large numbers of spare components need to be kept available, either in replacement modules or in component form such that modules can be manufactured, throughout the useful service life of the system. The supply of spares is often threatened when, because of the small production volume, the supplier no longer finds the component profitable to produce. When this occurs the manufacturer usually notifies customers of an ''end-of-life buyout." The customers then try to
project future needs and purchase enough components to satisfy them (not always at competitive prices). The DOD is known for keeping communications equipment in service far beyond the life span of equivalent commercial technology; typically, military systems are removed from service only after catastrophic failure. Defense equipment is often kept in use for 20 years, whereas component suppliers often set product lifetimes at less than 8 years; thus, the military needs to stockpile approximately 10 years' worth of components. Under normal circumstances, the DOD assumes that 25 percent of its equipment will need to be refurbished at some point.
The additional maintenance costs associated with traditional defense acquisition could be reduced if manufacturerswho can efficiently analyze all field failures, suggest redesign enhancements, and redesign components and modules to enhance cost effectiveness and other featuresprovided for maintenance and logistics support when appropriate. In addition, a reevaluation of military equipment maintenance practices may be warranted in light of the capabilities of advanced communications and transportation systems. Defense acquisition systems have historically provided logistics support for rapid equipment repair (i.e., within a few minutes of field failure) anywhere in the world by staging replacement modules near locations where critical equipment is in use. Such an approach may no longer be necessary.
The commercial sector achieves additional economies of scale in training. The expense of user, logistics, and support training is built into the cost of new product introductions, and training is subsequently converted from expensive formats (i.e., personal, face-to-face support) to videotape, interactive CD-ROM manuals, on-line help, and literature. By contrast, the training of defense maintenance and logistics support personnel offers few economies of scale. Training materials are developed, but they are neither as detailed nor as widely distributed as are commercial manuals. Indeed, defense support training can remain somewhat diffused and superficial because the military uses so many types of equipment and relatively small numbers of each type. As the DOD purchases more COTS products, the use of commercial training materials might be appropriate.
3.2.5 Cycle Time
Commercial product design cycles, which usually last from one to four years, are set by competitive pressures: The first company to market a product with a new feature can reap large increases in market share and
profitability.2A new commercial communications product is released every few months. Manufacturers therefore begin to field test and optimize the features of products before the designs are completed, accelerating the development process by several years. This environment fosters the introduction of new and improved technologies at a very rapid pace, often at a low incremental cost to consumers. Companies gain additional reductions in cycle time by designing products to accommodate new features on each new production run, often every six months.3These advanced commercial technologies are then available for rapid insertion into commercial or defense applications.
The military product design cycle is much slower. It begins when a contract is awarded and ends with delivery of the final product, which is not field tested or optimized until the design is completed. The developer does not have sufficient control over systems integration, testing, and evaluation to perform concurrent engineering that would reduce overall cycle time. Further delays are imposed because training, logistics, marketing, and distribution processes are not generally developed concurrently with manufacturing tooling equipment as they are in commercial systems.
A key design feature affecting cycle time is the ease of upgrading equipment. Commercial baseline products are designed to accommodate hardware and software extensions throughout the planned lifetime of the product. This is critical because of the high cost of the wireless infrastructure. The longevity of the infrastructure depends on a complex trade-off between the equipment offered by vendors and the pace of change in services. Typically service providers have a detailed road map that identifies when new services will be offered; these services are selected based on the equipment available at a reasonable cost and the market demand for a profitable service.
Commercial upgrades to accommodate new services and changes in market direction are generally implemented through software updates rather than more-costly hardware changes. Therefore, software plays a growing role in product development and cycle-time planning. Software is also often used to correct hardware problems, such as designs that were oversimplified to meet a price point. In such cases new software requirements are discovered late in the product development cycle, meaning that software is the last element to be developed and may be installed either just before or even after production. Yet software updates need to be thoroughly tested and tolerant of all environmental and loading factors. As a result, the software development process is now of great interest. The quality and timely release of software as well as software-defined infrastructure services are therefore becoming critical factors in the commercial communications industry.
Military radios, by contrast, are typically optimized to meet a single specification rather than designed for easy upgrading. Moreover, when upgrades are possible they usually can be implemented only by the original vendor. An exception is SpeakEASY, which was designed to support hardware and software extensions through its open architecture. Emerging military requirements (e.g., interoperability, open architecture) are driving the trend toward more flexible systems such as software radios. With the role of software expanding in both sectors, the commercial experience in software management may offer lessons for the DOD.
Both sectors are concerned about retaining the use of legacy equipment. As the complexity and cost of each new generation of infrastructure rise, the effects of obsolescence become more important to system planners, investors, and consumers. Concepts such as backward compatibility, while ill defined, have a very pragmatic meaning in the consumer electronics business: A technological advance should not result in denial of service to owners of older systems. Backward compatibility is also important to the military, which is likely to continue using legacy equipment for some time to come.
3.3 Barriers To Commercial-Defense Synergy
Although many defense communication needs can be met with commercially available equipment, a variety of barriers prevent a complete match across all systems. For example, commercial-defense synergy is not appropriate in the development of some highly specialized or classified command-and-control systems. There are organizational limitations as well. Traditional government acquisition has led to a large number of stovepipe systems, which are developed by a single contractor, meaning that other contractors cannot compete for follow-on development or production.
There are additional concerns regarding industry's current capability to meet the unique requirements of some military systems. In the wake of cutbacks in some procurement programs, many defense communications suppliers have begun to develop a commercial orientation to preserve their technology and manufacturing base. Commercial applications are attractive not only because of the vast market but also because they are not subject to federal and defense acquisition regulations: The supplier benefits from simplified accounting and has greater freedom to structure the details of contractual agreements. In the past, an increase in contracts was sufficient to rebuild capacity following a period of reduced defense spending. Now, however, action might be necessary to ensure that critical design and manufacturing capabilities are not lost altogether.
In view of the need to maintain surge capacity for those times when
sudden military action is needed,4the declining industrial interest in military communications systems makes it all the more important to understandand either overcome or accommodatethe barriers to commercial-defense synergy. These barriers include the risks of military dependence on the commercial sector, the contrasting approaches to making trade-offs between cost and complexity, and differences in communications infrastructures.
3.3.1 Risks of Dependence on Commercial Technologies
The DOD assumes certain risks when using commercial technologies. Most importantly, significant military use of commercial technology increases the possibility that a potential adversary could have or gain detailed knowledge of the systems. In the case of computer networks, the defense infrastructure is built almost entirely of widely available computers, software, and networking components. An adversary could exploit the known weaknesses of these components, a possibility that has prompted efforts to improve network robustness. Any weaknesses in the commercial communications infrastructure might be vulnerable to similar exploitation.
One solution is to use communications equipment that is unavailable to potential adversaries. Such products include those subject to U.S. export controls, which are intended to keep certain advanced technology products and weapons systems away from designated hostile countries. Several types of advanced commercial communications technologies, including those involving extensive computations or spread-spectrum techniques, fall into this category. It is difficult to gauge the impact of international technical awareness of military effectiveness because technology is only one of several influences (the others include strategy, tactics, training, and weaponry). Critics of export controls contend that many affected technologies are readily available from offshore sources and that the controls serve only to reduce sales for U.S. manufacturers and provide evidence of who is acquiring these commodities.5But with the rapid evolution of commercial wireless technologies, especially software-based systems that could be converted to implement new waveforms or military countermeasure capabilities, there is also an argument for strengthening controls on the export of certain advanced communications technologies. Although the issue is beyond the purview of this committee, a review of export controls may be warranted.
Another risk of commercial dependence is the sometimes-hidden difficulty of making what seem initially to be simple modifications to COTS systems. End-to-end encryption or AJ technologies, for example, might be added to a commercial system to meet military security needs. However, given the high commercial production rates, such modifications are
often very difficult to implement. For example, the key interface signals or connections could be embedded in an IC and unavailable for specialized wired connection to an add-on feature. The safest strategy for the DOD is to use a common baseline technology and common components but pursue a separate design effort. In addition, the DOD could participate in standards-setting activities to encourage the development of baseline commercial equipment designed to accommodate the addition of militarily useful features.
A third concern is the availability of commercial systems in wartime. Many of the largest satellite networks, including those shared by the U.S. military, are owned or operated by international consortia. Although these providers are reliable partners in peacetime, whether they would give DOD priority or expanded bandwidth during times of conflict is unclear. Such preferential service could be hampered by the operators' need to serve other customers or possible unwillingness to provide support in a controversial conflict.
3.3.2 Trade-offs Between Cost and Complexity
Perhaps the most obvious barrier to direct commercial-defense synergy lies in the contrasting strategies used to make trade-offs between system cost and complexity. As a potential user of commercial services the DOD has certain expectations, many of which are requirements if it is to fulfill its mission. In a number of dimensions, these expectations are at odds with the criteria used by commercial communications services in designing and deploying their products. Simply stated, the military has some extraordinary needs, whereas the commercial sector tends to focus on delivering reliable but ordinary service.
126.96.36.199 Performance Issues
For example, the military expects to use leading-edge technology. The present analysis is based on the assumption that the DOD cannot achieve its mission with technology that is inferior to that of an adversary. It is also assumed that every adversary's technology is state of the art. But commercial communications services are rarely based on the most advanced technology available. Rather, providers deploy technology based primarily on its cost effectiveness and affordability, that is, whether customers are willing to pay for the capability.6Over time, production volumes increase and costs decline, but the initial costs of an advanced technology can be a barrier to its commercial application.
The DOD also requires that certain functional capabilities (assuming they are technologically feasible) be deployed in any location where the
military needs to operate. Military operations are extraordinary events in which communications traffic is unpredictable, driven by the characteristics of the individual operation. Yet the cost of an inability to communicate can be very high, and so the probability of such a breakdown needs to be kept very low. By contrast, commercial communications providers have limited resources and cannot cover every service area that could be profitable. Commercial systems are designed to serve a particular area for many years with slow changes in technologies, features, and volumes. To minimize life-cycle costs, such systems are based on fixed facilities that cannot be deployed or shifted rapidly to meet an immediate demand. Commercial wireless systems are generally engineered to meet the peak traffic requirements of the average business day. They are not designed to meet the requirements of extraordinary events under emergency conditions, even regular and predictable ones.7
Similarly, commercial and defense equipment differ in their tolerance for unusual environmental conditions. Military units are likely to encounter extreme environments such as jungles, deserts, or polar regions and be subjected to the harsh conditions of battle. Therefore, regardless of the added cost and system complexity, defense communications equipment needs to be designed and built to tolerate extreme temperatures, submersion, high levels of shock, explosions, and vibration. By contrast, commercial manufacturers and consumers are unlikely to incur additional expenses for equipment that operates under extreme conditions. In fact, consumers often favor the least expensive product over one with the best performance, warranty, survivability, and advanced features. Consumer products such as telephones are designed to survive reasonable levels of wear and tear and perform under moderate environmental conditions.
188.8.131.52 Quality and Testing
Military communications equipment can be highly complex and pose difficult testing and diagnostic challenges. For example, a military system can encompass networks of computers, each running a real-time suite of applications in support of a system-level application. When a safety or mission-critical function is involved, elaborate procedures are followed to provide for multiple redundancies, failure detection, independent software development, and cross-checking to ensure reliability. But exhaustive testing to identify problems that have an extremely low probability of occurrence (e.g., once every few hundred thousand instructions) may not be cost-effective if there are no safety implications or critical effects on performance. In cases that may not justify exhaustive testing, new work in formal methods (i.e., mathematical techniques that obviate the need to
test every possible situation) holds promise for further reducing the likelihood of rare anomalies.
There is a widespread perception that defense equipment is tested much more thoroughly than are commercial products. In truth, commercial testing procedures vary widely but can be quite rigorous. For example, manufacturers of industrial-grade communications products have largely adopted high levels of integrated quality control. In the semiconductor industry, firms expend significant resources even after a chip is designed to create an exhaustive test pattern capable of catching both design and manufacturing flaws.8Many vendors have adopted processes that are comparable in function to those required for defense equipment, although the conditions are likely to be less extreme.9Likewise, commercial electronics systems have improved in recent years because consumers now demand unprecedented quality and reliability, even under environmental conditions that just a few years ago would have been considered strictly military grade. Table 3-1 displays the stringent test parameters for one modern commercial product, a car radio. Drivers now view the performance of a radio as reflecting on the quality of the vehicle. As a result, most of the environmental qualification tests for a car radio are actually comparable to tests conducted on a high-performance Navy jet fighter.
As a result of these trends, some commercial products may be tested thoroughly enough to meet defense needs without further testing under military conditions. In some cases, it may not be cost-effective for the military to test and inspect commercial components already shown to have very low failure rates. This is most likely to be the case for equipment that has few moving parts and is designed for human use under extreme conditions.
3.3.3 Infrastructure Differences
The differing expectations of defense and commercial customers affect communications infrastructures. Cellular customers want to be able to make calls whenever they wish. Military users have much more complex needs for security, LPD/I capabilities, concealment of network functions, message priority and preemption, tolerance of severe environmental conditions, and, perhaps most crucially, the capability to work without any fixed infrastructure. These needs generally preclude the use of antennas on fixed towers, highly engineered site installations, site-specific antenna selection, microwave links to central switching offices, and preplanned routing tables for switches. (Exceptions include permanent posts and fortifications and a considerable amount of mobile infrastructure mounted on trucks, trains, and large aircraft.)
In operations other than war, the use of existing local communications infrastructure (where available) can offer tremendous cost and performance advantages. It remains crucial, however, to provide military-grade security when using cellular communications or other wireless data services; such security measures need to be interoperable with the U.S.-based military communications infrastructure, including the Secure Terminal Unit III (STU III; the government standard in secure telephony) and related systems, and the array of secure networking products (e.g., the network encryption system [NES] equipment for strategic defense networks).10The DOD is addressing these issues through the CONDOR program, which is developing a cryptography module for cellular telephones used by the military (commercial versions are already on the market, and additional spin-offs are likely). The system will be interoperable with STU III. The security of local infrastructures remains an issue, however, as does the continuing need for interoperability among defense networks.11
In regions lacking a local communications infrastructure, substitutes will soon be available. The Iridium, Teledesic, and GBS satellite systems will provide worldwide communications. The military could also use unmanned aerial vehicles (UAVs; special aircraft that can fly along a programmed set of waypoints and perform programmed tasks) as high-altitude platforms for communications and reconnaissance services. Among their advantages, UAVs weigh less than manned aircraft (leaving more room for payload) and have no need to carry life-support systems such as oxygen and emergency ejection equipment. The UAVs could serve as relays between wideband satellite links and battlefield communications systems, carry sensors, jam or otherwise attack enemy information systems, and even provide positioning information in the event that the GPS is shut down during hostilities to prevent its use by adversaries.
The high-altitude endurance system (a UAV-based airborne sensor system) and the airborne communications node (ACN)12UAV-based platform are designed to be cost-effective, multipurpose platforms delivering C4I capabilities.
3.4 Designing Wireless Systems
As noted previously in this report, some features required in defense communications equipment are not generally available in standard communications products and could be difficult to add. Some advanced features, such as LPD/I and AJ waveforms, interoperability with the DOD's nearly two dozen legacy waveforms, or highly specialized spreading waveforms, may be attainable only in specialized military systems such as the SpeakEASY software radio. The following sections discuss the design of future wireless systems for defense applications, focusing on three key issues: network architecture; security; and multimode, multiband systems. Some commercial products could be adapted to meet defense needs in these areas, but specialized military research will likely be needed.
3.4.1 Network Architecture
184.108.40.206 Network Design Issues
Most commercial communications infrastructures use a base-station-oriented architecture, in which communication flows between a base station (equipped with a well-sited antenna and a high-power transmitter) and wireless terminals. Cellular, paging, trunked radio, and various data radio services use this model. This design is not entirely appropriate for military communications infrastructures, which need to allow for immediate deployment with no set-up time, no siting advantages, minimal antenna advantages, and continuous movement of all network participants.
The DARPA GloMo program has focused on peer-to-peer, multihop packet radio networks. However, there are valid questions about the suitability of a peer-to-peer model for military command-and-control hierarchies as well as the RF link penalty paid by this design.13The GloMo perspective was expanded recently to include other architectures, but the program has never assessed all the possible choices, even though network architecture forms the basis of any telecommunications system. Among the issues to be addressed are peer-to-peer versus base-station-oriented design, connection-oriented versus connection-free architecture, single-channel (e.g., Ethernet) versus two-channel structures, bandwidth issues, and protocol selection.
For example, the choice of a connection-free architecture is particularly important for applications that are intermittent in content, as is typically the case in battlefield situations. (The inefficiencies of a connection-oriented architecture are discussed in Chapter 2, Sections 220.127.116.11 and 18.104.22.168.) The choice of a two-channel architecture can be particularly important if, as is often the case in military applications, much more data are delivered to a client than are received from the client. This design enables maps and situational-awareness data, for example, to be disseminated in a battlefield without any multiple-access penalty to an unlimited number of receivers. Meanwhile, location information, such as identification and coordinates, can be sent from many field areas in short, spread-spectrum bursts over a separate multiple-access channel.
Numerous research issues need to be resolved before a military topology with optimum performance and overhead can be defined. Continuing R&D and demonstration projects over the next decade will help define the basis for commercially successful standards, which may or may not suit military needs. Current commercial systems treat each application separately, whereas the military probably needs to take an integrated approach. However, many commercial designs, such as the following eight examples, could have military applications. These designs would need to be analyzed and perhaps modified before their application to military systems. Even so, the effort would likely yield a considerable savings to the DOD.
Enterprise Networks. Enterprise networks are being deployed worldwide based on standards such as ATM, X.25 (a standard interface for packet network access), frame relay (a potential successor to X.25), and TCP/IP. The products include comprehensive WAN solutions that encompass multiple technologies such as LAN/WAN interconnections, dynamic routing, accounting, statistical information, and performance monitoring. Some new architectures unite connection-oriented WANs with connection-free LANs over both narrow and broadband channels. Critical questions remain to be answered about the suitability of this technology in the battlefield environment; these are issues that could be addressed as part of a broader assessment of military network architectures. System planners need to minimize the vulnerability of centralized control points and determine the bandwidth required to distribute routing information updates on degraded channels as units move around the battlefield.
Cellular Telephone Systems. Commercial cellular networks consist of both analog and digital systems conforming to various standards. With cell sizes of up to 10 km, these systems can cover a broad area, sometimes including wireless local loop service. The technical feasibility of using
cellular systems in a battlefield environment requires further study. For example, cellular systems introduce special complexities because of the need to hand off transmissions between cells. Performing a handoff when system coverage may be incomplete over the theater is an elaborate process; equally complex are the rules for assessing when a handoff might be advantageous in light of possible jamming. Furthermore, mechanisms still need to be developed for authorizing access in a mobile tactical network. In cellular systems, these activities are performed in the cellular mobile switching office using the home location register.
Low-Tier Systems. In contrast to cellular systems, low-tier systems use low-power microcells of up to a few hundred meters in radius, small rather than tower-mounted antennas, and 32,000-bps voice coding for high quality and low delay. These systems are designed to serve users moving at pedestrian speeds. They are also suited to wireless local-loop applications because the round-trip delay is under 2 milliseconds, the quality of speech is comparable to that for wired services, and the short distances between hub stations and users generally result in low fade levels. Given these features, low-tier technologies could play a role in military communications.
Authentication and roaming capabilities are also provided in current-generation PACS and PHS systems. Packet-transfer protocols are being developed that will enable low-tier systems to serve as the transport mechanism for a wireless LAN while simultaneously carrying voice traffic. Commercially, the most successful low-tier system is PHS, which serves several million subscribers in Japan. The PACS system, developed in the United States, has not attracted significant markets to date but is the focus of active R&D efforts. Areas of investigation include techniques to incorporate flexible antennas and channel equalization to extend the range of PACS systems, especially to high-speed mobile applications. To determine the applicability of PACS and other low-tier systems in the battlefield environment, system planners need to address various issues, including security, antennas, adaptive waveforms, and operating range.
Radio LANs. The PACS system provides a protocol that is ideal for wireless LAN applications. The PACS packet channel (PPC) protocol provides the user with a variable bandwidth and asynchronous, asymmetrical data service at rates up to 256 kbps per radio port. The PPC converts the physical layer of PACS from a circuit-switched protocol to a packet-switched protocol consistent with TCP/IP. The current voice-PACS architecture is capable of circuit-switched connections, with a radio port control unit at the hub providing connectivity into the public telephone system. The radio LAN PACS has a distributed hub architecture, appearing
to the user like a device connected to a wired LAN, when in fact it is connected to a radio transmitter. A radio port can be configured to provide packet data on some time slots and voice on other time slots. Radio ports are small units that are easy to install, require no special towers, and can support up to 238 users each. The LAN PACS system can provide a quick, efficient means of installing a wireless LAN. The PACS technology is an open standard, which makes it easy to obtain equipment.
Local Multipoint Distribution System and Related Technologies. Ka-band (SHF) frequencies are now being used in wireless cable connections linking end users, urban fiber loops, and local telecommunications bypass companies. The frequencies involved are the 18–19 GHz band, known as the digital electronic message exchange; the 27.5–30 GHz band, known as the local multipoint distribution system (LMDS), which is used in some locations as a one-way television delivery system (''wireless cable"); and the 38–40 GHz band, which can be used for line-of-sight (LOS) data transmission at rates of up to 155 Mbps. The last band is used extensively in Europe for backhaul transmissions of personal-communications signals from base stations to mobile switching centers. Because LMDS and related systems offer high throughput, they could be useful to the military in short-range (approximately 5 km), nonmobile elements of untethered communications systems. Further study is required, however, because the system would require a high SNR to be demodulated correctly and might not be suitable in the presence of jamming.
Line-of-Sight Relays. The introduction of wireless cable systems has led to the development of high-speed LOS relays. These devices provide wideband access through a remote hub in areas lacking direct LOS access to a primary hub station. The relays generally operate at low power because the antennas pointing in both directions (i.e., toward the primary hub and the users) are parabolic dishes with very narrow beam widths rather than the sectorized antennas used in wireless cable systems. Among other advantages, LOS relays operate continuously and therefore do not require burst-mode modulation. Systems based on relays are actually simpler in design than are wireless point-to-multipoint systems.
Satellite Networks. The DOD currently takes advantage of VSAT and telephony Earth stations for information gathering and two-way, transaction-oriented traffic. Approximately 40 hubs exist around the world, which could be used to backhaul PACS or cellular wireless nets deployed in the field. Regional and global mobile-personal-communications systems (e.g., Iridium, ICO, Globalstar) are being developed for deployment around the year 2000. These systems are being designed to operate in
either L-band (UHF) or S-band (UHF/SHF) frequencies. Broadband Kaband satellites (e.g., Teledesic) are being designed to use high-speed, interactive, low-cost Earth terminals.
Satellite traffic service levels for LEO and MEO systems tend to run an average of 0.05 to 0.2 erlangs (i.e., the channel is occupied 20 percent of the time) per square kilometer. Planned systems are expected to have spot-beam capabilities that will increase traffic service levels by several orders of magnitude, but DOD battlefield communications needs will likely exceed the service level of any single system. Because the link margin varies with the square of the range, shorter links provide greater advantage. Therefore, battlefield communications services will need to be supported with a hierarchy of links that graduate in altitude according to the transmission distance. Satellite communications will play a role but many types of systems will be required, including handheld units, vehicular and aeronautical platforms, and high-altitude UAVs.
Network Management Systems. Network managers enable a central controller to monitor and change system parameters using standard software, graphical user interfaces, and relational databases. The controller isolates faults, produces status summaries, keeps usage statistics, and changes configurations. As these functions become standardized, the DOD could adopt commercial network management systems as a means of simplifying enterprise network operations.
22.214.171.124. Bandwidth Requirements
Battlefield communications currently consist mostly of voice and a very limited amount of text-message traffic. Communications equipment is not broadly available to individual soldiers below the noncommissioned officer ranks.14Approximately 10 percent of soldiers now have voice communications, and only satellites, certain aircraft, and smart missiles carry sensors for still imagery or video.
The digitized battlefield of the future is based on the concept, verified by the Gulf War, that extensive real-time data gathering and surveillance can improve situational awareness and battle management. Military commanders want an accurate, real-time image of the total battlefield that indicates the positions of friendly and enemy forces; provides still and video images as well as data from infrared, radar, and other sensors; and is integrated with communications systems to ensure that the right information is distributed wherever needed. The realization of this vision will require significantly higher bandwidth, both on a link-by-link basis and in the aggregate (bits per second per cubic kilometer), than is now possible on the battlefield. Indeed, assuming that all future ground troops
are equipped with communications and image sensors, there will be order-of-magnitude increases in both the numbers of users and the bandwidth required for imagery, as the data will flow higher and farther in the military hierarchy than voice typically does. Moreover, users will not be willing to spend half an hour delivering a single image, nor will their batteries tolerate such loads. New systems will be needed to provide higher bandwidth and hierarchically sensitive store-and-forward networking that will minimize the power required to transmit wideband signals.
The commercial sector has focused primarily on providing as many narrowband voice channels as possible. Wideband applications such as remote surveillance systems are becoming common in the industrial sector, and video-on-demand and video games are potential commercial markets. But unless substantial markets emerge for high-bandwidth services, the commercial sector will be slow to produce high-bandwidth wireless communications products that exploit the results of the third-generation R&D efforts described in Chapter 1. Military system planners have one advantage over their commercial counterparts in that mobile soldiers often operate within a short distance of a vehicle. In the digitized battlefield concept, many military vehicles could be equipped with radio equipment that can serve as high-power repeaters and may have very-high-gain antennas relative to those on the handheld units. Although practical levels of transmit power and antenna gain depend on the carrier frequency used, the availability of well-equipped vehicle platforms will make it feasible to transmit imagery to and from soldiers with handheld units.
The DOD's combined requirements for real-time data traffic and high bandwidth suggest that ATM technology, which is expected to be popular commercially, might be appropriate for battlefield systems, at least within the wired network. This technology operates at high bit rates through fiber-optic interconnections. Fixed-size data units (53-byte cells) facilitate efficient switching while connection-based semantics (i.e., virtual circuits) enable bandwidth to be allocated before a connection route is established. In addition, ATM switches can provide sophisticated mechanisms for choosing how to multiplex flows, thus offering a variety of QoS levels for different types of traffic. Finally, ATM supports message priority, queue management, and admission control mechanisms to yield performance guarantees.
The seamless integration of ATM and TCP/IP technology into the wireless battlefield communications architecture would not be a trivial undertaking.15Gateway functions will likely be provided at hierarchical RAPs, nodes that provide a variety of cross-networking, repeater, and other information services. The implementation issues are widely debated in the network community (and extensively studied in the ATM
Forum) because ATM's mechanisms for end-to-end QoS cannot be supported directly by current Internet protocols. An ATM-only solution is also problematic for many reasons, most significantly because it is impossible to guarantee QoS for wireless systems in motion (the link margin is continually changing and the signal can be lost completely).16Thus, wireless-link-level protocols of increased sophistication will be needed to ensure that data can be delivered successfully across the link while still meeting the QoS guarantees given during connection setup. The commercial sector is likely to resolve these technical problems eventually, but the DOD would do well to stay close to these debates to ensure that its interests are represented. Support for research and prototype development, coupled with testing of emerging technologies in battlefield exercises, could be useful government roles.
The use of GBS services will enable the direct delivery of wideband services to the battlefield and provide terminal integration opportunities for the "smart push, smart pull" concept (in which the warfighter receives customized data in a timely fashion without expending much effort to define the needs). Real-time, in-theater sensors will provide more recent information and more immediate task assignments than will DOD's satellite assets; thus the realistic information flow model is within the battlefield, with copies back to the Pentagon.17In the Army's vision for the digitized battlefield of the future, bandwidth is allocated not only up and down the command hierarchy but also horizontally to cooperating formations.
126.96.36.199 Source Coding
In current voice coding technology, speech is compressed with many different compression algorithms to bit rates ranging from 300 to 64,000 bps. Voice quality and compression factors have improved over two decades of research to the point that, in moving voice traffic, linear predictive coding (LPC) and other voice coding technologies enable compression ratios of 26 to 1. Even greater compression is possible to provide additional LPI and AJ advantages but at the cost of reduced speech quality and increased delay. Encoding of data has also been reduced to a well-known process, thereby providing a standardized method (e.g., using the Lempel-Ziv algorithm [Ziv and Lempel, 1978]) for both commercial and defense applications.
There is now great interest in coding of not only voice but also sound, images, and video. Reasonable-quality compressed video suitable for conferencing applications can be achieved with video encoders and decoders operating at 64 to 128 kbps, the equivalent of a dozen digital voice channels in commercial cellular systems. Videoconferencing will require substantially higher bandwidth and will need to demonstrate operational
effectiveness if it is to be used extensively on the battlefield. Similarly, high-resolution still or sensor images (e.g., a small image of 480 × 640 pixels at 30 bits per pixel) will require representations of well over a megabyte, far exceeding the capabilities of existing tactical radios such as SINCGARS, which transmits data at about 2.4 kbps (requiring an hour to transmit the image just described).
Standards are under development and chip sets for standardized implementation have recently become available for defense and commercial purposes. The most notable standards are the joint picture experts group (JPEG) and motion picture experts group (MPEG), which encode images, including the high-energy, low-frequency components. Image coding was developed for military reconnaissance but is also used in full-scale video teleconferencing; many military users are mobile and communicate through wireless links, whereas commercial users are stationary and connected by fixed, high-quality transmission systems. (Commercial-defense synergy is a tradition in source coding, as described in Box 3-3.) Image coding and video source coding technologies can compress a typical still image by a ratio of up to 100 to 1; newer technologies offer up to four times the compression of deployed systems.
Commercial-Defense Synergy in Source Coding
Source coding technologies have often been shared by the commercial and defense sectors. An early example was Sigsaly, a World War II voice communications system that relied on a tractor trailer full of security equipment (which today would be referred to as a channel bank "vocoder"). After the war, AT&T explored many commercial spin-offs of this technology, leading to the spectrograph and several technologies for voice coding. Source coding became more sophisticated as semiconductor technology replaced vacuum tubes and greater functionality was achieved in a smaller package and at lower cost. One result was continuously variable slope delta modulation, widely adopted in defense and civil government applications for secure communications. The next major step was the development of linear predictive coding (LPC), which uses a mathematical model of a voice signal and enables speech to be represented at 2,400 bps, a rate low enough to be combined with available modems to provide real-time, secure voice communications over dial-up telephone lines. This discovery was widely used in secure defense communication in the form of the STU III telephone, which continues to feature this data rate. Extensions to the LPC source coding technology developed to the point that the speech quality at low bit rates was acceptable for commercial purposes. These approaches were adopted in several digital cellular communication standards, combined with multiple access techniques, and used to develop unique standards.
Compression ratios and image quality are likely to improve with time, but the evolution of the commercial technologies has been constrained by the wide proliferation of the JPEG and MPEG standards and the tools built around these standards. Furthermore, commercial data and video coding standards have not yet evolved to be robust in the presence of the bit errors introduced by wireless transmissions. Continued improvements in source coding technologies are needed so that the bit rate required for faithful reproduction of information can be limited and the DOD can control the growth of its bandwidth requirements. An additional concern is the vulnerability of commercial source codes to jamming, an issue that could be addressed by research on adaptive waveforms.
The extent to which the DOD will use compression and decompression techniques to reduce its bandwidth requirements is unknown, and so it is difficult to predict the impact of advanced data file compression. However, because imagery represents a large component of the data traffic supporting defense activities, at least some benefits are likely. There would be some costs involved: New software would be needed for workstations, and new compression technology would need to be deployed into sensor assets.18But as the use of advanced collaborative planning and intelligent databases grows, specialized or localized compression strategies on database file transfers will help control the pressure for additional communications bandwidth.
188.8.131.52 Highly Adaptive Systems
Advanced modulation and smart radio technologies offer the promise of flexible, dynamically changing communications systems that can adapt to almost any conditions. Such systems will be able to select modulation, spreading code, and FEC and interleaving algorithms that will perform optimally in any environmenteven in the presence of noise, jamming, or interferencewhile also meeting QoS requirements for delay as well as data rate requirements. When applications can detect prevailing channel conditions and the radio system and external networking options can support adaptation, the user can gain at least an order-of-magnitude improvement in range, bandwidth, and AJ or LPI capabilities.
Ongoing research in this area has several shortcomings from the military perspective. The commercial sector is pursuing extensive R&D but its interest in adaptive systems is motivated by profit (e.g., accommodating more users per hertz), whereas the military seeks functional advantages such as increased interoperability and AJ capabilities. The DOD also supports research on adaptive systems, in part through the GloMo program, but the technologies are generally not demonstrated and tested under military conditions.
For example, existing network protocols are generally designed for static configurations and high-quality broadband links. In packet radio networks, discovery algorithms (which identify neighbors as each radio moves about on a network) determine the proper store-and-forward sequence for moving communications traffic toward specific destinations. These algorithms are now being developed, assessed, and standardized. To minimize overhead and streamline the discovery process, routing information servers are provided by the network that mobile units can query to determine who is connected and select the latest optimized routing paths to a specific destination. These algorithms have not been tested in military networks where mobility and network survivability under degraded channel conditions are of primary importance. The relative amount of overhead in a highly dynamic environmentincluding degraded channelsneeds to be modeled in more detail.
Military systems will benefit further if source coding, cryptography, and antenna beam performance can also be adapted to prevailing channel conditions. The TCP/IP suite tries to deliver data without errors. In wireless military applications it is likely to be preferable to trade off bit errors (and the delay tolerated by the application) against the quality of representation of the original information. In voice applications, for example, users are often willing to tolerate occasional distortions of isolated words (but not entire sentences) but are intolerant of delay. Reconnaissance systems may require that no transmission errors will be acceptable, whereas other image delivery systems may allow a moderate number of localized image distortions, in exchange for more rapid delivery.
Communications and network security have attracted attention recently under the umbrella of information warfare issues. Enhancements can be made readily in the wired infrastructure but are more difficult in wireless networks, especially packet radio systems for which protocols are still in the formative stages. Commercial and defense communications networks face different threats. Commercial providers are most concerned about the fraudulent use (theft) of service, whereas users of these systems are most anxious about improper access to and manipulation of their data. The military, which is often both provider and user, seeks to protect all aspects of its communicationsnot only the message but also the source and destination information, the inner workings of its equipment, and even the existence of a network.
184.108.40.206 Availability of Service
The commercial and military sectors have different concerns with respect to service availability. For commercial systems the primary issue is interference among users. This type of interference is well modeled, usually has stationary properties, and can be countered with thoroughly studied solutions. Hostile jamming in a military conflict creates a totally different type of interference, which cannot be mitigated using ordinary approaches. Hostile jamming can create a situation in which no usable, undistorted parts of a message are received. Other forms of jamming include intentional disruption of key information bits in messages, playback of old messages that are no longer relevant, or transmission of noise sequences to trigger false receiver actions.
A determined electronic attack on a military communications network could not be countered by any existing commercial equipment or any simple modifications to such equipment. Furthermore, it is unlikely that any future commercial system could satisfy military AJ requirements to the degree offered by defense-unique systems. Current military systems with AJ capability include the SINCGARS and Have Quick radios and MILSTAR satellite system. Because equipment used for jamming is becoming easier for potential adversaries to obtain, there is a growing need for development of advanced AJ techniques such as nulling and scanning antennas, spread-spectrum modulation, approved secure-spreading codes, elaborate error detection and correction, time-stamped messages, adaptive jammer-sensing techniques, and adaptive jammer-responding modems.
Access to commercial communications networks generally cannot be denied to segments of the population that have the proper equipment and, if necessary, are willing to pay for service. This feature is unattractive to military planners, who would prefer that communications systems offer normal service to friendly customers while blocking access by adversaries. Weather broadcasting, for example, is a vital part of tactical planning, yet most standard commercial systems cannot simultaneously guarantee the delivery of weather reports to critical commercial services while denying such reports to the enemy. A variety of technical approaches, similar to those used to control access to DirecTV and DirecPC, are available to safeguard broadcast digital data. Cryptographic codes and secret information exchanged in advance can enable selective access to some broadcast information.
220.127.116.11 Confidentiality and Integrity
Defense systems secure both the message data and, in a separate process, the routing information. Existing security systems cannot prevent
an adversary from detecting the presence of military communications. Information about the battlefield, important targets, and plans of interest can be inferred based on the volume of communications to and from particular locations, and much can be learned about routine military activities from the modeling of communications traffic. As stealth aircraft, ships, and other vehicles are deployed, it is critical to guard against the detection of these platforms based on their radio communications.
The DOD uses LPD/I systems to hide the evidence of radio transmissions. These systems rely on unusual transmission frequencies, spread-spectrum techniques, narrow-beam antennas, low-power transmission, or very brief messages. By reducing an adversary's awareness of transmissions, LPD/I systems also minimize jamming efforts and their impact. Communications systems with these features can be of great value in many military applications because they deprive adversaries of information about deployment of troops, routine versus unusual operations, the communications hierarchy, and level of alertness or activity. However, advanced LPD/I communications techniques have not been strongly supported in recent years.
Commercial systems offer considerably less protection. For example, cellular communications can be easily detected, jammed, or demodulated and user location can be pinpointed readily. Equally vulnerable are commercial CDMA systems, which are based on direct-sequence spread spectrum with published spreading codes and do not provide LPD/I capabilities. Current digital wireless standards make provisions for privacy and authentication to block unauthorized use, but cellular wireless standards make no provisions for traffic security, meaning that information about the routing, content, and significance of the data can be intercepted. Signaling information is sent in the clear, although the identity of the caller is protected by an identification number known only to the user and the system (or is temporarily assigned by the system if the user is roaming to a new location). In packet-switching protocols, ATM, frame relay (an interface for packet network access), and fiber-optic networks, all addressing and signaling information is left unprotected so that switching or routing equipment can also read and interpret addressing, routing priorities, and other information contained in the headers.
Commercial service providers are not expected to expend significant resources to harden commercial infrastructures against attack, although they want to prevent losses resulting from fraud. Careful design strategies will be required to deter both fraudulent use and attacks, threats that will be nearly indistinguishable on a mobile packet data network. The growing use of such networks will motivate commercial research aimed at solving these problems. Yet even if the commercial sector achieves significant security advances, the DOD is likely to prefer at least some of
its own approaches because COTS technologies would be readily available to adversaries.
One area where military R&D could be helpful is encryption, which protects voice, data, or video in an information frame or packet. A typical encryption system for real-time communications involves a cipher that encrypts the relevant bit stream one bit at a time. The decryption process requires both the relevant decryption key as well as synchronization between the sending and receiving process. (Synchronization refers to the assurance that a particular bit being decrypted in fact corresponds to the bit that was originally encrypted and sent to the receiver.) In principle, synchronization requires only knowledge of the starting point of the incoming bit stream, but in practice, establishing and maintaining synchronization throughout the duration of the transmission is complex. Current cryptographic systems are often inefficient because the synchronization consumes bandwidth and because synchronization may be performed packet by packet. (Any user of the STU-III secure telephone system can attest to the time it takes to achieve end-to-end synchronization.) The design of improved synchronization algorithms would streamline the security system and free up bandwidth for other uses.
3.4.3 Multimode, Multiband Communications
Interoperability has long been a goal of military systems. There are more than 17 different U.S. defense communications networks, and the sharing of messages among them requires the deployment of many unique information gateways or bridges. Furthermore, compatibility among U.S., North Atlantic Treaty Organization (NATO), European, and United Nations systems is increasingly important to military operations. For example, one report on the Bosnia Implementation Force notes that close air support missions can involve British Harriers, NATO E3-A (airborne warning and control system) aircraft, Norwegian forward air controllers, and Swedish-led brigades (Allard, 1996). The establishment of fully interoperable radio networks will require multimode, multiband communications capabilities, which are the focus of several DOD-funded research and demonstration programs.
The military ideal of a multimode, multiband radio implementing many different waveforms over a broad frequency domain does not have a commercial counterpart, although commercial technology that would support multiple standards is being explored under the European RACE and ACTS efforts. Existing multimode commercial systems have at most six different types of waveforms, each one generally restricted to a narrow frequency range, sufficient to access all of the wireless services that an international traveler is likely to need. The commercial sector is unlikely
to support additional waveforms because of the costs involved, whereas the military would almost certainly pay for them to gain the added functional flexibility.
18.104.22.168 Software-Defined Radio
Software radios are evolving in both the defense and commercial sectors. The military version is intended to enable interoperability among defense networks, reduce logistics support costs, and provide the capability to add new functions to fielded equipment through software updates.19The commercial work is driven by the need to accommodate the large number of standards used in mobile telephony. The design of common hardware for a wide range of applications would offer convenience to consumers and simplify manufacturing; however, the ultimate popularity of these systems will depend on whether they prove to be cost competitive with multiple dedicated implementations.
Several DOD-funded experimental models have been built. In field demonstrations, SpeakEASY was shown to be capable of receiving communications from the Air Force and translating them for the receivers and networks used by Army ground forces. The four-channel radio is compatible with some legacy waveforms and spans frequencies from 2 MHz to 2 GHz. The ACE, JCIT, and Millennium programs are not yet completed. Software radios are also being designed under the GloMo program to have adaptive interference-rejection capabilities. It is not yet clear whether any of these systems will offer the performance and cost effectiveness needed to initiate a production program.
Most commercial dual-mode digital cellular and personal-communications units can implement multiple transmission and reception formats using DSP software. Information about commercial radios still in development is typically not publicly available. There are undoubtedly plans to make software radios, which will likely be less flexible than are military versions. The commercial radios may contain software that is not intended to be changed after manufacturing. Furthermore, they will likely not offer the frequency range, extent of waveform synthesis, or sophisticated security expected for military applications.
Meanwhile, the commercial sector has focused intensive R&D efforts on various radio components to achieve incremental, practical advances. The DOD can expect to take advantage of the rapid commercial progress in many componentsA/D converters, DSP chips, RF amplifiers, display elements, processors, batteries, and storage deviceswhich will probably drop in price over the next several years. However, as discussed in Chapter 2 (Section 2.4), the DOD will likely need to develop its own specialized filters that can accommodate a broad range of frequencies and bandwidths,
as well as antennas that offer both frequency and beam-shape agility.
When all the functions of a radio are defined by software, the ''intelligence" and network services offered by the radio can be extended to greatly enhance military applications and perhaps eventually lead to intelligent radio services in commercial applications as well. Smart radios (i.e., radios capable of optimizing frequency, modulation, and protocols for a given purpose and signal environment) can incorporate the rules learned by an experienced communications specialist. Many simple rules define how to minimize interference. These rules can be applied in real-time, packet-based communications systems much more effectively than in traditional voice systems. Through real-time evaluation of each communication link and the spectrum in which the system operates, new levels of intelligence can be achieved to avoid jamming or to optimize transmissions under a wide variety of conditions (e.g., by minimizing battery drain, reducing traffic in the vicinity of hostile jamming activities, maximizing bandwidth or network capacity).
The introduction of the multimode software radio creates a significant opportunity for the convergence of many different systems and functions. Traditional defense platforms have separate systems for communication, navigation, identification, data exchange, signals intelligence, electronic warfare, and other functions. A software radio could be rapidly configured to perform any of these functions in any combination required. This convergence of technology will reduce the numbers of military systems procured while also increasing the cost effectiveness and utility of equipment. The resulting lightweight, agile platforms will be capable of rapid response to support the small units of fast-moving military forces now evolving. The increased availability, utility, and power of radio devices will create a new paradigm for military communications (see Table 3-2).
22.214.171.124 Co-Site Interference
Co-site interference, which is already a problem for military communications platforms, will worsen with the introduction of multimode, multiband radios unless new mitigation approaches are developed. Current technology designed to reduce the effects of co-site interference on radio performance is quite limited. Power combiners can connect up to five transmitters to a single antenna, but only if the frequencies are sufficiently separated. Receive co-site filters can suppress the carrier of colocated transmitters, but broadband signals are not suppressed adequately, and the broadband noise of transmit power amplifiers is not suppressed sufficiently at frequencies near the transmitting frequency.
Moreover, receive co-site filters become complex when the number of co-site transmitters is three or more, and receiver noise performance is degraded, resulting in reduced transmission range and an increased error floor. Co-site problems extend to antenna beam shape, which changes when antennas are used in close proximity to each other or to metallic structures. Because of the unique conditions on military communications platforms, R&D in this area will likely need to be supported by the DOD.
3.5 Defense Technology Policy Issues
The government influences private-sector technology development in a variety of ways. The instruments of government policy include indirect methods, such as investment tax credits, or direct methods such as federal funding for R&D and technology testbeds. Sometimes these policies are implemented to accelerate the development of strategically important technologies; at other times the motive is to ensure that equipment will be available for procurement by the government in a timely fashion.
Government policies supporting the development of appropriate defense technologies have always been a special case. In the past, when defense requirements generally guided private-sector technology advances (e.g., transistorized components), federal investments in R&D were not controversial. Now that sophisticated consumer and industrial products are developed independent of defense requirements, the need for federal investments may seem less pressing. However, the DOD needs to maintain a competitive advantage over potential adversaries with respect to warfare capabilities, including communications systems. The technology policy issue for the future is how to encourage innovations in electronics and communications technology that will dominate world markets while also ensuring that the U.S. military retains capabilities that exceed those of potential adversaries.
3.5.1 Implications of Changes in Military Tactics
The Gulf War demonstrated the way in which high technology permeates warfare. Advanced sensing, imaging, and targeting capabilities in the Patriot missile defense system, stealth aircraft, and other systems provided extensive advantages for U.S. forces. For example, Patriot missiles were aimed using surveillance satellites controlled from the United States. Liftoffs from Iraq were observed by these satellites within seconds, and critical targeting information was relayed through controllers in Colorado to the front-line Patriot batteries. This orchestrated activity demonstrated the capabilities of the U.S. military's existing global communications network, which required the support of high-bandwidth data links
to move sensor information both to and from the theater of action. But the Gulf War experience also suggests that communications advances are needed to enable rapid infrastructure deployment, logistics enhancements, and increased protection of technologies to prevent their exploitation by adversaries.
3.5.2 Rapid Infrastructure Deployment
During the ground war, the mobile forces moved so quickly that the communications infrastructure could not keep up with the front lines. Future communications systems will likely need to be rapidly deployable (and redeployable) so that they can keep pace with rapidly developing battles. Because Iraq did not react when U.S. troops first began arriving in Saudi Arabia, the coalition forces were able to build up an overwhelming combat strength in the Middle East as well as the logistical stockpile needed to pursue vigorous modern warfare. Adversaries in future wars are unlikely to be so accommodating, meaning that forces will need to be projected rapidly from the U.S mainland. Future conflicts are likely to be "come as you are," and communications infrastructures will need to support immediate action.
The recognition of this need has heightened interest in "instant infrastructures" based on satellite communications and mobile elements. The RAP has been proposed as a basis for a moveable front-line infrastructure with sophisticated, on-the-move antenna systems able to maintain high-bandwidth, point-to-point links with the rear-area infrastructure. To avoid the latencies inherent in satellite communications, hybrid systems that consist of DBS downlinks and UAV uplinks are being investigated. In general, these systems are viewed as backups to the terrestrial trunk linkages.
Continued military R&D investments will probably be needed because there seems to be little commercial interest in moveable infrastructures. One example of a commercial system with moveable elements is the Metricom multihop packet radio network, which operates in the unlicensed ISM bands in the San Francisco Bay and Washington, D.C., metropolitan areas. Although the infrastructure radios are in fixed locations, the multihop architecture makes it possible to add coverage in an incremental fashion through the addition of relay radios within the service area; bandwidth can be added also.
Future military communications systems will need new features corresponding to the reduced size of U.S. forces. Current planning provides
forces that are only sufficient to fight two regional conflicts at the same time. Instead of stationing so many troops overseas in areas of high tension, a split-base approach will be used, with advanced echelons overseas and the bulk of the forces on the U.S mainland. This approach will require high-quality, high-bandwidth connectivity worldwide, complete with access extensions that can be rapidly deployed, torn down, and reestablished as troops move.
Logistics tracking and management will be especially critical, given the growing need to transport materiel from the United States to the scene of the conflict. Many commercial systems are available. For example, OmniTRACS makes it possible to track vehicles continuously as they move and to plan routes efficiently. Package delivery services such as UPS and Federal Express have deployed sophisticated logistics systems for tagging packages and tracking them en route while also providing user-friendly on-line services that enable shippers to find their shipments. Wireless LANs were originally developed partly for warehousing applications. Finally, wireless tagging technology could provide the DOD with automatic inventory and location-identification capabilities, providing the basis for a complete logistical information system that could track the location of every item shipped.
3.5.4 Preparing for Unsophisticated Adversaries
There is some uncertainty about the technical requirements for communications during future confrontations with unsophisticated adversaries. Recent U.S. actions in Haiti and Somalia are examples of these types of operations, which may become more common as the United States plays an expanding role in peacekeeping and peacemaking missions. These countries tend to have little modern communications infrastructure, although this situation is changing as worldwide markets evolve for advanced technology. When deployed in less-developed countries, the U.S. military could bring along state-of-the-art commercial infrastructure technology. These systems would need to be shipped, installed, and operational within days, with military systems sufficing in the meantime. The commercial systems could transport the bulk of noncritical traffic, making it accessible to a smaller number of military-specific systems in the field.
In many ways, peacekeeping and other nontraditional military operations are similar to law enforcement activities, and many of the same communications issues need to be addressed. Even an unsophisticated adversary could disrupt service to U.S. forces using commercial systems. For example, cellular infrastructure is difficult to hide and could easily be targeted for sabotage. Although stealth and LPD are not always critical to defense communications, steps need to be taken to prevent adversaries
from learning of upcoming operations, performing traffic analyses, and intercepting specific types of communications traffic. The basic security and authentication mechanisms in the latest commercial systems can reduce interception by the technically unsophisticated; they are sufficient for nontactical communications traffic such as logistics support. Military-specific systems will continue to be needed for transmissions that require complete security.
The DOD might need cooperation and technical information from U.S. or foreign manufacturers so as to monitor the traffic of adversaries, track specific telephones, or infiltrate existing communications systems in particular countries. The U.S. military therefore needs to maintain a technical awareness of foreign-made equipment, perhaps as part of the effort to demonstrate, test, and procure COTS wireless technology (see Sections 3.2 and 3.3).
3.5.5 Preparing for Sophisticated Adversaries
Sophisticated communications technology is rapidly becoming a commodity. During the Gulf War some military specifications and procurement procedures were abandoned in an effort to get new capabilities, such as GPS, into the hands of the troops. Any adversary could buy the same sophisticated technologies; the threat is measured by how much the adversary can afford. Indeed, one of the implications of the Gulf War as a model for future conflicts is that the United States might not prepare sufficiently to recognize or defend against sophisticated adversaries.
A sophisticated adversary can be defined as one with the technical capability to build advanced communications systems or the financial resources to purchase what it needs on the global arms market. The greatest immediate threats are countries that can buy technologies from the countries that make them; for instance, the SCUD missiles used by Iraq in the Gulf War were based on the Chinese Silkworm missile.
To maintain a competitive advantage against these adversaries, the U.S. military could add military-specific modifications, such as security or waveform hiding, on top of commercial core systems. The military can leverage many commercial technologies, among them advanced ICs, DSP chips, and protocols. The advantage gained will depend on how these capabilities are integrated into defense systems and the choice and performance of the added military-specific capabilities.
The DOD has many reasons to use commercial communications products and practices whenever possible, building on a long tradition of
synergy between the two sectors. Many COTS technologies offer cost and performance advantages, and their quality is better than ever. The economies of scale achieved in mass production provide additional benefits and lessons that can also be exploited by the military. The selective use of commercial products and practices in DOD systems could help accommodate growing needs for global, untethered communications systems in spite of declining defense budgets.
However, the military will continue to have some unique needs that cannot be met by consumer products, or even future commercial R&D programs, because the motivations and interests of the two sectors differ. The DOD has unusual needs in three fundamental areas: network architecture, which influences all other aspects of a communications system; security, which encompasses confidentiality, data and system integrity, and service availability; and multimode, multiband systems, which can enable interoperability among diverse systems. The DOD needs to examine its needs in these areas carefully and probably pursue its own R&D in selected technology areas. All of these issues are addressed further in Chapter 4.
1. For example, advanced coding (Cacciamani, 1970, 1971, 1973) has been used in commercial satellite communications since the early 1970s for both data and highly compressed digital imaging, enabling the use of antennas on the order of 18 inches in diameter for digitally compressed video signals with link BERs less than 10-9. The best known of these technologies is probably CDMA, which has been widely adopted for cellular and personal communications systems worldwide. Encryption, along with data mining and RF fingerprinting, is increasingly being used to protect against fraudulent use in cellular systems, video entertainment subscription receivers, and business data. Finally, on-board digital processing will be used in the planned mobile telecommunications satellite and high-speed data satellites such as Teledesic.
2. A short lead time in a growing market can result in a large increase in market share. In addition, because prices can fall quickly after a new product is introduced, the first to market is often the only competitor to make a substantial profit. Yet a release date is often difficult to predict. Companies can be punished by the market if they fail to meet predicted release dates, as often happens, for example, with software upgrades.
3. The internal design cycle may actually be much longer because the basic equipment architecture is more likely to be on a two-year design cycle paced by the evolution of new semiconductor components. During the baseline design cycle of up to four years, anywhere from one to four design teams may be working on the next baseline architecture.
4. Many U.S. commercial wireless communications suppliers include divisions that have historically been involved in defense work. Within these companies,
cross-fertilization between the defense-related and commercial units may provide a mechanism for meeting military surge needs using the company's commercial products. However, this type of crossover is not always straightforward because of the differences between defense and commercial markets.
5. For example, current regulations regarding processors, A/D converters, and cryptography appear to reflect technologies that are nearly a decade old. The advent of common high-performance microprocessors enables the widespread development and use of cryptographic algorithms, which are often distributed on the Internet. The export of A/D converters is limited to technology of less than 8 bits, but advanced sigma-delta technology has only 1 bit (noise shaping and DSP techniques are used to increase dynamic range). Thus, the number of bits no longer seems like a useful metric for A/D converters; the metrics used to evaluate microprocessors seem equally outdated.
6. This is a simplified description of the decision-making process. More precisely, throughout the design, fabrication, and deployment of commercial products, trade-offs are made among performance requirements, standards requirements, cost goals, and design approaches to define a product that would be the most attractive and competitive in the marketplace. International, national, and regional standards determine many commercial design parameters, including off-axis emission from an antenna, maximum power flux radiated to Earth from a satellite, the capability of system users to coordinate or coexist with other users of a frequency band in the same geographic location, and numerous electrical safety regulations (e.g., related to wiring, batteries, radiation hazards, and chemical exposure).
7. Customers understand and expect this and are generally not willing to pay for a capacity that sits idle most of the time. Even during the busiest hour of the average business dayconditions that the systems are engineered to handlethere is a measurable probability of blockage that is calculated based on customer willingness to pay. Because the cost of a blocked call is usually only the effort required to try again shortly, there is little incentive to reduce the probability of blockage to zero. An interesting demonstration of the customer's acceptance of blockage and delay is the phenomenal growth of the Internet, where service is provided on a best-effort rather than guaranteed basis (although data services continue to come under increasing pressure for better service access).
8. Intel Corp., which after marketing its Pentium microprocessor found a design flaw in the precision of certain mathematical operations, uses a test suite comprising of billions of instructions to validate each possible instruction, register, arithmetic function, interrupt process, and instruction trap as well as sequences of events to prevent any surprises in complex applications. Only now are academic researchers considering more sophisticated theoretical techniques for dealing with testing processes of such enormous complexity. This research is critical to the future success of complex systems.
9. For example, commercial processes might take place at temperatures ranging from 0 to 50 degrees Celsius (°C) rather than -55 to 125 °C as in military processes. Or, commercial processes might involve 30 G of force rather than 1,000 G.
10. The NES is an encryption system certified by the National Security Agency
that enables clusters of defense computer networks to interconnect through the unclassified Internet. The NES provides high levels of assurance that a system communicates only with other systems that have comparable security levels.
11. Several military initiatives, including the Multilevel Information Systems Security Initiative and the DOD Goal Security Architecture, are intended to deal with various aspects of infrastructure in an effort to enable interoperability among systems. However, these programs have yet to field functions that enable communication between independent defense networks.
12. The ACN is designed to provide hierarchical communications over a broad theater of operations. Cross-linking and networking will enable various networks to communicate and access services through satellite links worldwide. The ACN will also serve as a repeater, picking up signals and rebroadcasting them over and around terrain obstacles, thereby extending the range of low-power equipment used on the ground.
13. In a base-station-oriented architecture, a greater investment is ordinarily made in the base station than in terminals. In such a network, both the transmit and receive link equations can benefit from the improved performance of larger antennas, more powerful transmitters, and more sensitive receivers. In typical systems the link advantage relative to the peer-to-peer design is approximately 10 dB.
14. Security is an issue in equipment deployment: The use of systems with cryptographic security requires procedures for securing clearances and equipment controls.
15. The TCP/IP protocol suite would need to be supported on top of ATM because the DOD has identified TCP/IP as the means for ensuring interoperability across heterogeneous military networks and because the entire system is unlikely to be constructed from native ATM technology.
16. An additional drawback is ATM's strong connection orientation, which makes it difficult to support mobility because existing connections need to be broken and reconstructed repeatedly. Furthermore, the ATM cell (i.e., data packet) structure was designed for the extremely low BERs of fiber-optic communications, whereas a radio fade can persist for several cell durations, making it difficult to use standard coding techniques to improve link quality. The loss of even a small number of ATM cells in a highly stressed network can dramatically reduce packet throughput.
17. The importance of distilling source information prior to transmission over a network is well understood in the commercial sector but remains an issue for the military, especially the Army, where communications, command-and-control, and intelligence functions are separate. There is no financial incentive on the part of the command-and-control and intelligence communities to spend resources to distill data at the source. Often the problem is passed off to the communications community, which is forced to transmit whatever is provided. For example, in situation awareness (SA) reports, positions are reported every 12 seconds regardless of motion. As a result the communications system is overloaded with SA reports. A more efficient approach would be to project positions based on direction and velocity and only send reports when the trajectory or velocity changes. But such an approach would require the development of software at a cost to the command-and-control
community. Instead the practice has been to blame the communications system for failing to support the traffic load. This situation would never arise in the commercial cellular industry, where providers take a systems approach and make trade-offs between bandwidth costs and source compression costs.
18. Mobile code, such as Java, might eliminate the need to agree on a compression standard because the delivery of executable code (along with the transmitted data) would allow the receiver to adapt to the sender's coding scheme.
19. An alternative approach would be to implement new functions in ASIC chips, which offer efficiencies in terms of power consumption. However, this approach would not provide an open architecture and might not be adaptable to future radio waveforms.