The primary objective targeted the development of a software defined radio prototype platform applied to regenerative satellite communications. Such a platform must be capable of upgrading and reconfiguring itself to adapt to current and emergent signalling standards.
Additional goals include:
- To use established standards to develop 'waveforms' that the SDR platform is capable of implementing. These chosen waveforms were made compliant to the Software Communications Architecture (SCA). The SCA is a set of standards (for hardware and software) designed to ensure portability of complaint waveforms between SCA complaint radios. In principle so long as the SCA radio possessed sufficient resources, any SCA waveform may be deployed on it. In this project the DVB-S waveform was fully implemented and a limited implementation of the DVB-S2 waveform was also developed.
- To use the Software Communications Architecture - Core Framework to:
- To control the overall functionality of the radio,
- To implement the radio waveforms,
- To show scalability of computing resources by connecting multiple processing nodes and controlling them through the SCA.
- To validate the Software Radio Paradigm. Specifically to show that Arrays' prototype SDR solution offers:
- To move away from ASIC solution and perform as much signal process as possible on software reconfigurable DSPs and FPGAs,
- Radio operations that are now controlled though software. Therefore the capabilities of the device are more limited by its software library rather than by its physical layer,
- The ability to upgrade and to adapt to multiple evolving standards. This helps to 'future-proof' the product,
- The ability of the device to reconfigure itself to be able to operate on different networks as needed,
- A reduction of both hardware and software development and deployment costs for communication satellite payloads.
- The above goals are subject to the underlying requirement that the prototype platform was composed of generic Commercial Off-The-Shelf (COTS) components.
The SDRRCS prototype is designed to offer a reconfigurable and upgradeable wireless communication system. The goal is to perform as much signal processing as is possible in reprogrammable devices such as FPGAs or GPPs. Figure 1 shows the general layout of the receiver and which hardware components perform what functionality.
Figure 1. Signal Processing Blocks Required at the Receiver, as well as Types of Devices Which Perform the Processing
In developing the architecture, the goal was to develop a standardized OnBoard Regenerative Processor (OBRP) for the deployment of the DVB-S and DVB-S2 waveforms using COTS elements. It was also deemed desirable to perform as much of the signal processing as possible on general purpose processors, while still retaining the use of the Application Specific Integrated Circuits (ASICs) where bandwidth requirements dictated. The prototype is designed using Commercial Off The Shelf (COTS) components. Essentially four COTS components are used in its construction:
- Digital receiver board
- General purpose PC
- Linux operating system
- Software Communications Architecture
COTS components enable small-to-medium (SME) companies to rapidly prototype generic SDR hardware platforms as well as perform the system integration. Designs become modular and can easily be upgraded. For example, as individual components are improved upon such as faster GPPs, then these components can easily replace the existing processors. The net result is a system that now offers greater functionality than before.
The project plan is to develop an integrated hardware and software SDR solution that can be used as a regenerative transponder for satellite communications. The prototype transponder will show compatibility for the SCA-CF, which is a set of operating standards dictating how SDR should internally function. The prototype will be constructed using COTS components, both hardware and software that are (or can be made to be) SCA compliant. This will then guarantee that they will function properly when incorporated into the prototypes' architecture.
Since the SDR paradigm (including the use of COTS for product development) is still new and unproven. There are many issues and concerns that must be overcome in order for it to achieve wide scale acceptance and success.
For the SDRRCS development project the two most important issues are:
- A lack of system reconfigurability and upgradeability, arising from the non-standardization of COTS equipment and unforeseen design complications. This concern can be reduced by the use of the SCA standards to ensure components are SCA compliant and therefore standardized.
- Real-time issues primarily from bottlenecks arising from inadequate processing power from the processing components (CPU, etc). This is important since multimedia applications for satellite require a large spectral bandwidth, and thus a high degree of on-board processing power is needed.
An SDR, SCA complaint radio was built using COTS equipment. In addition SCA complaint versions of the DVB-S and DVB-S2 waveforms were also built. The radio successfully demonstrated the required traits of upgradeability and reconfigurability by uploading, and correctly deploying variants of the DVB-S and DVB-S2 waveforms.
However, when the waveforms were started and their throughput evaluated, the performance was below estimated values (approximately half target values).
Initial targets were for 1Mbps for the DVB-S2 waveform. It was apriori estimated that the bottlenecks in performance would be the matched filtering for estimating the phase error and the iterative process of decoding required by the LDPC codes. Benchmarking of the two individual processes showed that both were capable of exceeding a target rate of 1 Mbps; however, the PC (represented under the SCA as a single node with one executable device - the GPP) was overwhelmed when simultaneously running all signal processing components of the waveform.
Analysis revealed the aggregate real-time performance of the radio platform was based on four components:
- PC capability (primarily determined by clock speed of the processor).
- Efficiency of drivers provided for the radio card.
- Complexity of signal processing requirements for the waveform to be run.
- The SCA implementation and its interaction with the CORBA (facilitates communications between components) distribution.
The most significant limiting factor was the performance of the PC host platform itself. The PC was responsible for phase and all baseband waveform processing. Additionally, the PC must provide resources for the radio management (the SCA, the Linux OS, and control of the Pentek card). These results indicate that although COTS components and in particular PC processors have rapidly developed, dedicated signal processing is still best left to FGPAs and DSPs.
The main benefit offered by SDR technology (and is validated by this project) is that the prototype, constructed using COTS components was made reconfigurable and upgradeable to different waveforms. These two abilities allow SDR devices to ?future-proof? themselves against changing standards, something conventional satellite transponders systems cannot do.
- Only need to upgrade individual components, not the whole device.
- Use of SCA-CF provides interoperability between radios and helps to force standardization.
- Easy to develop new (unexpected/unforeseen) applications for the device since they are written in software and then uploaded to the system as needed.
- Generally cheaper to develop with COTS components than to develop the radio from scratch.
- COTS components offer a rapid prototyping solution and thus provide a quicker time-to-market solution.
Perspective on Future SDR Development
Developing SDR technology is clearly an iterative ongoing process, but initial steps such as this development of a prototype OBRP system is a necessary first step. It should be emphasized that SDR is more of a concept than a physical entity such as a piece of software or hardware. The benefit of SDR is the gestalt benefit that is realized from the confluence of these components. There is currently no universally accepted definition of precisely what is an SDR device; only that after a device possesses enough SDR traits (upgradeability, reconfigurability, etc.) may it be considered as an SDR device.
The individual components comprising SDR devices are rapidly maturing - if only because they had so far to go. These are the FPGAs, the software development tools, the reference architectures for the SCA, etc. Nevertheless, complete SDR devices are still in the technology development / demonstration stage (January 2007).
SDR must address three primary areas of risk which focus around the SDR core technology itself and its relation to the current wireless value chain.
- Technological: SDR technology is still unproven and therefore suffers from credibility issues. Additionally to design the interfacing between individual components still requires a large effort on the part of the developer.
- Information Security: Large organizations are only too aware of the dangers associated with hackers and network security. Since SDR technology is essentially predicated on the fact that software is used to control almost all aspects of the communications device, security concerns become a significant obstacle, impeding widespread adoption of SDR technology.
- Regulatory and Standardization issues: SDR technology has/will redefine the communications value chain. As such, many of the previously successful incumbent players will stand to lose much of their dominance as regulatory issues change. For example, the spectrum allocated to some inefficient television channels has been proposed for re-use by spectrum efficient WiMAX signals. This is also true of the desire to replace Standard Definition Television (SDTV) on the VHF bands with High Definition Television (HDTV) on the UHF bands. Some television service providers have understandably enacted legal challenges for these kinds of encroachments on their licensed spectrum. Note that while the Federal Communications Commission (FCC) has endorsed the research and use of SDR technologies, it may still be some time before all the legal issues surrounding SDR are resolved.
Last Update: 15 Sep 2008