Software-Defined Radio: High Performance, Flexible Technology for Spectrum Monitoring
The software-defined radio (SDR) paradigm replaces hardware with flexible software-based components that are inexpensive and easily upgraded to meet current and future needs in spectrum management.
The radio spectrum is used for a wide range of radio services, including mobile phone communications, police and other emergency communications, radar and satellite services, broadcast radio and television. To ensure equitable, economical and efficient use of this scarce resource, government regulatory authorities allocate frequencies for different radio communication services and assign specific frequencies to users. Since it is important for users to operate strictly within their allocated frequency bands, regulators employ various techniques to monitor and record spectrum use.
Spectrum monitoring and recording are critical components of spectrum management. Spectrum monitoring enables spectrum regulators to plan frequency use, avoid incompatible use of frequencies and identify sources of unwanted electromagnetic emissions that may impair the services provided by licensed spectrum users. This involves measuring various performance parameters and analyzing the data to ensure users are complying with regulatory requirements. Measurements monitor spectrum occupancy, occupancy rates and scan frequency bands to identify unknown transmitters. In locations where the RF spectrum is crowded, monitoring and recording systems capture large amounts of data. Spectrum monitoring stations in these areas require high performance monitoring and recording equipment. Such sophisticated, hardware-based instruments are traditionally expensive and designed for single-use applications.
As the number of parameters and services increase, so does the demand for more flexible and higher performance measurements and instruments to make them. A spectrum monitoring and recording system typically consists of an antenna system, a high performance radio receiver, a data storage unit and a signal processing unit (see Figure 1); the signal processing may be incorporated into the radio or data storage unit.
BENEFITS OF SDR SPECTRUM MONITORING
As noted, spectrum monitoring and recording applications demand flexible, high performance solutions. Traditional spectrum monitoring instruments are based on dedicated hardware components, typically expensive and difficult to upgrade for new applications. So they quickly become obsolete.
The SDR paradigm replaces hardware components with flexible and inexpensive software-based components, which are easily upgraded. These software-based components include modulators, demodulators, mixers, filters and amplifiers and since system upgrades require no hardware modification, SDR technology is more suited for rapidly changing spectrum monitoring applications.
Traditional spectrum monitoring systems using only dedicated hardware components greatly limits performance. However, a SDR system can achieve wideband operation, high channel count and wide bandwidths. SDR technology also reduces the cost to produce high performance RF instruments while reducing development time.
Probability of Intercept
Probability of intercept (POI) is a measure of the minimum duration a signal is required to be detectable for an instrument to intercept it with 100 percent probability. It is usually specified in milliseconds, microseconds or nanoseconds. When a monitoring system is described as a high POI instrument, it means the device requires a shorter signal duration for 100 percent POI; a low POI monitoring system requires a longer signal duration. Using a high POI instrument increases the probability of capturing the signal of interest.
Most traditional hardware spectrum monitoring and recording instruments scan for unwanted signals by sweeping from low to high frequencies. This technique yields a low POI and makes these instruments unsuited for capturing short duration signals. Modern SDR instruments are capable of continuously measuring the spectrum for frequencies within their specified spans. This technique yields a higher POI and can capture very short duration signals.
Radio systems are generally categorized by their bandwidth (see Figure 2), i.e., narrowband or wideband. Systems that can tune a large portion of spectrum are referred to as wideband and are commonly used for video streaming, surveillance systems and other applications that need wide bandwidth to carry high data rates. In addition to bandwidth, wideband spectrum monitoring applications require instruments capable of acquiring, storing and processing large volumes of data.
One of the key digital features of a modern SDR spectrum monitoring system is the digital signal processor, which are typically implemented in FPGAs or ASICs. Using reconfigurable devices, such as FPGAs, provide flexibility for upgrades to maintain the longevity of the system, by enabling new features to be deployed in software, rather than requiring new hardware.
Spectrum monitoring and recording instruments capture large amounts of data, especially in locations where the spectral environment is crowded; however, most of the processing is not done in real-time on the radio. The captured data is usually transferred to a storage bank for processing and analysis later. In most spectrum monitoring instruments, transferring captured data to the storage unit uses traditional data buses and protocols, even though traditional standards are highly susceptible to packet losses, meaning some critical data is lost. New and emerging interfaces such as 10G, 40G and 100G Ethernet have high speed with data integrity, making them suitable for high performance spectrum monitoring and recording instruments.
A broad array of analysis methods and techniques are used to characterize signals of interest. The Gabor spectrogram, for example, is commonly used to observe changes in the frequency, amplitude and duration of a signal to identify sources of interference. To employ advanced analysis techniques, such as modulation analysis and the Gabor spectrogram, rich datasets of waveforms acquired over long periods of time are required. Traditional spectrum monitoring solutions typically lack the throughput to transfer large amounts of high bandwidth data to storage, and these limitations make them incapable of continuously storing waveforms acquired over long durations.
For most RF systems, data storage is determined by many factors, including the digital interface between the radio and the storage solution, the network interface controller (NIC), the random access memory (RAM) and the hard drive interfaces and configuration. A system that can store the high bandwidth data acquired by current radio receivers requires a digital backhaul that supports high speed data transfer, linking the radio to the storage. 40G and 100G links are the best choices to meet the high speed data transmission requirements.
The maximum data transfer rate is also determined by the NIC. Unlike traditional network interface controllers, today’s devices are engineered to support high speed data transfer. The rate data is written to storage depends on the RAM. High performance RAM is required, which also affects the reliability of the data transfer.
To store the streamed data, storage based on redundant array of independent disks (RAID) technology or enterprise grade hard drives is required. RAID virtualization technology delivers high storage performance with multiple ways to configure the hard drives: RAID 0, RAID 1, RAID 5 and RAID 10. RAID 0 provides excellent performance, RAID 1 has excellent redundancy and the RAID 10 configuration combines the performance characteristics of RAID 0 and RAID 1. RAID 10 is, therefore, the best choice for achieving performance and redundancy. Of course, the hard drive interfaces must be capable of supporting high speed data transfer.
By combining high performance RAM, a high speed data bus, a high speed NIC and the latest hard drive technology, a high performance stream-to-disk solution capable of storing several terabytes of data can be achieved (see Figure 3). Considering the spectrum monitoring challenges posed by today’s crowded spectral environments, the capability to continuously stream data to a storage solution for a long duration offers significant benefits to spectrum managers and users.
The factors to consider when choosing an SDR solution include frequency of operation, number of independent channels, RF bandwidth, digital backhaul and the availability of pre-configured storage and playback options. The solution must be flexible and adaptable to accommodate new and emerging technologies, including the ability to tune into desired frequency bands and adjust the bandwidth to suit the application. Having multiple channels enables data to be captured at different frequencies or time intervals. Connecting a radio and a recording and playback system for spectrum monitoring can be daunting. To avoid implementation delays and performance issues—dropped packets or configuration challenges—working with a company offering a complete solution can reduce the time to implement a system and minimize risks to the project.