What is EVM (Error Vector Magnitude)?

In the world of RF communications, when information travels wirelessly using radio waves, maintaining a clear and undistorted signal is absolutely crucial. EVM, or Error Vector Magnitude, steps up as a simple yet powerful metric to assess this very aspect.

Imagine a map: Picture a constellation diagram, a map for digitally modulated signals. These signals are essentially represented by points scattered across this map, with each point's location determined by its in-phase (I) and quadrature (Q) components. Ideally, all the points would occupy specific designated locations on the map.

Question comes to mind – what is Constellation diagram?

A constellation diagram is a visual representation in the complex plane of the discrete signal points used in a digital modulation scheme.

In the real world, things aren't perfect. Signal imperfections like noise, distortion, and phase noise act like roadblocks, causing the actual points to deviate from their ideal positions on the constellation map.

This is where EVM comes in. It acts like a surveyor, meticulously measuring the distance between each received symbol's location and its closest ideal spot on the map. By calculating the root mean square (rms) of all these distances, EVM provides a single, comprehensive value that reflects the overall amount of signal degradation in the system.

As per 3gpp 38.101

EVM is typically expressed as a percentage, it can also be represented in decibels (dB)

EVM (dB) = 20 * log10 (EVM (%))

Note:

• Lower dB means better quality (weaker errors).
• Higher dB means worse quality (stronger errors).

In the above the question comes to mind What is RMS?

Root mean square (RMS) is a way to find a single value that represents the average strength of a set of numbers that have positive and negative values, or a fluctuating signal.

Imagine a rollercoaster ride - the average height doesn't tell the whole story, but RMS captures the overall "intensity" by considering both the ups and downs.

There's a close link between EVM and another crucial metric: Bit Error Rate (BER). BER essentially tells you how often errors occur in the transmitted data bits.

• When symbols land far from their ideal locations on the constellation map (high EVM), the chance of them straying into the decision boundary of another symbol's designated area increases.
• This "trespassing" is interpreted as a bit error, leading to a higher BER

Difference between BER and EVM

It's important to differentiate between EVM and BER. BER is calculated based on the original bit pattern that was transmitted. EVM, on the other hand, focuses on the received symbol's location relative to its ideal position on the constellation map.

Let’s understand it by simple example –

BER tells you how many wrong turns were made on the data delivery route, while EVM reflects how far off course the delivery vehicle strayed at various points along the journey.

Reason for bad EVM:

• Noise: Additive White Gaussian Noise (AWGN)
• Distortion: Non-linear distortion (IMD, harmonic)
• Phase noise: Fluctuations in carrier signal phase
• Hardware imperfections: Limitations in components like amplifiers, filters
• Synchronization issues: Symbol timing errors, carrier frequency offset.
• Environmental factors: Multipath propagation, fading.
• System design choices: Inappropriate modulation/coding schemes

-Ravi

#EVM #RMS #errorvectormagnitude #constillation #wireless