Microwave Journal
www.microwavejournal.com/articles/610-next-generation-radios-new-paths-for-communications

Next-generation Radios: New Paths for Communications

October 1, 2005

Anyone who deals with military communications systems today is certain to come face-to-face with the concept of network-centric warfare. The heart of this concept is a reliable and ubiquitous communications system. Success in combat requires that the necessary information and communications links always be made available exactly when and where needed (see Figure 1).

The Challenges

Traditional radio communications systems do not meet this requirement. They are usually designed to meet the needs of a largely closed group of users. Accordingly, they use waveforms that have been optimized for the mission at hand and whose characteristics reflect the state of technology at the time they were developed. In today’s terms, a waveform means the entire functionality involved in a radio between the input and output of voice, data and video at the user interface and the signal on the antenna. This primarily includes protocols, coding and modulation. A few examples of waveforms are SATURN, Link 16 and TETRA. The main weaknesses in communications systems and conventional waveforms are a lack of interoperability, insufficient networking capability and low data throughput. A large number of incompatible waveforms are in use today. Current systems can handle only one or, at most, a few of these waveforms. This leads to user groups that can communicate with each other only to a limited extent. However, armed forces are increasingly collaborating in various groups (task forces). Traditional systems cannot meet the requirements of these mission concepts. In addition, most waveforms do not allow information switching via relay stations, as is customary in telephone networks and in the Internet. They lack mechanisms for networking (routing; layer 3 in the Open System Integration (OSI) model) and transport security (that is ensuring the reliability and integrity of the information; layer 4 in the OSI model). They are limited to providing point-to-point, point-to-multipoint and broadcast links (layer 2 in the OSI model). In addition, the data throughput of the waveforms is too low to support the applications necessary for network-centric warfare. Solutions are already taking shape. In all probability, a new standard for the architecture of radio-based communications systems will meet the requirements for interoperability not only at the air interface but also inside the radios. The Software Communications Architecture (SCA) developed on behalf of the United States Department of Defense pursues this objective. The Software Defined Radio (SDR) Forum and the Software-based Communication Domain Task Force of the Object Management Group (OMG) support the SCA as a common standard. The SDR Forum includes approximately 120 companies from the radio communications industry, research institutions and government offices. The OMG consists of about 800 members from the software industry and end users.

Fig. 1 Next-generation radios must satisfy requirements for networkcentric warfare.

The future Internet Protocol, version 6 (IPv6) standard will form the basis for networking the systems. To increase data throughput, broadband waveforms will be used.

Interoperability Via Software

Modern, high end radios will have an architecture similar to that of a PC, with highly complex and highly specialized plug-in cards. Since the radio’s central control and monitoring functions will be integrated into its software, functionality can be programmed as needed to the extent allowed by the hardware. Such radios are referred to as software-defined radios (SDR). Versatile software of this type is made usable by implementing abstraction layers and standardized software components. The advantage of such an architecture is that the number of waveforms that the radio can handle is limited only by the internal storage capacity. The user can also modify the functionality of the loaded waveforms when necessary by installing software updates. Of course, the hardware must be powerful enough to handle the waveforms. Performance depends on the possible channel bandwidths, the linearity and power of the amplifier, the possible hop rates, the frequency range covered, the timing or antenna characteristics. Interoperability is achieved by installing waveform software. Today’s SDRs can already do this. However, this does not satisfy the objective of being able to install a waveform on any and every radio. The waveform will function only on instruments of the same product family, because this family has a proprietary software platform. The platform prevents the waveform software from being used on just any instrument. This is similar to the PC world, where Windows applications cannot be run on UNIX machines. The only way to overcome this problem is to create a common standard. This standard must standardize the radio’s central control, define the interfaces between the common software components used and make hardware control uniform. The SCA, which is pursuing this objective, has four main principles. First, the interfaces to the operating system will be standardized. A subset of the POSIX standard, which is derived from UNIX, will be defined. The waveforms must use exclusively this subset for system accesses. At the same time, the operating system must make these interfaces available. Second, CORBA™, which is an objected-oriented communications standard for distributed systems, will generally be used for communicating between software components. Third, the mechanisms for loading and changing waveforms and other applications will be precisely defined. The instance that handles this task is the core framework. Fourth, building blocks will determine the basic modularity for the waveforms and will define the application programming interfaces (API) for some of the software components.

The SCA has limitations in two areas that are currently being further developed to ensure easy portability of waveforms. First, the specifications on how radio-specific hardware is mapped in the software are insufficient. Parts of the radio will remain proprietary unless a uniform hardware abstraction layer (HAL) is defined. The US Department of Defense has recognized this weakness and has expanded the standard in cooperation with Rohde & Schwarz and other interested companies. Second, some interfaces between the software components have not yet been made uniform. Solutions have already been proposed, however, for some areas.

Despite these weaknesses, the SCA represents a giant step forward in standardizing radio architectures. It is currently the only proposal that holds the promise of industry-wide standardization of radios and thus easy portability of waveforms. Interoperability on the air interface, that is the ability to load several waveforms on one radio, is already possible today with an acceptable amount of effort by using multi-band SDRs. For some SDR families of instruments, the migration path to SCA has already been plotted out and is highly recommended.

Networking Via the Internet Protocol

When it comes to networking, there is no getting around the Internet Protocol (IP). This is particularly true when information networking is involved. The Internet is not limited to the transport of data. Solutions such as voice over IP and video streaming make it possible to use the Internet to transmit information previously restricted to telephone and broadcasting networks. Thus, it is only logical to select the Internet Protocol as a standard for networking in radio communications. The new IPv6 version of the protocol offers several characteristics not available in version IPv4 that are relevant to the integration of mobile user equipment (mobile IP, IPSec). As a result, some radio system customers have already adopted IPv6. However, a fair amount of effort is involved in order to transfer Internet functionality to highly mobile networks, that is to networks with a rapidly changing topology that has no infrastructure. Before further steps are taken, solutions must be found for three areas in which radio-based networks differ from the classic Internet. At present, the transmission and connection set-up protocols on which the TCP/IP are based — layers 1 and 2 in the OSI model — are designed for cable connections. Thus, the mechanism for collision avoidance in data transmission, for example, is not suitable for radio links, because additional measures must be taken in order to detect and prevent collisions. This means that data throughput fails when collisions occur. One solution is to use radio-specific protocols. In addition, the mechanisms for setting up routing information, such as for setting up path information for the efficient forwarding of data packets, are designed for networks based on a fixed infrastructure. They are not suitable for highly mobile networks. Solutions are currently being developed in the area known as “ad hoc networking.” The objective is to develop algorithms that can determine the best routes for the data packets to be forwarded without involving extensive communication effort. Third, both the Internet Protocol as well as the transmission of routing information take up a large portion of the bandwidth of a radio link. It is often necessary to route multiple links through one node. As a result, the bandwidth of a single link can drop below an acceptable limit. To ensure optimum use of the bandwidth, corresponding boosters are combined under the term Quality of Service (QoS). Since most solutions are designed for high bandwidth networks, they are only of limited use with mobile networks. Intensive research is also being done in this area. The available bandwidth can be used efficiently by applying the described mechanisms for automatic network management. However, the bandwidth is insufficient for the numerous command support applications that are planned (transmission of videos, synchronization of databases). Only broadband waveforms can offer the required data throughput.

Quality Means High Data Throughput

Waveforms are currently being developed that permit a data throughput of 2 to 5 Mbps over a distance of up to 10 km for ground-to-ground links. Only waveforms with a bandwidth of several megahertz can do this. Thus, the lower frequency bands are not useful for transmission, because they would not provide enough channels having this bandwidth. Yet, higher frequencies have the drawback that the range is lower for the same transmit power. If you want to increase the transmission rate from, for example, 16 kbps to 3.2 Mbps under the same general conditions such as carrier frequency, range or receiver sensitivity, the transmit power must be 200 times greater, because the transmission rate and required transmit power are proportional to each other. Likewise, if the carrier frequency is increased from, say, 200 to 2000 MHz, the transmit power must again be increased by a factor of 50 (in a rural area). These examples provide a very rough idea of the challenges associated with high data rates. However, the actual range is highly dependent on the current environmental conditions, which can easily cause the calculated values to change. For example, when field trials were conducted in a rural area, the actual results from the R&S M3TR tactical radio (see Figure 2) were always twice as good as the values projected by the theoretical models. However, this does alter the main relationship between data rate, transmit frequency, and transmit power and range. The ratio between data rate and required transmit power can be improved through a variety of measures such as using optimum modulation methods, channel coding and directional antennas. Nevertheless, the amount of energy for a mobile user and thus the associated maximum transmit power usually remains limited. As a result, relay stations will be unavoidable for long-distance links at high data rates despite their drawbacks. In contrast, narrow-band waveforms require far fewer relay stations. Consequently, the new broadband waveforms are more likely to complement rather than replace waveforms with narrower bands.

Fig. 2 Software-defined radios of the R&S M3TR family in actual use.

The Future Starts Today

Many of today’s radio systems carry the disadvantages of proprietary waveforms, insufficient networking capability and inadequate throughput for future applications. As the networking of armed forces becomes more common, it will become increasingly necessary to replace these systems with solutions that can handle the future. One approach is to restrict the number of waveforms to just a few that are each optimized to a specific mission condition. The requirements are clear: A waveform is needed that is suitable for the efficient transmission of large amounts of data (high data rate (HDR)). A further waveform should be capable of transmission at a comparably high data rate even if a jammer is present (anti-jamming (AJ)). Yet another waveform should minimize the probability of detection (low probability of detection (LPD)) and interception (low probability of interception (LPI)). Additional waveforms should ensure efficient communication across long distances as well as communication via satellite. If these various waveforms are combined in a single packet, the result is described as a waveform made up of multiple signal shapes (signals in space). However, this interpretation appears useful only if the waveform can be scaled as a whole, such as by adapting it to the mission conditions at hand by setting a few generally valid parameters.

The characteristics of the next generation of radio communications systems are of particular interest for military planning. The current research projects, the capabilities of today’s software-defined radios and the standards that have been adopted suggest that the next generation will fulfill the SCA. The next generation will be able to handle IPv6 via radio, can be networked ad hoc, will offer a broadband waveform with a high data rate and will support an acceptable number of waveforms currently in use. Of course, the next generation will not meet all requirements, but the SCA will provide the radio communications system with an open and flexible architecture. New technologies can easily be added to this architecture, and existing system parts can be reused. This will have a positive impact on logistics, life span, performance and migration effort, ultimately making the system more economical.

Boyd Buchin is involved in software R&D and design in the Radiocommunications Systems Division at Rohde & Schwarz.

Rüdiger Leschhorn is head of technologies and studies in the Radiocommunications Systems Division at Rohde & Schwarz.