Main / Radio
“The single biggest problem in communication is the illusion that it has taken place.” ― George Bernard Shaw Glossary
This is an interesting website for a guy that likes to teach wireless comms via pictures: https://wirelesspi.com/ System Components
AntennaeThe dipole is the common simple antenna with two equal length conductors (like rabbit ears). The electrical signal is connected to the feedline between the two rods, or received there. The radiation pattern is maximum perpendicular to the conductor. If the rods are aimed vertically then you get an "omnidirectional" antenna - not up and down but 360 degrees around. The Yagi antenna (Yagi-Uda) is a directional "beam antenna" or "parasitic array" used for high gain in HF, VHF, UHF bands usually for fixed-frequency applications. Radiation pattern is unidirectional along the primary axis. Generally speaking, the more elements an antenna has, the narrower the beam will be (more directional) and the higher the gain on the signal. With electronically controllable phase and gain you can steer the main beam in different directions without moving the apparatus. Exactly the same thing for receive. This is called "phased array". https://www.tutorialspoint.com/antenna_theory/index.htm EM SpectrumRadio waves take up the 3 Hz to 300 GHz range of the electromagnetic spectrum. This is the lowest frequency category of EM waves. Most common bands:
The reason for the divisions by powers of 3 is that the bands are based upon powers of 10 for wavelengths. S-band (also UHF), X-band, K-band, and others are IEEE designations of pieces of the SHF band. The 420 to 450 MHz range is allocated for amateur radio/satellite (ham) use, called the "70cm band". Proximity-1 Space Link Protocol used for lander/orbiter comms is in this band. FM and ATC radio are VHF. AM is MF. Sampling and ProcessingA filter's bandwidth increases proportionally with the signal frequency, so there's an advantage to converting to a lower frequency for filtering. This is one purpose of an intermediate frequency (or processing frequency). Low sampling rates of an ADC could require down-conversion to a lower IF. It's also easier to amplify a lower frequency signal. The superheterodyne circuit design in combining an oscillator with the input signal for down-conversion to an IF is common in modern radios. Put another way heterodyne or superheterodyne refer to an RF receiver that uses an intermediate frequency. A baseband PLL is used to generate all baseband related clock signals, like ADC/DAC sampling clocks, the data clock, and data framing. Modulation and KeyingModulation is merely multiplication, and results in shifting a signal to higher frequencies. AM is the original way of modulation for radio broadcasts, and along with FM and PM these are examples of analog modulation. FM and PM are called angle modulation, so between angle and amplitude you have the two major groups of analog modulation. Angle modulation can perform better against interference and noise. PM is the basis for WiFi, cellular GSM, and satellite TV signals. PSK, FSK, ASK, and all those are digital modulation schemes. When digital signals are used to modify the carrier wave, this is called keying rather than modulation. QPSK offers double the throughput of BPSK because of the four symbols instead of two (need to research how that works) but the tradeoff is QPSK has more possible phases: What this means with I and Q data is that when you've done the BPSK demodulation, you can actually throw away the Q data, whereas with QPSK you'll still need both to finish the process. The hint is in the name (QPSK = quadrature PSK). What is a phase ambiguity anyway? WIP case PHASE_AMBIGUITY is when "00" => -- 0 deg DATA20_I <= DATA19_I; DATA20_Q <= DATA19_Q; when "01" => -- +90 deg DATA20_I <= not DATA19_Q; DATA20_Q <= DATA19_I; when "10" => -- +180 deg DATA20_I <= not DATA19_I; DATA20_Q <= not DATA19_Q; when others => -- +270 deg DATA20_I <= DATA19_Q; DATA20_Q <= not DATA19_I; end case; I and QOrthogonality or quadrature are terms that refer to the use of two carrier waves of the same frequency that are 90 degrees out of phase. Raw RF data is often called IQ data and consists of a real and imaginary component. I is for in-phase, not imaginary, and Q is for quadrature. I and Q refer to the two carrier waves, in other words one is I and one is Q. I lags Q by 1/4 cycle, or 90 degrees, and that difference may explain the name quadrature. I/Q data is a complete representation of how a carrier is modulated: amplitude, phase and frequency. Given complex values of the signal, the I is real, and Q is imaginary. A real-valued signal is just a complex signal where all the imaginary components of all the complex values are strictly zero. Real valued signals have one degree of freedom. Complex signals are often used to represent signals or data with 2 degrees of freedom. Channel Band TermsbasebandA baseband signal is a lowpass signal with low frequencies very near zero. The highest frequency is the cutoff or upper bound of the signal or system. A baseband channel is a comm channel that can operate at very low frequencies, like serial cables or LAN. In some cases this could be called the information, as in the audio of an AM baseband signal. It's the signal base, as in it doesn't get demodulated (reduced) any further. passbandRange of frequencies that can pass through a filter, or in telecomm systems it's the portion of spectrum that is transmitted. A bandpass filter removes anything outside the passband. sidebandFrequencies higher or lower than the carrier that exist due to modulation and hold the signal information. Holds all the spectral components of the signal besides the carrier. All forms of modulation produce sidebands. USB = upper sideband, LSB = lower sideband IFAn intermediate frequency is between the carrier (high freq) and baseband (low freq) frequencies. Lowers the cost and complexity of processing for a wide range of carrier frequencies. near-zeroRelatively very low baseband frequency, like audio in kHz as compared to high mHz carrier frequencies. half-band filterwidely used for their efficiency in multi-rate applications. A half-band filter is a low-pass filter that reduces the maximum bandwidth of sampled data by a factor of 2 (one octave). When multiple octaves of reduction are needed, a cascade of half-band filters is common. And when the goal is downsampling, each half-band filter needs to compute only half as many output samples as input samples. ACLR = adjacent channel leakage ratio ApplicationsAntenna size increases inversely to the frequency, so the lowest frequencies require the largest antennas. They also have the least bandwidth. Anything below the range of tens of kHz (10 kHz) is rare, data slow, and expensive. This is the ELF (extremely low frequency) range, used for one example with submarines. Submersion in sea water weakens radio communications, but VLF and ELF signals can penetrate further. Low frequency, long-wavelength signals can penetrate better than high frequency signals. ELF transmissions require large amounts of input power, and have low output power. But due to the strength (low attenuation) of the output, the small amount of output power can cover tremendous distances. Satellite TV dish antennas use the 950-1450 MHz range. Traditional TV antennas are in the 50 to 870 MHz range. A diplexer is often used to allow both, which functions as both a lowpass and highpass filter onto a single channel. AM radio is in the MF band, FM radio in the VHF band. Television occupies VHF and UHF bands. Satellite radio is in the UHF or SHF band, as are mobile phones, wifi, Bluetooth, etc. SDRA digital down-converter (DDC) takes as input a quantized RF signal from an ADC and uses a LPF + downsampler to get a baseband signal. This is typically implemented in HW, an FPGA or ASIC. Specs and MeasurementsNote that antennae and RF receivers might have different bandwidth specs. It generally follows to have an RF receiver that covers at least the antenna bandwidth. Antenna gain is a key spec that is a combined measure of directivity (how radiation is concentrated directionally) and electrical efficiency (proportion of wave power delivered as electric signal). Gain is a comparison against a hypothetical lossless isotropic antenna, expressed as dBi. dBc dBi dBd dBiC dBFS dBm C/No = ratio of carrier power to the noise power density in dB-Hz; it determines whether a receiver can lock on to the carrier and if the information encoded in the signal can be retrieved, given the amount of noise present in the received signal. Signal Power Ratio Equivalence Chart
Digital Comm LinksOrder of operations: NCO = A numerically-controlled oscillator (NCO) is a digital signal generator which creates a synchronous (i.e. clocked), discrete-time, discrete-valued representation of a waveform, usually sinusoidal. Direct Sequence Spread SpectrumDSSS technology breaks down the transmitted stream of data into small pieces across a frequency channel. A redundant bit pattern (known as a chipping code) is generated for each bit transmitted. Generally, the longer the chipping code, the more likely it is that the original transmitted data will be properly received. The process in demodulating is that "code acquisition" happens before "code tracking". The former is expensive and involves matching the reference signal to the received signal, and the "code" part is the DSSS part that does the signal spreading. What about the generator polynomials? Those can be fixed and won't change dynamically. Almost like choosing an encryption key because it defines the spread spectrum, and can't lock onto the carrier without having the right one. Polynomial is key to the de-spreading, and you have to de-spread to get the carrier lock. Soft BitsSoft bits or more generally speaking, soft metrics, are usually taken at the output of the demodulator. A hard bit (AKA the transmitted data bit) is obtained by performing a hard decision on a soft metric. You usually have multiple soft bits that would represent one hard bit, and the soft bits express a probability as to the likely value of the hard bit. For example, a demodulator may have a 4-bit or 8-bit output and each set of 4 or 8 bits is actually a representation of one hard bit. For an unsigned output in the 4-bit case, you could say that if your output is 0-7 your corresponding data bit is probably 0 but if your output is 8-15 your corresponding data bit is probably 1. The BER then comes from comparing the hard sent data to the decision output here. Wi-FiWhat is the relation between number of WiFi antenna and number of spatial streams? https://superuser.com/questions/791481/what-is-the-relation-between-number-of-wifi-antenna-and-number-of-spatial-stream You need at least as much antennae as you have streams. Besides, even with single stream more antennae can help in case of poor signal/noise ratio. DAC and ADCA loopback sending a signal from a DAC to an ADC for testing is not necessarily always the best test method because of various non-idealities (like finite quantization) introduced by both sides. Sometimes it's better to use some other independent unit to generate the signal. PEP = peak envelope power, a way to adjust the power setting for your signal you generate from the DAC; when this is reduced, the intermod products also go down (at a faster rate than the signal power, so it's not a linear relationship) Interleave offset and gain are related to a specific architecture in which multiple converter components are combined to get higher sample rates. These values can be predicted mathematically. A spurious element is one that is not harmonically related to the signal, meaning it won't necessarily go away if the signal does. Occurs in a band or at a frequency unassociated with expected harmonics or intermod products. Derived from the quirks of the system components themselves, rather than the signal. TIS = time interleaving spurs, OIS = offset interleaving spurs You never want to overload the converter with too much gain from your AGC, and could be a reason to include an AGC. THD = total harmonic distortion, which is a sum of harmonics power Total dynamic range and high dynamic range are a little different in that the former represents the strongest signal to receive which the latter indicates ability to receive both strongest and weakest signals. Intermod ProductsThe result of nonlinear behavior of an amplifier. These are kind of similar to harmonics, which can come from the same root causes, but intermod refers specifically to multiple signals. Harmonics can be present with a single tone signal. What about different orders of intermod products? The order of the product is given by the sum m+n where the mixing rule is (m*F1 +/- n*F2) with input signal frequencies F1 and F2. Therefore a second order product is when m+n=2 or F1+F2, 2*F1, etc. A third order product could be 2*F1+F2, for example. And so on. Although this goes on infinitely, each successively higher order product is reduced in power. The second, third, and fifth orders are generally the only ones of significance, and often the third order is used for receiver data sheets because some of those fall close to the signal frequencies. For example, with signals of 10mhz and 15mhz, 10+10-15=5mhz and 15+15-10=20mhz. Second order products are further away. Fourth order products are largely ignored because they are the same as the signal or very far away. The height of these elements is a sign of how well the device performs... higher is worse and ideally there would be none. The baluns don't affect these. DPD = digital pre-distortion Code RateGenerally the "code rate" or "information rate" is just the ratio of information bits over transmitted bits. In other words, the proportion of the data-stream that is useful. With FEC, you'd have some redundant bits which take up useful information bits. Expressed as 1/n meaning one bit of useful data for every n bits transmitted. Does that mean the highest possible theoretical code rate is 1/1, or simply 1, in which every transmitted bit is a data bit and no error control is done? Local Oscillator (LO)The output of the RF stage is one input of a mixer. A Local Oscillator (LO) is the other input. The output of the mixer is at the Intermediate Frequency (IF). Parts DiveThe AD9364 series of transceivers:RF -> AD9364 -> baseband processor What are the RX components? There is also AGC and channel independent RSSI, and decimation factors produce the desired output data rate. The DATA_CLK is for RX data (output by the PHY), but the FB_CLK is for TX data (input to the PHY). The FB_CLK signal must have the same frequency and duty cycle as DATA_CLK. Phase Shifterhttps://www.rfmw.com/data/telemakus_teleapp_003.pdf Phase Shifters are a critical component in many RF and Microwave systems. Applications include controlling the relative phase of each element in a phase array antenna in a RADAR or steerable communications link and in cancelation loops used in high linearity amplifiers. The phase of a single tone is meaningless until it is compared to another signal. So phase is therefore a relative measurement. The circuit for a phase shifter can be achieved using varactor diodes (also called a tuning or varicap diode). A varactor diode behaves like a voltage variable capacitor when it is reversed biased. As the voltage is increased the capacitance deceases and with it the phase of the reflection Single-Ball Signal GeolocationDigitally modulated signals generally appear as noise. To detect them, we use a cross-ambiguity function (CAF) which is a cross-correlation. CAF processor cannot simultaneously optimize for multiple things, like wideband TOA and FOA detection. Instead, we break the process into steps, and start with narrower bands. Radio RX -> Channelizer (is this the same as sub-band tuning?) -> Detection CAF -> Measurement CAF -> Metadata extraction -> Linker -> Geolocation The Detection is 95"% of the work, with all the rest being smaller portions. Wideband, short-pulsed signals, like traditional RADAR, support good TOA measurement accuracy and poor FOA accuracy. Narrow band, long signals, like comms, support good FOA measurements and poor TOA accuracy. Garage Remote ConceptThere are some pretty small form factor fingerprint modules: Fingerprint Sensor Modules ? Mouser, if we want to go that route. I really like the NFC idea. Allows the leveraging of highly developed fingerprint/face id security infrastructure on smartphones, and improves segmentation of the transmitter functions. NFC antennas are really easy and off-the-shelf ICs are low-power and flexible. Not sure how much overhead there is in garage door transmitter protocols, but if enough can be offloaded to a mobile app, potentially the entire embedded solution could run off an NFC IC. They usually have an ARM core built in. This would be extremely low power and good battery life. Essentially the battery would just power the transmitter. Microcontroller can be powered entirely by NFC when present. In the past I used this NXP part for a ?smart label? for Johnson&Johnson that logged temperature of the drug vial it was adhered to: NTAG SmartSensor temperature monitor (nxp.com). There is surprisingly a lot in there for such a small part. It has an ARM M0+ core A quick search found that some transmitters are OOK/ASK to send 12-bit codes. Found a couple that use M0 MCUs. This FCC test report seems to indicate that liftmaster is using OOK? 7964 FCC Part 15 Remote Control Transmitter Test Report EMC Chamberlain Group The (fccid.io). Microchip has some application specific parts for this: MICRF114 Low-Power Integrated Sub-GHz Wireless RF Transmitter Data Sheet (microchip.com) https://ww1.microchip.com/downloads/en/DeviceDoc/50002416A.pdf https://fccid.io/HBW7964/Test-Report/Test-Report-1853266 https://www.nxp.com/docs/en/data-sheet/NHS3100.pdf https://www.mouser.com/c/embedded-solutions/sensor-modules/fingerprint-sensor-modules/ Units manufactured by Chamberlain from 2011-Present use ?Security+ 2.0 (rolling code) on 310, 315, and 390 MHz? Found a Github repo from someone that was able to replicate the system. GitHub - argilo/secplus: A software implementation of the Security+ system used by garage door openers Check the Protocol Details section. It details the Security+ 2.0 protocol. There is also a link in that section to the US Patent, which outlines implementation details as well. Units manufactured by Overhead Door from 2011-Present use ?Intellicode 2/CodeDodger 2 (rolling code) on 315 and 390 MHz?. |