SDR provides a qualitative leap in frequency agility and protocol standard independence, and solves the problem of system incompatibility in a highly fragmented communication environment.|By Ron Hickling, Oleg Panfilov, Tony Turgeon, and Michael Yagi

Cognitive radio systems in general, and TSR products in particular, soon may revolutionize telecommunications. By providing all signal processing in software, TSR systems asymptotically approach the ultimate SDR systems goal of seamless system operation in highly fragmented, multi-terminal, and multi-frequency communications environments.

What Makes a TSR?
These advancements are achieved when the received signal is converted into digital form at the antenna input and software is downloaded dynamically to correspond with the specifics of the received signal. Progress toward the SDR's ultimate goal of seamless operation for TSR has to cope with rapid evolution that includes multiple standards driven by the rate of advancement in microelectronics, particularly the increase in microprocessor computational power and reduction in power dissipation. The SDR Forum stipulates that real SDR products must have two fundamental features: flexibility toward operational standards and independence from carrier frequencies.1 TSR systems, by definition, have both characteristics.

Glossary of acronyms
A/D — Analog to Digital
ADC — Analog-to-Digital Converter
CDMA — Code-Division Multiple Access
CISC — Complex Instruction Set Computer
CMOS — Complementary Metal-Oxide Semiconductor
D/A — Digital to Analog
DDCR — Direct Down-Conversion Receiver
DSP — Digital Signal Processing
DSPP — Digital Signal Processing Primitives
EDGE — Enhanced Data GSM Environment
FPGA — Field Programmable Gate Array
GIPS — Billion Instructions Per Second
GPRS — General Packet Radio Service
GPS — Global Positioning System
HDTV — High Definition Television
IEEE — Institute of Electrical and Electronics Engineers
IEICE — Institute of Electronics, Information and Communication Engineers
I/O — Input/Output
I/Q — In-phase and Quadrature
LAN — Local Area Network
LVDS — Low Voltage Differential Signaling
MESFET — Metal Semiconductor Field-Effect Transistor
MIPS — Million Instructions Per Second
PECL — Pseudo Emitter Coupled Logic
PLD — Programmable Logic Device
RF — Radio Frequency
RISC — Reduced Instruction Set Computer
SAR — Software Assisted Radio
SDR — Software Definable Radio
SINAD — The Ratio of Signal-Plus-Noise-Plus-Dstortion to Noise-Plus-Distortion
SNR — Signal-to-Noise Ratio
SR — Software Radio
TSR — True Software Radio
WCDMA — Wideband CDMA
Illustrtaion by Roy Scott
But SAR cannot be reconfigured for both carrier frequency and communications protocol, because only a portion SAR signal processing is done in software or otherwise digitally. SAR fails to provide both features, which are demanded by real SDR systems.

SAR systems, shown in Figure 1, may be viewed as an intermediate step on the evolutionary path toward TSR. The path to that goal includes direct down conversion from RF to baseband immediately following the antenna input (see Figure 2). As with any ultimate goal, SAR can be approached asymptotically while each generation of new semiconductor processes brings the elusive SDR goal of multi-terminal and multi-frequency operation successively closer.

Depending on where the signal processing starts in their microprocessors, SDRs are made differently for good historical reasons.2, 3 In a TSR, the ADCs are configured as close as possible to the antenna, which places great demand on their performance. Signal processing of the digitized antenna output is done by fast logic circuits and fast microprocessors using downloadable signal processing software selected according to a system operational environment.

Effective quantization of the radio signal at the antenna enables the communications terminal's air interface parameters to be reconfigured quickly. Dynamically switching frequencies and communications protocols in the user's terminals add or remove system software components to enable remote reconfiguration of terminals and achieve greater flexibility; thus, TSR is the only technology that currently shows promise in delivering the ultimate SDR goal of a truly "universal" radio terminal. Other approaches are vulnerable to changes in applied standards and to the introduction of new functionality requirements.

SDR System Requirements
Protocol-specific system parameters define the requirements imposed upon SDR systems. Initially, wireless communications systems are expected to benefit the most from SDR and thus are expected to show the greatest demand as reflected by shipping volume. What, then, are the SDR implementation requirements for these systems?

click the image to enlarge

Figure 1. A conventional analog receiver typically uses a double-conversion design. The architecture requires multiple external analog components (shown as red boxes), contains many analog interfaces (shown as red lines), and can only decode one type of waveform.

click the image to enlarge

Figure 2. The generic architecture of the TSR system.

Processing requirements for different communications protocols depend on factors including the application (voice, video, and multimedia, for example, require different bandwidths) and the way in which signal processing algorithms are implemented. Table 1 estimates the resource demand in processing different protocols.4, 5

click the image to enlarge

Figure 3. A plot of the effective resolution of a delta-sigma converter for the first-, second-, third-, and fourth-order modulators using one-bit quantization. The x-axis represents the ratio of clock rate to maximum input bandwidth (over sampling ratio). The y-axis represents resolution in bits.

click the image to enlarge

Figure 4. Performance of new A/D converters: CMOS technology converter at a clock rate of 10 MHz.

Estimated processing power may vary widely among different communications protocols. Table 1 shows that current generation GSM phones require about 10 MIPS. GPRS systems require processing power on the order of 100 MIPS. An additional order of magnitude is required for EDGE systems. WCDMA systems with a signal bandwidth of 5 MHz push power requirements an additional order of magnitude to 10 GIPS. The new generation of orthogonal frequency division multiplexing LANs is in the same league of processing requirements: 5 GIPS. From a processing standpoint, the challenge in software radio is to exploit the three basic processor types — fixed architecture processors, FPGAs, and programmable DSPs/RISCs/CISCs — to optimize the three-way trade-offs among speed, power dissipation, and programmability. Regarding programmability, the issues of high-level language interfaces, portability, and reprogramming speed must be considered.

click the image to enlarge

Table 1
The Technology
SDR technology is eventually expected to offer complete programmability and reconfigurability to both multi mode and multi-functional communications terminals and network nodes. A current lack of sufficient processing power prevents SDR from becoming a full-scale reality. However, processing power continues to increase at an unabated rate — doubling every 18 months — which assures that SDR-based multi mode and multi-functional communications terminals and network nodes soon will come to life.

click the image to enlarge

Table 2
For now, the goal is to enable TSR systems that can implement a couple of protocols per unit installed where a steady supply of tens of watts of power supports the required signal and data processing. The power requirement limits the range of current possible TSR implementations to basestations and motor vehicles, aircraft, boats, and trains. Next generation TSR components will implement DSPP directly in the silicon to make lower power consumer devices feasible. An SDR suitable for commercial narrowband and broadband applications will typically cover the frequency spectrum between 400 MHz and 6 GHz. This range embraces most of the existing and emerging standards and the most likely future developments.

click the image to enlarge

Table 3
Basic ingredients of TSR hardware design include mixers and ADCs, reconfigurable hardware such as FPGAs and PLDs, DSP boards, and general-purpose computers.6 Embedded software can reside in all of the programmable entities used in the design. Any TSR unit design must address several issues, including:
• A transceiver partition between hardware and programmable hardware entities
• Deciding which type of programmable hardware should be used
• The ability of the designed architecture to adapt to evolving communication protocols
• Interfacing the various entities used in the design of the SR unit for real-time operation of the platform

These issues involve optimizing the design to achieve the TSR platform's objectives. They have a direct effect on system performance because they select:
• How much RF bandwidth the TSR platform can process
• The degree of programming flexibility in the design
• How much time it would take to reprogram the hardware
• The choice of hardware architecture given that different components must be able to operate in real-time
Following the direct conversion front-end, a general-purpose workstation platform is preferable because it is by far the most flexible and cost-efficient programmable hardware.7

click the image to enlarge

Figure 5. Performance of GaAs MESFET A/D converters at a clock rate of 1.75 GHz with 50 dB dynamic range.

click the image to enlarge

Figure 6. A software radio basestation used as a relay/translator for a single-protocol handset.

TSR Architecture
The architecture of TSR must be able to accommodate operation in different environments characterized by different standards, carrier frequencies, power levels, and bandwidths. Architecturally, TSR is best defined as "the software implementation of the radio transceiver receiving digitized, down-converted signals from an antenna."

click the image to enlarge

Table 4
Patented delta sigma loop circuitry makes the direct down-conversion from RF to its baseband equivalent. Digitization of the wireless signals' functionally at the antenna, in TSR systems, dramatically simplifies the implementation of transmitters and receivers. It uses DDCRs instead of traditional superheterodyne receivers. The result is low cost and simplicity, especially for broadband receivers typical for CDMA and WCDMA cellular systems. It can be implemented in a single integrated circuit instead of the bulky discrete component filters required for superheterodyne receivers. The architecture requires few external analog components and can be programmed to process any type of signal or multiple types of signals.

Delta-sigma Converters
Delta-sigma converters digitize signals by modulating the analog input into a high-speed, one-bit digital data stream that is subsequently processed digitally to produce a high-resolution word stream at a slower data rate. The converter is a closed-loop system in which the order of the loop and the input bandwidth may be traded for resolution. Figure 3 plots the ideal resolution for a given relative bandwidth and loop order.

To push the envelope, new delta-sigma converters capable of operating at clock rates in excess of 5 GHz have recently become available. Additionally, this converter architecture provides the ability to simultaneously extract the modulation from an incoming wireless signal and digitizes it with extremely high resolution. This particular converter is expected to realize a dynamic range of 55 to 100 dB (depending on bandwidth) using this architecture. The performance analysis of high-order delta-sigma ADC converters operating at 5 GHz shows that expected SNR depends on the operational spectrum width DF and the order of delta converters. The estimated values of SNR in decibels for different converter orders and for operational bandwidths, measured in megahertz, are shown in the Table 2. The evolution of A/D and D/A converter parameters is shown in Tables 3 and 4. Figures 4 and 5 show 100 dB dynamic range using CMOS technology at a clock rate of 10 MHz with roughly 50 dB dynamic range in a preliminary system using GaAs technology at a clock rate of 1.8 GHz.

Evolution of Applications
Increasingly powerful processing capabilities will allow TSR products to evolve to serve the wide range of new applications that software radio makes possible. For example, Figure 6 shows a single protocol cellular phone connected to a variety of networks through a software radio basestation that serves both as a repeater and (when necessary) as a protocol translator. This application does not require software radio handsets for universal nationwide access.

Figure 7 shows a similar concept for automobiles. The car's "dual mode" transceiver transmits and receives using a proprietary protocol and also receives GPS navigation signals. The car can transmit position information to the basestation and has universal access similar to what is available to the handset shown in Figure 6. Figure 8 shows a car with software radio technology. The automobile can access any service and has the potential for international roaming. Furthermore, specialized services and capabilities (such as automobiles serving as repeaters for other automobiles and/or handsets) can be implemented without disrupting access to "standard" services.

click the image to enlarge

Figure 7. An SDR basestation used as a relay/translator for a vehicle with single-protocol capability.

click the image to enlarge

Figure 8. A software radio communicator used to enable an automobile to access any wireless network.

In this configuration (where both the basestation networks and the access device use software radio technology), any wireless access device can communicate with any other wireless access device with multiple choices of network services. In such an environment, the user can utilize the service best able to meet his needs whether the priority is the bandwidth, cost, or latency. Such flexibility would create a competitive cognitive radio environment where limited bandwidth use is efficiently allocated so the user pays only for the bandwidth he needs, and no usable bandwidth remains idle.

click the image to enlarge

Figure 9. Dynamic distribution of signals in licensed and unlicensed frequency domains.
In bandwidth-intensive broadband applications, next generation software radio technology allows service providers to dynamically use unlicensed frequency bands to meet additional client needs when saturation has been reached on licensed frequencies. For example, Figure 9 shows how residential customers could receive video feeds on unlicensed frequency bands while business customer locations concurrently receive video feeds on licensed frequencies.

Ideal SDR products must have two fundamental features: flexibility toward operational standards and independence from carrier frequencies of received signals. Two quite different lines of SDR products are emerging.

The first, SAR, is a hybrid, where a portion of signal processing (usually at high or intermediate frequencies) is done in hardware, and the rest in software. Such systems can satisfy only one criterion: partial independence from standards. Meanwhile, they depend on analog components for filtering and other signal processing, which makes them reconfigurable but not frequency agile.

The second, TSR systems, use direct down-conversion from RF to baseband immediately after an antenna. They enable the two fundamental features. Such systems correspond to the spirit and expectations put on SDR.

As with any ultimate goal, TSR can be approached in an evolutionary manner, with each iteration of new processes in microelectronics coming closer to multiterminal and multifrequency operation. The rate of introduction of these new processes, especially those enhancing microprocessor performance, defines the rate of TSR evolution.

1. SDR Forum web site
2. IEICE Transactions on Communication, special issue on software defined radio and its technologies, June 2000.
3. Mitola, "The Software Radio Architecture," IEEE Communications, Vol. 33, No. 5, February 1995, pp. 26–38.
4. Helmschmidt, Jurgen; Eberhardt Schuler, et. al., "Reconfigurable Signal Processing in Wireless Terminals," IEEE Proceedings of the conference Design Automation and Test in Europe, DATE-03.
5. Kokozinski, Rainer; et. al., "The Evolution of Hardware Platforms for Mobile Software Definable Radio Terminals," IEEE PIMRC2002.
6. Laddomada, Massimilliano; Fred Daneshgaran; Marina Mondin; Ronald Hickling, "A PC-Based Software Receiver Using a Novel Front-End Technology," IEEE Communications, August 2001.
7. Bonner et. al., "A Software Radio Platform for New Generations of Wireless Communications Systems," CNIT 12th Tyrrhenian Int'l. Wksp. Digital Commun., Italy, Sept. 2000.


About the Authors
Ron Hickling, Oleg Panfilov, Tony Turgeon, and Michael Yagi are employed by TechnoConcepts Inc., 15531 Cabrito Road, Van Nuys, CA 91406, USA.
E-mail: ronh, panfilov, tony,