As a new technology in the field of wireless communication, software radio (SDR) is attracting more and more attention both inside and outside. In the field of communication, it is a new radio communication system after analog technology to digital technology, fixed communication to mobile communication. With the development of communication technology, equipment compatible with various standard types has increasingly shown its demand. Compared with traditional radio systems, software radio systems have a series of advantages such as general structure, software-based functions, and interoperability. .
Ⅰ. The origin and development of software radio
The reason for the emergence of software radio is related to the Gulf War. At that time, the multinational forces led by the United States used a variety of communication equipment of different standards, which caused difficulties in communicating with each other. After that, in May 1992, Jeo Mitola first proposed the concept of "software radio" at the American Communication Systems Conference. The basic idea is to make all tactical radios based on the same hardware platform, install different software to form different types of radios, and complete functions of different natures. So it has software programmability. This concept quickly attracted the attention of countries around the world, because military communications now have higher requirements for the reliability, interoperability, flexibility, anti-jamming, survivability, confidentiality, and security of radio communication systems. The U.S. military and Hazcltine have developed a software radio station called "speakeasy" (easy to talk), which realizes a multi-band and multi-functional radio platform commonly used by the U.S. military. More than 4 different modulation waveforms. This radio can be called a "handheld computer" with an antenna that can transmit voice and data. Communication services include voice, data, and video images.
At present, software radio has been paid more and more attention in the civilian field. The main reason is that the technical standards of the current communication system are diverse, and the various technical standards and the corresponding systems are difficult to be compatible with each other, and it is difficult to realize it with a unified device. And the third generation mobile communication system still has a standard battle, if software radio is used to adapt to different standards, it is a feasible way. On the other hand, the development of communication technology is very rapid, the old system is continuously improved, and the new system is emerging rapidly. People need a system upgrade method that is more economical than completely eliminating the old equipment, and the programmability of the software radio is better. adapted to this need.
Ⅱ. the software radio architecture
The radio frequency part, up/down conversion, filtering and baseband processing of the traditional analog radio system all adopt the analog mode, and the communication system of a certain frequency band and a certain modulation mode corresponds to a special hard structure; while the low frequency part of the digital radio system adopts digital circuits ( For example, the local oscillator uses a digital frequency synthesizer, source coding and decoding, and modulation and demodulation are completed by a dedicated chip), but its radio frequency and intermediate frequency parts are still inseparable from analog circuits. Compared with the traditional radio system, the A/D/A conversion of the software radio system is moved to the intermediate frequency, and as close as possible to the radio frequency end, the whole system frequency band is sampled, that is, the digital processing is performed from the intermediate frequency (or even the radio frequency), which is A prominent feature of software radio. Digital radio uses dedicated digital circuits to achieve a single communication function without programmability. The software radio replaces the dedicated digital circuit with the programmable DSP device, which makes the system hardware structure and function relatively independent. In this way, based on a relatively common hardware platform, different communication functions can be realized through software, and the operating frequency, system bandwidth, modulation mode, source code, etc. can be programmed and controlled, and the system flexibility is greatly enhanced.
The hardware platform of the software radio adopts a modular design, which must be a communication platform with openness, scalability and compatibility, and is made into the form of a bus with a modular standard. Based on this relatively common hardware platform, we implement different communication functions by loading different software (the card can be replaced when needed). The hardware platform of software radio is much more demanding than PC, it needs wideband radio frequency front-end, wideband A/D/A converter, high-speed DSP devices and so on. In order to perform high-speed A/D/A conversion and digital signal processing, software radio systems must work in parallel with multiple CPUs. In addition, for digital signal processing data to be exchanged at high speed, the system bus must have a very high T/O transmission rate. Among the current system buses that meet the requirements, the VME bus has the most mature technology, the best versatility, and the most extensive support. VME provides multiple CPU parallel processing, supports independent 32-bit data bus and address bus, and the rate reaches 40Mb/s (or even 320Mb/s), which basically meets the requirements of software radio and is the preferred bus method for software radio. Three, the key technology of software radio
1. Multi-band down-conversion and broadband RF
For the antenna of the software radio system, it should have a multi-band antenna and a programmable radio frequency conversion function. On the basis of satisfying the antenna gain, physical size and price, it should have a working bandwidth of 2MHz-3MHz. In radio engineering, it is not necessary to cover the full frequency band, but only need to cover several windows of different frequency bands. Therefore, a combined multi-band antenna can be used. The US military's speakeasy is a solution that uses multiple sets of RF antennas. For broadband RF, tuning, energy control, and low-noise preamplifier (LNA) configuration is also a key technology, and computer-aided design (CAD) can be used to optimize the design of the system.
2. Broadband A/D part.
The key to determining the performance of wideband analog-to-digital conversion is sampling and the number of bits. The sampling rate is determined by the signal bandwidth, while the number of quantization bits requires a certain dynamic range and DSP precision. Because the existing single-chip ADC cannot meet these two requirements, multiple ADCs can be used in parallel.
3. High-speed parallel DSP part.
In the digital processing operation of the system, the most difficult is up-conversion, filtering and sub-sampling. The high-speed parallel DSP includes digital baseband processing, modulation and demodulation, bit stream processing and decoding functions. For FM and spread spectrum systems, this part should also have despreading and dehopping functions. To achieve this part of the function, it is necessary to use high-speed parallel DSP to form a multi-processor parallel computing system, including more multiple access calls, wider program bus and data bus, single instruction multiple data, multiple instruction multiple data. The structure and the use of super-instruction structure, etc., this part can be realized by a dedicated digital integrated circuit chip ASIC (for example, the DDC chip HS P50016 of Harris Corporation of the United States).
4. Leaving the general route structure of openness and scalability.
In the traditional system structure, the pipeline is generally used, which is characterized in that each functional unit is connected by a circuit. If the function of a certain part is to be added, deleted or modified, the corresponding functional module must be adjusted. Therefore, this structure Not open. In order to realize the interconnection of various functional units in the system, an open and extensible hardware platform is formed, and at the same time, it has a high data throughput rate. The software radio system must adopt a new interconnection structure, which is characterized by relatively simple implementation and can directly apply a variety of bus standards (such as VME, bus, PCI bus, etc.). , bus-based interconnect structure.
5. Software protocols and standards.
Since the mid-to-late 1990s, foreign countries have been studying how to implement software plug and play (Plug & Play), and have proposed based on it. JAVA/CORBA software protocols and standards. The idea based on "software bus" is to establish a standard-based, open and easy-to-use architecture. The so-called "software bus" is similar to the "hardware bus" that is often said. The application module is made into a bus according to the standard, and the combined operation can be realized by inserting the bus, thereby supporting a distributed computing environment. This design idea is consistent with the reusability of software in software systems.
6. Power consumption, volume and cost of the system.
This is the key to commercialization of software radio, and its solution largely depends on the development of hardware technology. Fourth, the development and prospect of software radio
Since the 1990s, with the rapid development of various wireless communication systems, the differences in radio communication standards, and the advancement of digital signal processing technology, software radio technology has attracted more and more attention, and it is expected to become a future global communication network. new system.
According to the ideal structure, all signal processing tasks of the software radio station from RF to baseband are carried out in full digital form, so it is completely programmable, and its structure is also reconfigurable and reproducible. However, because there is no A/D converter that can be applied to the radio frequency band, another topic that is being researched now is the digital RF front end of the software radio station, which is the key to the digitization of the whole frequency band.
Existing digital signal processing devices (DSPs) have been widely used for signal processing in parts such as IF, baseband or terminal, which has brought the technical performance of radio equipment to a new and modern level, but its RF front-end is still narrowband. For the software radio station, the A/D converter in its RF front end must be able to handle the entire communication frequency band, generally from 2MHz to 3GHz. In addition, the typical characteristics of mobile communication signals are fading and shielding, and there may be strong blocking and interference. As a result, the dynamic range of mobile communication signals appearing at the receiving RF end is as high as 100dB or more. If considering different standards of mobile communication signals , its dynamic range will be greater. For a system with a bandwidth of 10MHz, the sampling frequency is greater than 25MHz, which requires 2500MIPS of computing processing capability, which is far from meeting the signal environment requirements to be processed by the RF front-end. Even with A/D converters capable of meeting bandwidth and dynamic range requirements, their power requirements can still hinder the use of mobile terminals. The software radio station developed at the present stage cannot adopt the ideal structure of full frequency band digitization, but adopts the practical structure of partial frequency band digitization of Rf front end.
To sum up, software radio contains two meanings: one is the radio frequency (RF) front end, and the other is digital signal processing; its core components are broadband A/D/A converters and high-speed DSP chips. The biggest advantage of software radio is that it can complete various signal processing tasks on the hardware platform by defining various working parameters and reorganizing the channel structure according to the wireless band and channel access method. Therefore, a dedicated digital front end for each working standard can be designed on a common hardware platform, or the commonality of different working standards can be utilized. The former design not only achieves the greatest degree of freedom, but also minimizes the number of gates used, while the latter design requires the development of dedicated algorithms for digital front-end functions, but can be implemented with ASICs, allowing us to Take advantage of the concept of software radio.
In general, although software radios were originally developed for military short-wave over-the-horizon communications, because they offer a high level of fidelity not found in analog receivers, they also offer excellent flexibility. With a little modification, it can adapt to and meet the requirements of different users, and its cost is very low. Coupled with the rapid development of digital signal processing technology, it has been increasingly used in the field of civil communication, especially in mobile communication. The application in the system is more extensive.
