Conclusions and Recommendations
The history and challenges of wireless communications, as outlined in previous chapters, suggest a variety of strategies that could be pursued to fulfill the vision for untethered military communications systems. This chapter summarizes and integrates key points made in the preceding chapters to provide a set of 12 recommendations directed to the DOD and DARPA. Organizational changes are recommended that would provide an environment conducive to the development and military application of state-of-the-art commercial technology. To meet defense-unique needs, specialized R&D and demonstration efforts are recommended that focus on various aspects of wireless technology, from the highest network level down to individual components.
4.1 History And Challenges Of
Voracious consumer demand is stimulating many advances in commercial wireless communications technology, particularly cellular and cordless telephones. The portfolio of wireless services now available in the commercial marketplace includes a wide range of telephony, paging, and data applications delivered over a variety of service offerings ranging from land mobile radio to cellular to satellite communications. Each service offers a unique combination of coverage region, bandwidth, subscriber equipment properties, and connectivity. In the aggregate, commercial wireless capabilities are considerable, yet many technical challenges
remain. The cost of wireless voice systems needs to be reduced and their quality improved. Specialized wireless data networks have not taken off as yet, perhaps because they are not powerful enough or because mass market applications have yet to emerge. Considerable research to address these and other issues is under way, both in the United States and overseas. Industry road maps suggest that, by early in the twenty-first century, commercial wireless communications will meet the long-term goal of enabling users to communicate ''anytime, anywhere."
The DOD uses a variety of wireless systems that are based on 1970s and 1980s technology and designed to serve specific needs. The DOD no longer drives the evolution of state-of-the-art communications technology but still needs access to it, perhaps more than ever. As threats to peace change from global to regional conflict, a transformation is taking place in military roles, missions, and communications needs. The vision for military communications stresses C4I and the protection of the lives of U.S. personnel, who will be based principally in the United States but will need to be prepared to move quickly throughout the world to carry out a variety of missions, including noncombat roles such as peacekeeping and humanitarian response. Such missions are nontraditional in the sense that coordination with foreign partners may be essential, whereas national survival will not be at risk as was anticipated during the Cold War. In addition, the need for U.S.-based logistical support will grow, and new systems will be required to counter terrorism. Thus the accurate, timely transmission of information will be perhaps more essential than ever in meeting military objectives. Effective global communications systems will be critical.
The civilian and military sectors have a long history of interaction in the design and deployment of wireless communications technology. In the Gulf War, DOD used commercial wireless equipment such as GPS receivers and found that the performance was comparable to that of equipment designed explicitly to meet military needs. Yet current commercial technologies and practices cannot meet all military needs. For example, the military cannot tolerate the long lead timeson the order of months to yearsthat are typical in the building of commercial communications infrastructures. Commercial wireless companies carry out elaborate advance planning and measurement operations, whereas the military requires networks that can be organized quickly and can adapt rapidly to changing operating conditions (including spectrum availability). These networks will also have to be compatible with other military communications systems, both new and old.
Differences between military and commercial needs also have implications for network architecture. Commercial research on integrated (i.e., multimedia) systems is oriented toward network architectures based on the base-station-oriented model. It is not clear whether that approach or
the peer-to-peer design will be more appropriate in future military settings. Previous DARPA research on packet radio networks has encompassed base-station-oriented, peer-to-peer, and multihop networks.
The evolution of technology is also influenced by organizational differences between the two sectors that encompass market drivers, acquisition practices, and maintenance and repair arrangements. Commercial R&D and manufacturing are oriented toward mass markets of tens of millions of units. Adding functionality to designs has to be justified by consumer demand. By contrast, military equipment is designed to provide the functions required to fulfill missions. The number of units produced is tiny relative to commercial markets. Commercial customers acquire equipment and subscribe to services as they are ready to use them, expect effective function within minutes, and rely on equipment manufacturers for repairs. The military acquires equipment on a contingency basis, operates its own repair facilities, and is prepared to train its personnel to use communications systems.
Many research efforts are under way to realize the commercial and military visions for wireless communications. Fueled by the success of cellular communications and projections of ever-expanding markets for wireless services, the commercial sector is pushing ahead in various areas. One objective is to enable portable devices to communicate at the high bit rates needed for advanced information services. Another objective is to advance the state of the art for software radios as a means of fostering economies of scale in R&D and manufacturing in a world of diverse and changing technical standards. By using multiple types of operating software, such radios can serve as single hardware platforms capable of transmitting and receiving signals that conform to a variety of standards. Meanwhile, DOD is taking a dual approach to wireless technology development by both conducting its own research, focusing primarily on components, while also relying increasingly on commercial technologies to ensure interoperability and systems integration. The DARPA GloMo program has initiated a broad range of coordinated R&D efforts that will provide enabling technologies for future military systems. Among the commercial technologies that the military expects to use are Internet protocols and ATM.
These observations lead to the following general conclusions:
These conclusions provide the basis for the recommendations presented in the remainder of this chapter. The 12 recommendations are organized in a hierarchical order from the general to the specific. The first three focus on organizational changes, or meta-issues, that need to be addressed by DOD to help align military requirements with commercial products and services and provide an environment conducive to the absorption of state-of-the-art technologies. The other nine recommendations identify R&D project that should be carried out by DARPA to advance the synergy between military and commercial systems while also meeting specialized military needs that will exceed anticipated commercial developments. The R&D recommendations are presented in order of priority, reflecting the committee's view that high-level systems issues are of paramount importance. The recommended research excludes subjects that will be adequately covered by the commercial sector.
4.2 Standards Development
1. The DOD should participate in standards-setting activities for wireless communications technologies and systems.
With commercial demand for wireless technology growing worldwide and DOD budgets flat or falling, incentives for future commercial-military synergy need to be provided by the military side. Consequently the defense community needs to gain a deep understanding of technology trends so as to obtain advance notice of new concepts and influence the development of cost-effective equipment that meets military needs. Although new technologies can originate in diverse settings that include industry, academia, and nonmilitary government laboratories, the features of available equipment are determined to a large extent in the process of standards setting.
Current R&D and standards activities are focused on enhancing wireless communications technology to make it possible to move many types
of information (including data, video, and images) to and from portable wireless devices. Although this work is certain to produce new technology, the commercial deployment of these innovations is not assured. The availability of innovative technology in the marketplace depends on business, social, and government policy factors. Military planners need to maintain a continuing awareness of the difference between what is possible technically and what is available in the market to meet military needs.
Because military requirements often exceed those of the commercial market, opportunities for the DOD to use commercial products depend on equipment details, which are reflected in standards. Such details might be of little initial interest to consumers. Yet the defense community is often ahead of industry in recognizing features that will eventually be important to all users; if these needs are not addressed in the standards-setting processes then they might have to be met later in commercial settings at great cost, both financially and in terms of network performance. An example is the poor state of network security in analog cellular systems, an issue not considered in the initial design process. Newer digital systems help alleviate this problem, but adding security features to older equipment remains an expensive and difficult challenge. The military, by contrast, always plans for system security in advance.
Standards also influence whether "hooks" or interfaces are designed into commercial technologies to enable modifications that would meet specialized military needs. By participating in the standards process, government agencies will make it possible to embed standard devices in military-specific system architectures and generally promote a capability for cost-effective systems integration.
The DOD could influence standards setting by participating in activities such as the ATM Forum, the IETF, and the Multimode Multiband Information Transfer System Forum. Effective participation will be constrained by the rules governing standards organizations and by the abilities of DOD management and participating individuals to influence technical decisions and political processes. They will need to understand how standards for wireless communications are established, a complex process that has influenced the very different evolutionary paths of wireless technologies in the United States, Europe, and Japan. In addition, the DOD could benefit from the analysis, simulation, and laboratory testing of candidate technologies to support defense interests in the establishment of specific standards. These activities could reveal the limits of some technologies that would not be apparent in more benign commercial settings but are likely to be of long-term significance to consumers. By conducting such tests and sharing the results with the commercial sector, defense agencies will improve the chances that adopted standards
will serve immediate military needs and also benefit the commercial sector in the long run. Participation in standards creation could ultimately prove to be more cost-effective than commissioning equipment explicitly designed to meet military needs.
4.3 Demonstration And Testing
2. The DOD should pursue a vigorous process of technology demonstration and testing prior to development and procurement. In particular, the focus should be on system concepts based on commercial technologies and specialized military enhancements.
As the defense budget is reduced, less money will be available for major procurements. Nevertheless, an inventory of advanced technologies can be maintained and deployed as needed if military equipment is adapted whenever possible from commercial technologies. Examples of this strategy include the Condor project, in which DOD is supporting the development of a cryptography module on top of the cellular telephony system, and the Army's plans for the digitized battlefield. The Condor project is focusing on core noncommercial technologies such as on-the-move, high-bandwidth, phased-array antenna technology. The Army's digitized battlefield network will include switches, routers, and hubs based on commercially available technology. The Army is not building its own ATM switches, which are widely available in the private sector, but instead is developing high-speed encryptors and decryptors that are compatible with commercial switches.
Military planners need to understand the thrust of commercial developments and identify critical technologies that are unlikely to emergesoon enough or everfrom the commercial sector. These technologies need to be developed by the DOD. For all other elements of wireless systems, new commercial developments need to be tested and evaluated on a continuing basis in military exercises to determine and verify their suitability for military use. Resources available to support this approach include the federal defense laboratories, which can provide a critical bridge between the military's operational needs and private technology development, and the defense industry, which is well equipped to collaborate in the development of prototype military-specific equipment. In addition, both the DOD and industry will continue to rely heavily on U.S. colleges, universities, and technical schools to provide competent engineers, technicians, programmers, operations staff, and scientists as well as key technological breakthroughs and innovative ideas vital to U.S. military and commercial competitiveness.
The DOD's technology demonstration and testing efforts need to take into account the unique communications needs of each military service. A mechanism is needed to help integrate the requirements of these varying systems and find a common architecture that would cost-effectively support the majority of those requirements. This task might be carried out by an R&D subunit such as DARPA or an existing Pentagon-level joint program office. In the past the military has focused on one radio subsystem at a time, with the result that interoperability has been minimal. The vision of the future calls for the development of an overall system concept with a standard set of multimedia protocols and waveforms to assure interoperability at all echelons, with individual radios specified and procured to work within this architecture.
3. The DOD should plan a new approach to procurement that will identify now commercial infrastructure systems and subscriber equipment can best be used for military purposes and how to purchase commercial equipment in the most productive way.
The DOD needs to develop models to analyze how best to use commercial systems and equipment. For example, a planning system could be created to keep track of commercial communications infrastructure and service access points within a given geographical region. Such a system needs to be comprehensive, monitoring telephony, data, microwave, satellite, and fiber-optic services. The analysis could suggest the most effective means of information delivery for each source-destination pair, identifying the gateways and networking software required for interoperability and backup paths. This information could be used for planning or real-time robust network management, performance optimization, and repair.
A system is also needed to evaluate the potential for defense equipment based on commercial technology by comparing the cost-effectiveness and performance of legacy hardware to those of COTS products. In many cases COTS components may represent improvements for the military in terms of cost-effectiveness, power effectiveness, or other important features. The planning system needs to be capable of determining how to apply these internal components to anticipated defense applications. As part of this process the DOD needs to develop the expertise necessary to translate its operational needs into requirements for commercially available equipment.
Once the DOD identifies how best to use commercial technologies, cost-efficient acquisition strategies need to be pursued. Governmentwide
purchases might achieve greater economies of scale than would purchases by individual military services. The DOD could use this opportunity to become a more effective customer. Government acquisitions often add specialized requirements that prohibit commercial suppliers from offering bulk rates. As an alternative, DOD could explore how to purchase bulk quantities of standardized equipment and then separately acquire enhancements that align a system with military requirements. It might also be possible for designated agencies to acquire equipment and then distribute it to other government users. For example, the Army Signal Corps could acquire all DOD communications equipment and provide for all maintenance and upgrades on the basis of annual agreements. Consolidated purchasing might enable cost-effective reliance on commercial providers for maintenance, logistics, and training.
The DOD can also foster commercial-defense synergy by allowing multiple vendors to build subsystems that meet open hardware and software interface standards and by selecting, from the field of possible vendors, those most capable of creating an ongoing competitive production over a long, sustained product life. When procurement programs keep at least three vendors competitive over the course of the system, the competition encourages the evolution of advanced features and improvements in cost-effectiveness. The STU III is an example of a fully competitive acquisition in which three vendors competed for production of several hundred thousand secure telephones. In this controlled market, each vendor was required to meet open interoperability standards, but each was also allowed to implement unique features and functions to attract market share and compete on price. Each competitor's model updates kept the other competitors busy matching features and prices, a process that benefited both the users and the government while also motivating ongoing technology insertion. The SpeakEASY radio is another example of open system architecture. An "open system forum" allows contractors to participate in setting hardware and software standards that specify open interfaces between system components.
4.5 Modeling And Simulation
4. DARPA should build on current research in modeling and simulation to incorporate the communications traffic, mobility of network elements, and radio propagation encountered in mobile military information networks.
The performance of wireless communications systems depends on three phenomena: communications traffic, mobility of network elements, and radio propagation. Accurate assumptions about these phenomena
need to be made if the DOD is to design, deploy, and operate effective systems. Modeling and simulation are the best available tools for optimizing system design and predicting performance.
Various commercial packages are available for simulating communications systems consisting of established components and subsystems. The quality of the results they produce depends on the accuracy of the models of traffic, mobility, and radio propagation that they incorporate. Good models of narrowband radio propagation are available. However, models that incorporate site-specific characteristics of buildings, terrain, and foliage are not commercially mature, and there are no radio propagation models that incorporate adaptive antennas. Furthermore, available "teletraffic" models apply to only a few simplified conditions, and mobility models are confined to abstract formulations (e.g., fluid flow, Brownian motion) that have not been validated with reference to practical conditions. As a consequence, existing tools cannot provide realistic analyses of complex DOD systems. Much more is needed, especially in the site-specific channel modeling, traffic, and mobility areas, to provide military planners with the tools necessary to rapidly deploy wireless systems in battlefield scenarios in a wide range of modern theaters (e.g., built-up urban areas). DARPA should stimulate research to derive and validate models of wireless channel effects using modern modem and antenna technology so that appropriate protocols, applications program interfaces, and optimization algorithms can be developed.
Research leading to techniques for accelerating the run-time of complex wireless system simulations would benefit the military and commercial sectors alike. Simulation of communications systems is often performed by signal-processing workstation tools, which enable programmers to define a complex communication system by connecting icons representing predefined building blocks. This approach enables programming to be automated, but the resulting simulation code is often inefficient to run. DARPA research in this area could build on the impressive progress already made in the S3 program using parallel processors to simulate communications systems. This program includes a wireless component that could be enhanced to simulate the traffic, mobility, and radio propagation conditions of military communications.
DARPA could also establish libraries of code representing communications waveforms, protocols, source coding techniques, and network interfaces. This research would accelerate system design while also reducing the expense of simulation software maintenance, which typically costs 10 times more than software creation because of the longer time frame involved and rapid advances in technology. Research on system development and maintenance methods will help identify the most effective strategy.
Finally, integrated tools are needed to assess the performance of the subsystem elements of a software radio operating within a large-scale network in motion over a broad geographic area. These tools would enable researchers to explore network quality, connectivity, stability with respect to multiple performance criteria, and unexpected problems. Some research groups are developing their own tools to investigate specific aspects of performance. DARPA could attempt to provide an integrated capability built on a common application interface that would reveal the effects of each system component on the performance of the entire system. This capability would enable the optimization of an entire system rather than just the individual components.
4.6 Network Architecture
5. DARPA should initiate research to produce network architectures that incorporate commercial products in a manner that meets military requirements.
The performance of a wireless communications system depends in large part on the coordination of network elements. The architecture of a network defines these elements, identifies pairs of network elements that communicate directly, and specifies the protocols for that communication. Architectures for wireless systems differ according to how terminal modems are connected (peer-to-peer versus base-station-oriented design), how infrastructure elements are connected (hierarchical versus distributed), and the nature of communications with other networks. Given the special needs of military wireless communications networks, designers need to adopt a system architecture that takes maximum advantage of the capabilities of commercial devices and subsystems yet also provides unique interfaces and protocols as dictated by military needs. DARPA research in this area is likely to reveal new network architectures that use commercial products and services in an innovative way to serve military aims and make it possible to gracefully absorb new technology at the subsystem and component level.
Within a new network architecture, many key issues remain to be resolved with respect to protocols for wideband data services. Current commercial systems treat each application (e.g., Internet, telephony, video dial tone) separately instead of taking an integrated approach. Recent discussions of nomadicity underscore the importance of considering wireless and mobile components and conditions as part of heterogeneous and interoperating network contexts. Military wideband packet radios, which are designed from the outset for integrated information services, could provide a useful model for future commercial developments. However,
research is needed to determine network topologies and protocols that make full use of the capabilities of advanced radios.
In creating a new network architecture, protocols and algorithms need to be optimized to meet the objectives of military operations. This is a complex task. As an example of the complex interdependence of algorithms and architecture, the design of effective routing techniques needs to aim for the following objectives:
The difficulty of achieving all these objectives is compounded by the need to run several different applications on mobile terminals at any given time. Therefore, the routing protocol needs to process each packet individually to meet the QoS objective of the transmitting application.
In a mobile wireless network it should not be necessary for every terminal to assume the size, weight, power, or cost burdens associated with critical network services such as database servers, image servers, speech recognition and synthesis, transcoding, transcrypting, position location, and health and safety reporting. Research on network architecture can identify the best way to distribute services among network elements with the aim of minimizing the weight and power consumption of devices carried by personnel.
4.7 Network Security
6. DARPA should conduct research aimed at understanding and bridging the differences between security needs in commercial and military networks.
The commercial sector is developing a variety of data networking products and services that are likely to be integrated into single-hop base-station-oriented architectures. However, as noted in Chapter 3, the requirements for security in commercial networks are likely to be different from those in defense applications. For example, troops on the ground or ships at sea that do not wish to disclose their location require high
LPD/I performance levels, a requirement not faced by most commercial users. These differences need to be understood and accommodated if the military is to make use of commercial wireless technologies.
Research is needed to improve the information-processing features of encryption techniques. For example, the end-to-end encryption of all traffic would obviate the need for interfaces or bridges between interacting systems. However, current end-to-end encryption algorithms require considerable network overhead to establish the cryptographic synchronization. DARPA could seek improved methods of end-to-end cryptographic synchronization through multiple networks with lower overhead than is currently possible. With such methods, the time required to establish cryptographic synchronization between active network members would be reduced significantly.
DARPA also needs to design network protocols that allow for commonality of hardware and software interfaces as well as security differences that meet the needs of all applications. These protocols need to provide multilevel security, long identified as important for defense communications. The concept entails the shared use of a network by individuals with differing authorization to access information of differing levels of sensitivity. Multilevel security implies the management of access to computing and communications systems and to information transmitted or stored in those systems commensurate with individuals' authorization. In general, the design and implementation of multilevel security systems have been imperfect, and wireless and mobile applications compound the challenge.
4.8 High-Density Communications Platforms
7. DARPA should conduct research aimed at reducing co-site interference.
Military ships, combat aircraft, UAVs, and mobile battlefield systems all operate a variety of communications systems in close proximity to one another. The placement of numerous radios in the same general location typically results in interference and many compromises. Current technology designed to reduce the effects of co-site interference on radio performance is quite limited. For example, 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.
DARPA needs to catalog the co-site enviromental of military platforms, identify both common and unusual problems, and design more effective
solutions that are useful over a broad spectrum, such as 2 MHz to 2 GHz. Specific co-site problems that require attention include the following:
This R&D would be timely because the full deployment of software radios operating in broad frequency ranges will depend on the development of techniques to overcome co-site interference and high-power in-band interference. When serving several subscribers in the same band and mode at the same time, SINCGARS radios would experience significant interference. Significant technical challenges are associated with co-site interference generated within the software radio itself. Additional issues arise in the effort to enhance mobility of forces. For example, consolidating multiple mechanized vehicles will produce a single command track, which now has six SINCGARS radios and the usual problems associated with multimode radios, regardless of whether the radio architecture is analog, digital, or software based.
Research is also needed to minimize the interference caused by the spurious emissions (spurs) and harmonics of co-site transmitters and receivers. When many radios are co-located or networked, it might be possible for each transmitter-receiver pair to agree on a common frequency plan to minimize spurs and harmonics or on common filtering plans to minimize co-site degradation. Additional research topics could include time and frequency management (to prioritize transmissions) and reception time slots (to minimize the impact of transmissions likely to cause interference to other receive channels). Such research would need to take into account the interleaving and coding present in each channel and the likelihood of interference. Finally, a new generation of highly linear filters is needed with improved agility and a wider range of operating frequency than is currently possible.
4.9 Software Radios
8. DARPA should carry out research and demonstration projects designed to field software radio technology for military applications.
Software radio is a far more versatile technology than the term ''radio" implies. These radios can operate as part of a network and perform a vast array of electronic and computational functions (e.g., database management, transcryption) through the downloading of software. A networked radio can function in ways not envisioned when the component is manufactured. In a military scenario, some radios can function as active interrogators while others act in a passive manner, undetected but coordinated by active units. Similarly, software radios can perform services that until now were unique to each application.
A software radio could be a leading part of the C4I infrastructure: With appropriate software it could be applied to signal intelligence, electronic intelligence, communications navigation and identification, electronic warfare, information warfare, electronic countermeasures, missile tracking, guidance, or commercial paging or telephony. The cost-effectiveness of software radios increases with the number of available software functions, the ease of performing new tasks, and the ease of cooperating with other network systems to accomplish larger tasks. The multiple roles that can be played by software radios could have an impact on DOD's organizational structure, as services provided by individual organizations and procurements are combined in one system.
Commercial software radios are likely to have more limited capabilities (e.g., changing signals and bandwidth on a single frequency) than are military versions, which are being designed to span large frequency ranges and implement many legacy waveforms. However, the commercial sector is achieving rapid advances in many software-radio components, such as A/D converters, DSP chips, RF amplifiers, displays, batteries, and data-storage devices. The DOD can use these COTS products to good advantage. At the same time, DARPA needs to undertake specialized R&D focusing on antennas (Sections 4.10) and filters (Section 4.12) for military applications. Exploratory research on novel components and designs could also be beneficial (Section 4.13). In addition, to identify any necessary improvements and make optimal use of this promising technology, DARPA needs to demonstrate software radio technology on defense platforms where density, power, and weight are critical.
4.10 Smart Antennas
9. DARPA should conduct the research needed to adapt smart antennas for mobile military applications.
Commercial applications for smart, adaptive antennas are limited to relatively low-cost and unsophisticated single-band units with limited flexibility in beam pattern. Moreover, virtually all existing adaptive antennas for mobile radio applications are designed for use at base stations rather than mobile units. Military applications require antenna functionality for several orders of magnitude of frequency coverage as well as electronic tuning, the coupling of more than one transmitted and received signal, and a diversity of beam shapes ranging from omnidirectional to pencil beams. Several technical challenges need to be overcome before such antennas can be produced. DARPA research needs to pursue the following objectives, among others:
4.11 Smart Waveforms
10. DARPA should conduct research to produce transmission techniques that adapt to a wide range of operating conditions.
Commercial communications systems tend to operate under predicable, stable conditions. Therefore, commercial waveformscharacterized by frequency band, bit rate, modulation method, and source coding and channel coding techniquesare designed for operation over a limited range of conditions. For example, high-tier systems are designed for one set of conditions, whereas low-tier systems are designed for a different set of operating conditions. The operating environments and transmission conditions encountered by military communications systems are more unpredictable and subject to change. Military systems would therefore benefit from transmission technologies that adapt to changing conditions.
Recent research has produced theoretical results on the optimization of bit rate, modulation method, source coding, and channel coding techniques. New DARPA research should be aimed at uniting the theory of adaptive modulation and coding with the emerging technologies of advanced software radios (Section 4.9), which promise a practical means of implementing adaptive schemes. The results of the recommended modeling and simulation research (Section 4.5) would provide valuable tools for performing this work. In addition to meeting military needs for adaptable
systems, this work could serve the long-term needs of the commercial sector by providing core technologies for integrated personal communications systems that combine the advantages of several of today's separate systems.
One objective should be to develop modulation and coding strategies that allow the demodulator to acquire the properties of interference in the time domain, the frequency domain, or the constellation domain so that the interference can be avoided or canceled. These techniques would concentrate energy in the regions of frequency, time, or code space that provide the greatest link margin. Another objective should be to produce transmission techniques that establish a data rate consistent with interference and fading conditions. These techniques should allow for selection of a data rate over two or more decades of bit rate. They should also be designed for rapid carrier synchronization time, rapid timing synchronization, and minimum training time for learning channel conditions. The choice of data rate should be consistent with adaptive source coding, which gracefully degrades the perceived quality of the end-to-end application as the channel quality declines.
In addition, for defense applications new waveforms are needed that allow for very high spread-spectrum processing gain. Typical spread-spectrum techniques require considerable computational power to be detected by the desired receiver. The high signal-processing gain required to jointly discover frequency offset, baud boundary (or baud boundary offset), and spreading code alignment at these high processing gains is inconsistent with traditional portable radio design. In addition, new spread-spectrum waveforms are needed that allow for further reductions in the detectability of transmissions, direction, and properties of the spread signal.
4.12 Filter Technology
11. DARPA should conduct research to overcome the limitations of current filter technology for use in military software radios and high-density platforms.
Both software radios and high-density platforms require advanced filters. Filters currently constitute 25 percent of the volume of a typical software radio and are used in receive preselectors, power amplifier output filters, local oscillators, and mixers. To a great extent the radio receiver's sensitivity and dynamic range are determined by the selectivity and the losses of the preselectors, and the radio's co-site performance is determined by the selectivity of output filters and the filters in the local oscillators and upconverters. Active and digital filters are not appropriate for these functions because they introduce noise that degrades performance. Handheld
software radios would be improved by the miniaturization of filtering functions and improvements in frequency tuning range and selectivity.
There is also a need for filters that cover wide frequency ranges. Commercial radio equipment spans either a small frequency band or a few selectable bands. Military radios that span wide frequency ranges (such as 2 MHz to 2 GHz) require lumped filters built of inductors and capacitors to handle the low end of the frequency range and transmission-line techniques for filtering at the high end. Research is required to identify new circuit and materials technologies that allow for tunable, highly selective filters that span this entire range, operate at higher power levels, and take up less physical volume. In addition, new processing techniques are needed to enable the monolithic integration of highly selective filters with semiconductor devices and to produce multisection filters in more sophisticated shapes.
4.13 Novel Components
12. DARPA should develop novel components to enhance the flexibility of software radios.
Software radios have been significantly enabled by novel components, notably DSPs and FPGAs, which allow new waveforms to be added to fielded systems through the installation of new software and hardware. Additional novel components could further enhance the flexibility of software radio architectures. For example, components could be designed to reduce the equipment size, weight, and power needed to accommodate military designs that incorporate a wide range of potential future waveforms and large numbers of legacy waveforms. Radio performance could be improved through research into new DSP architectures that have adaptive resolution, clock speed, instruction sets, memory architectures, and arithmetic functions designed to the specific signal processing of communications systems.
Similarly, novel FPGAs that either interconnect analog circuit elements or integrate analog and digital operation could be of great value. A chip containing analog circuit elements and FPGAs could be applied to the analog front-end functions of many communications systems while also providing a broad range of interfaces to other systems. Research is required to identify the semiconductor processes and circuit topologies that provide sufficient isolation between the various analog functions, interconnection with minimal loss, and circuits with sufficient versatility to be configured for many applications.
Finally, the monolithic integration of nearly the entire software radio function could provide experience in combining analog and digital signal
functions on a common substrate. It is currently difficult to keep digital noise from coupling into analog circuitry and degrading performance. Research on how to implement wide-dynamic-range analog circuitry monolithically with digital circuitry would provide the basis for implementing an entire software radio on a single component. Such an effort needs to encompass the monolithic implementation of high-performance filters that are tunable over large frequency ranges.