Other DoD SDR Programs

2.7.1 Joint Combat Information Terminal (JCIT) [10]

There is a long history of how the Naval Research Laboratory’s (NRL) Navy Center for Space Technology got into the business of developing communications equipment. NRL placed transponders in the multi-spacecraft dispenser, the upper stage vehicle that inserted payloads into orbit, so as to provide near real-time reports. From this effort grew the tactical receive equipment (TRE) and eventually a flight box, the multi-mission advanced tactical terminal (MATT), capable of simultaneously receiving and processing intelligence reports for tactical receive applications (TRAP), tactical data information exchange system broad-cast (TADIXS-B), and tactical information broadbroad-cast service (TIBS). A number of MATT-like devices had been made for Special Forces. The Army had a requirement to build an Army aviation command and control system (A2C2S), which required dozens of discrete radios to meet worst-case scenarios. The development of a multi-channel system to adapt to meet various communications needs led NRL, leveraging their efforts under MATT, the improved data modem (IDM passes digital data for targeting or situation awareness) and the

Commander’s situation awareness workstation (CWAS), to initiate development of the enhanced communications interface terminal (ECIT) in 1993. ECIT had five multiband receivers and covered HF, VHF, and UHF. It employed a number of programmable modules and its transceivers supported multiple waveforms, ECCM, and FEC techniques. ECIT (used on the UH60C helicopter) evolved into an eight-channel system that was capable of covering 2–512 MHz, and expandable to 2.5 GHz, and is now known as the joint combat information terminal (JCIT).

The JCIT configuration is shown in Figure 2.17 and its architecture in Figure 2.18.

JCIT was configured using SEM-E modules and an IEEE 1394 backplane. JCIT has an open system hardware and software architecture developed under NRL and is owned by the government. JCIT implements many modulation formats (AM, FM, SSB, BPSK, SBPSK, QPSK, SQPSK, SOQPSK, MSK, CPFSK, DSSS, FH, GPS, CDMA, TDMA), and many protocols (ATM, TRAP, TADIXS-B, AFAPD, TACFIRE, TIBS, JTIDS, TRIXS, TADIL-A/B/J). The system is software reprogrammable to support a variety of mission scenarios and is portable operating system interface (POSIX) compliant, uses VxWorks, and its software is coded using both Ada and C. The JTRS program has leveraged much of the software structure used in the JCIT effort.

Figure 2.17 JCIT configuration

2.7.2 Global Mobile (GloMo) [11]

DARPA’s GloMo program had the vision to ‘‘Develop and integrate technologies and tech-niques at applications, networking and wireless link levels that will enable wireless users to access and utilize the full range of services available in the Defense Information Infrastruc-ture.’’ The GloMo program addressed mobile application support techniques, mobile networking, and wireless nodes and was comprised of more than two-dozen efforts by industry and academia. Only a few are summarized below.

2.7.2.1 Wireless Internet Gateways for Internets (WINGs) and Secure Protocols for Adaptive, Robust, Reliable, and Opportunistic WINGs (SPARROW)

The University of California, Santa Cruz, developed and analyzed packet-radio networking protocols and architectures that would provide guaranteed service for multimedia traffic and to provide a seamless extension of the Internet. They used and improved the C11 protocol toolkit (CPT). Packet radio prototypes were used to test and demonstrate results using modular, high-speed, low-cost, commercial, spread spectrum radio hardware. Under SPAR-ROW, secure protocols were designed for exchanging control information to protect nodes against compromise. The WINGs protocols were leveraged by a number of other DoD programs.

Figure 2.18 JCIT architecture

2.7.2.2. Advanced Signal Processing & Networking (ASPEN)

The prime objective of the ASPEN project was to enhance packet radio networking protocols using digital programmable radio technology in order to demonstrate quality of service networking. ASPEN utilized the WINGS protocol and focused on enhanced throughput, a wideband waveform with Doppler tracking, and variable processing gain.

The program’s objectives were to:

† demonstrate an advanced wireless network, link protocols, programmable multi-channel wideband radios, including enhanced throughput, utilizing GloMo radio APIs and a wide-band waveform (with Doppler tracking and variable processing gain);

† investigate and demonstrate adaptive signal processing techniques;

† develop tactical Internet traffic models and demonstrate a wireless network using ASPEN radios.

2.7.3 Ultracomm

Ultracomm was a DARPA effort that was targeted toward reducing the footprint of a SDR by employing size reducing technologies such as: silicon carbide amplifiers and microelectro-mechanical-system (MEMS) devices and filters.

Raytheon E-Systems of Falls Church, Virginia was awarded a contract to develop a multi-band, multimode communications system based on a family of modular components, pack-aged in the widely used PC card format (PCMCIA). One Ultracomm system provided four independent communication channels, with each channel capable of covering the RF commu-nication bands between 20 MHz and 2.8 GHz. The system had a wide 60 MHz instantaneous bandwidth, was low power, and was completely software reconfigurable throughout the spectrum. It accomplished this via MEMS filtering, which provided highly integrated prese-lection across the entire spectrum on a single silicon chip. Silicon germanium ASICs were used to produce a high performance RF analog signal processor that could be closely inte-grated with the MEMS filter devices. State-of-the-art analog-to-digital converters were used to achieve an impressive instantaneous bandwidth and allow much of the signal processing to be done using inexpensive digital components. Silicon carbide power microwave devices were used to develop a high power density amplifier for the transmitter. Power MEMS devices provided for efficient power transfer by exactly matching impedance across the entire spectrum. All of the technologies were integrated together using advanced packaging tech-niques, high-density multi-chip modules, and optical interconnects.

2.7.4 Wideband Radio Networking (WRN) Program³

As indicated earlier, the US Army Communications & Electronics Command (CECOM) at Fort Monmouth, New Jersey, was an original member of the Joint SPEAKeasy Multiband Multimode Radio (MBMMR) Program, along with the US Air Force and DARPA. A SPEAKeasy objective of major interest to the Army was the program goal to evolve and

3This section was authored and provided by Mr Donald W. Upmal, WRN Project Leader, US Army Commu-nications & Electronics Command, Fort Monmouth, New Jersey.

enhance the engineering prototype radios over the 4-year program period to support not only narrowband low data rate waveforms, but also to support wideband high data rate waveforms.

The Army’s special interest in this capability was driven by their extensive modernization plans and future vision to ‘digitize the battlefield’ of the twenty-first century. Lessons learned during the Task Force XXI tactical field exercises in March 1997 further emphasized the Army’s urgent need to develop new higher data rate waveforms to supplement the low data capacity of today’s legacy radios, such as the SINCGARS, EPLRS, and NTDR radios. This same need was later reflected in the Operational Requirements Document for the JTRS, specifying the requirement to develop a new wideband networking waveform for the future JTRS family of radios. When the SPEAKeasy program was restructured in 1997, due to the lack of required additional funding after TF-XXI-AWE, all the development tasks related to wideband capability were eliminated. Subsequently, the Army decided to shift its FY98–99 funds to an effort that would better support their need to develop high data rate wideband waveforms. Thus, the Army WRN program was born.

In November 1997, CECOM issued a broad agency announcement (BAA) entitled Wide-band Radio Networking, with the program objective to provide an enhanced experimental capability within the CECOM Research, Development & Engineering Center (RDEC) to facilitate the development and evaluation of new wideband radio network waveforms and technology. The BAA solicited the development and delivery of four products:

† a modular software based programmable wideband network radio (WNR);

† a computer automated wideband radio network testbed (WRNT);

† a wideband WNR test waveform; and

† a software development environment for the WNR.

The WNR was to be used to host new candidate wideband waveforms and, along with the WRNT, provide the CECOM RDEC with an enhanced network radio testbed capability to develop and experiment with third party/industry network waveforms, adaptive protocols, and radios, as well as provide a testbed for candidate JTRS waveforms and DARPA program networking technology. The WRN contract was awarded in April 1998 to prime contractor Raytheon Company, Fort Wayne, Indiana, requiring the delivery of three WNR radios.

The contract specified that the WNR architecture be mainly software based and repro-grammable so that performance and functional improvements could be upgraded and controlled through software only, with minimal hardware changes. With the intent to leverage as much currently developed software radio technology as possible, the contract also required the WNR architecture conform to the latest PMCS reference model to the greatest extent possible, and that the ‘radio APIs’ developed under the DARPA GloMo program be used during software development. To reduce cost, it was also required that the design be COTS/

NDI based to the maximum extent feasible. The final Raytheon design produced the radio shown in Figure 2.19, having an open hardware architecture using a standard COTS 19‘‘ rack chassis and dual 3U compact PCI (cPCI) card cages, with standard cPCI data busses, control processor (CPU) cards, and Ethernet I/O cards.

As shown in Figures. 2.20 and 2.21, the open modular architecture conformed exactly to the top level PMCS reference model, including separate Red and Black data busses, allowing implementation of secure INFOSEC features and operation during future upgrades. Due to contract schedule constraints, the programmable INFOSEC module design was not completed prior to WNR delivery, therefore a standard Ethernet interface

was used as an interim gateway between the Red and Black data busses. As a result of the flexible hardware and software design and numerous spare card slots, the WNR design could support multi-channel operation (as illustrated in Figures. 2.20 and 2.21), but the WRN contract only required the single ‘Raytheon wideband modem/RF’ channel be implemented and delivered.

Figure 2.20 WNR block diagram Figure 2.19 Army wideband network radio (WNR)

The WNR could be controlled by any standard PC HMI running the host network manager (HNM) software and connected to the COTS ‘Ethernet I/O interface’ card plugged into the red side cPCI bus. The HNM software provided simple GUI screens and menus, and allowed easy user control of all radio system and waveform functions and features.

The majority of the radio system control, bus control, networking, and high-level wave-form functions were perwave-formed in (and divided between) the Red and Black control proces-sors. The two COTS PEP modular computers CP-310 processor cards were identical and include a 166 MHz Pentium processor, 64 Mbytes DRAM, 80 Mbytes flash ROM, full complement of serial and parallel ports, and a cPCI bus interface. All system and waveform executable software, which resides on the control processors, were downloaded and/or repro-grammed directly from the HMI PC via the Ethernet I/O and cPCI bus interfaces.

The wideband (WB) modem card and RF front end circuit card were two custom Raytheon designs, and were sandwiched together to form a double-width cPCI 3U module. Again, to leverage current technology and reduce cost, the WB modem/RF module design was a previous Raytheon development under the DARPA ASPEN radio program and re-used on the WRN program.

The programmable elements of the WB modem included a Texas Instruments TMS320C548 DSP and two dense Xilinx XC4085XL FPGAs. A Qualcomm Q1900 Viterbi decoder and DA/AD converters were also located on the card. This combination of base-band processor elements performed all the data modulator/demodulator functions necessary to process a wide variety of digitally modulated waveforms. All modem and waveform related executable software, which resided on the DSP or FPGAs, was downloaded and/or

Figure 2.21 WNR open modular architecture

reprogrammed via a JTAG connector located on the edge of the WB modem card by simply lifting the WNR top cover.

The RF front end card provided all the required receiver/exciter functions in two selectable frequency bands; 225–500 MHz (lowband) and 500–1000 MHz (midband). The maximum RF output was nominally 23 dBm, with 20 dB of output power control (attenuation) in 1 dB increments. An internal R/T switch provided a single RF interface to an external antenna port.

When needed, RF power amplification was provided by an external power amplifier. A highband 1.3–2 GHz RF module was also developed under the WRN contract, but was not fully implemented in the delivered radios.

At the time of the WRN contract award, Raytheon had already been an active member of the SDR Forum and was preparing their technical proposal for JTRS. Raytheon used a consistent software approach for all three efforts, the WNR architecture design and imple-mentation was a stepping stone (with lessons learned) toward the development of the JTRS architecture. The WNR software architecture therefore was an early version of today’s JTRS architecture and its basic structure is shown in Figure 2.22.

From top to bottom, this layered architecture consisted: of an application layer, where the major radio/waveform functional software resided; a core framework layer, which provided an essential set of open application-layer interfaces and services; a common object request broker architecture (CORBA) layer, which was standard commercially available ‘middle-ware’ facilitating message passing and a distributed processing environment; a POSIX OS

Figure 2.22 WNR open software architecture

layer, where commercial VxWorks OS software resided; a network & serial I/O services layer, where commercial components provided these functions and protocols; and a board/

bus layer, where the standard cPCI software resided for the WNR.

The WNR software uses a distributed control structure in support of the object-oriented design. Therefore, as mentioned earlier, computer software configuration items (CSCIs) were distributed across several processors within the WNR system. This distribution is shown in Figure 2.23, with the four major WNR CSCIs being the core CSCI, INFOSEC CSCI, HNM CSCI, and ASPEN waveform CSCI. The fifth waveform-A CSCI shown in the figure was for an additional WNR waveform, but was not completely implemented in the delivered WNRs.

The WNR software architecture provided isolation and standard interfaces between the underlying software and hardware layers and the upper level application layer software, assuring enhanced portability, reconfigurability, and interoperability of software and hard-ware components developed using this architecture. It also provided a building-block struc-ture for defining APIs between application software components.

The ASPEN waveform implemented on the WNR was also a re-used product from Raytheon’s previous DARPA ASPEN program and had a two-fold purpose:

† provide a test waveform robust enough to sufficiently stress the WNR to assure its final design could host future experimental wideband waveforms and associated adaptive networking features;

† be evaluated as a candidate JTRS network waveform.

The ASPEN physical layer modem/signal-in-space was a quasi band limited (QBL) DSSS waveform, using DPSK data modulation, convolutional FEC coding, and block interleaving.

It used a 20 Mcps chip rate, with individually selectable spreading ratios and packet lengths.

Figure 2.23 WNR software distribution

During WNR acceptance testing the ASPEN waveform, with its very robust RF character-istics and adaptive networking features, was found more than adequate as a test waveform. At the time of this writing, the WNRs were being used by US Army CECOM to implement and test additional wideband waveforms as possible candidates to meet the Army’s future JTRS operational requirements.