KF6HI --- You can contact me through the following methods:kf6hi@arrl.net
Signal-to-noise ratio in radio, definition and applications of radio communication
You and your partner go out for breakfast. Another couple is in the restaurant. The conversation between you two flows easily. By the time dessert arrives, the restaurant and bar area are almost full. The conversation becomes much more difficult. You find yourself shouting at your partner to be understood. What's happening is that the background (ambient) noise level has increased. To maintain communication, you must compensate by increasing the volume (more signal power) in order to keep the speech level of the conversation at a level that can overcome the increased noise level (more noise power). This is similar to radio. (RF) The situation is the same around the world, in the field of radio frequency. (RF) Worldwide, the signal is your radio signal, and noise is atmospheric, electrical, or other interfering radio signals.
Signal-to-noise ratio (SNR) It is a quality factor that compares the level of the desired signal with the background noise level. It is calculated as the ratio of signal power to noise power. SNR is expressed in dB (decibels).
SNR (dB) = 10 * LOG [Signal power / Noise power, both in watts]
For strong signals, you can estimate SNR using the S table. If the received CW signal is S9 and the meter reads S7 without a signal, then the SNR is 12 dB (typically each S unit represents 6 dB). Assuming this signal is received through a 2400 Hz wide receiver filter bandwidth. Now apply a narrow 500 Hz CW filter. The noise reading should drop below S6, improving by 6.8 dB. This is the ratio of the two bandwidths. A narrower bandwidth filters out more noise, thereby increasing your SNR. The SNR is now 18.8 dB. However, it's important to note that the desired signal must still pass completely through the filter. Passing a nominal 2300 Hz wide SSB signal through a 500 Hz filter will make the signal difficult to understand.
10 * LOG[2400 Hz / 500 Hz] = 6.81 dB
Noise permeates the entire radio spectrum. It can be caused by external noise (RF signals) received by the antenna. < (Common scenarios for 30 MHz and HF frequencies) or internal noise added by the receiving system (signal >(Common scenarios for 30 MHz, VHF, and UHF) are dominant.
From Figure 1, you can see that noise is a broadband signal. To measure its power, we must specify the bandwidth over which it is measured, which is called the noise bandwidth. (NBW)。
Figure 1: Comparison of different SNRs. The power of the signal (red) is compared to the noise power in the received bandwidth (blue).
SSB signals are typically received through filters of 2600 Hz or 2400 Hz. Within the amateur radio community, 2500 Hz has been used as a benchmark for assessing receiver performance. The "modes" I'm referring to include SSB, CW, RTTY, JT65, and others. Each mode has its own advantages and disadvantages. For example, SSB is a signal with a bandwidth of 2300 Hz. If you use a narrower filter, you not only reduce noise but also start losing signal energy. This loss of signal energy will quickly make the SSB signal difficult to understand. Bandwidth is a limiting factor for SSB modes.
CW is a much narrower signal. Common filters include 2400 Hz, 500 Hz, 250 Hz, and 100 Hz. Using a DSP (Digital Signal Processor), you can even narrow the range further.
Essentially, this is what all new DSP modulation schemes are doing. They are using DSP to narrow the noise bandwidth, making it as close as possible to the signal. In addition, they are also encoding the signal. Encoding means that they are adding redundant information and/or reducing the alphabet or known results of the decoded signal *1. Encoding improves the data recovery of the signal at a given SNR. The range of encoded signals is outside the scope of this discussion.
DXers, who focus on HF amateur communication over long distances, typically deal with weak signals. Different modulation modes have different performance characteristics for weak signals. Let's look at some common modes used for DXing. In the amateur radio literature, SNR (Signal-to-Noise Ratio) is often expressed in dB relative to the noise within a 2500 Hz bandwidth, which represents the ability to hear and decode weak signals. This is what we show in Figure 2.
Figure 2: SNR Performance Reference for a Randomly Replicated SSB Signal at 2500 Hz
As shown in Figure 2, temporary SSB communication requires a 6dB SNR. A nearly unintelligible SSB signal (identifying your callsign and obtaining the signal report) would be 0 dB SNR. Please note that when we perform these SNR comparisons, it is on a clear channel. There is no QRM, no static interference, and no fading. To compensate for these channel characteristics, we need a higher SNR.
Refer to Figure 2, we can see that for casual CW operation, the signal-to-noise ratio (SNR) can be as low as -11 dB or -5 dB, while maintaining the same clarity as casual SSB. But what does -5 dB mean? Does this mean the signal is below the noise floor? How can you hear it! Remember, this is relative to a 2500 Hz bandwidth. You would typically use a 500 Hz filter, and now -5dB becomes 1.8dB. This is sufficient for CW. Is it too noisy? Click on a 250 Hz filter, now SNR = 4.8 dB.
In Figure 3, we re-calculated the data using a 100Hz filter. The CW signal with -5 dB SNR at 2500 Hz now has a 9 dB SNR. For excellent DXers who want to confirm their callsign and signal reports, they can reduce it to -1 dB SNR (allowing the brain to perform its own DSP on the signal). Furthermore, when you are simply trying to find your callsign and receive the return signal report, you are actually reducing the alphabet – and gaining some decoding advantage. For a discussion of which CW SNRs are reproducible, see Ray Soifer's "The Weak Signal Capabilities of the Human Ear" written for W2RS.
Figure 3: SNR performance of a reference 100 Hz minimum-repetition-rate CW signal.
To achieve better SNR performance, you need to use one of the DSP modes. The JT65 was originally developed for EME communication with very weak signals. The JT65A has been adapted for use with JT65-HF. As shown in Figure 2, if your JT65A SNR report is -15 or lower, the signal you are receiving is too weak for traditional CW reception. By using the JT65, you can increase the signal reception by 9 dB. (24-15)Please note that when using this mode, it will never report signals greater than -1 dB. The software limits the reporting of such strong signals. This can be traced back to the original purpose of JT65: EME (Earth-based Mobile). The DSP code was not designed to handle positive SNR values. According to my experience, any reported SNR value of -10 or higher is suspicious.
WSPR 4 is not a two-way communication mode, but rather a beacon mode. The narrow DSP filtering and encoding increase the reception of weak signals by 8 dB.
JT9 was developed after JT65 5. It is similar to JT65, but narrower and with better performance. Unlike JT65, it is specifically designed for high-frequency operation. It will report a positive SNR.
FT8 is a relatively new mode developed to reduce signal degradation during Sporadic-E communication over ranges up to 6 meters. However, it has been widely adopted by the amateur radio community in the HF and VHF spectrum.
Please note:
*Think about the time you spent learning Morse code. You probably started with letters, then numbers. When you're just testing on letters, it's easier because the alphabet is reduced without the numbers. You don't have to worry about confusing B with 6.
References:
1 EME using JT65
2 JT65-HF
3 JT65A maximum SNR -1 dB
4 WSPR
5 JT9
ACRONYMS
SSB = Single-Sideband, narrowband voice communication mode
CW = Continuous Wave, a Morse code communication mode.
HF = High-frequency radio spectrum, from 3 to 30 MHz
VHF = Very High Frequency radio spectrum, from 30 to 300 MHz
UHF = Ultra High Frequency radio spectrum, from 300 MHz to 3 GHz
RTTY = Radio Telegraphy mode of communication
闽ICP备2021006864号-7
闽公网安备35020602002794号